Front Cover
 Title Page
 Table of Contents
 Chapter 1: Rural migration and...
 Table of contents
 Introduction and summary
 Migration: Some estimates of its...
 Migration and "labour shortage...
 Remittances and agricultural...
 Some special considerations of...
 Migration and rural poverty
 Migration, land tenure, village...
 Conclusions and policy recomme...
 Chapter 2: Agricultural mechanization...
 Table of contents
 Historical background to the mechanization...
 The mechanization of water...
 Mechanization of crop producti...
 Mechanization of post-harvest...
 The relative impact of agricultural...
 Interrelations between land tenure,...
 The influence of government pricing...
 The influence of multinational...
 The institutional back-up for agricultural...
 Chapter 3: Rural women and the...
 Table of contents
 Progress during the last ten...
 The continuing crisis in agric...
 Women farmers and important...
 Constraints to increasing women's...
 Conclusions and recommendation...
 Chapter 4: Organization and management...
 Table of contents
 The small farmer
 Institutional set-up for providing...
 Production inputs and services...
 Agricultural extension workers
 Farmers' organizations
 Conclusions and suggestions
 Back Cover

Title: Selected issues in rural development in the Near East
Full Citation
Permanent Link: http://ufdc.ufl.edu/UF00084629/00001
 Material Information
Title: Selected issues in rural development in the Near East
Alternate Title: WCARRD, ten years of follow-up, selected issues in rural development in the Near East
Physical Description: v, 251 p. : ; 30 cm.
Language: English
Creator: Richards, Alan, 1946-
Food and Agriculture Organization of the United Nations
Conference: World Conference on Agrarian Reform and Rural Development, (1979
Publisher: Food and Agriculture Organization of the United Nations
Place of Publication: Rome
Publication Date: 1989
Subject: Rural development -- Middle East   ( lcsh )
Migration -- Near East   ( ltcsh )
Technology and technological change, Agricultural -- Near East   ( ltcsh )
Labor and laborers, Agricultural -- Near East   ( ltcsh )
Agricultural development -- Near East   ( ltcsh )
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
Spatial Coverage: Egypt
Bibliography: Includes bibliographical references.
Statement of Responsibility: by Alan Richards ... et al..
General Note: At head of cover title: WCARRD World Conference on Agrarian Reform and Rural Development, ten years of follow-up.
 Record Information
Bibliographic ID: UF00084629
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 30135955

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
        Page v
    Chapter 1: Rural migration and its impact on agricultural production and rural development in the Near East
        Page 1
        Page 2
    Table of contents
        Page 3
        Page 4
    Introduction and summary
        Page 5
        Page 6
    Migration: Some estimates of its magnitude
        Page 7
        Page 8
        Page 9
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
    Migration and "labour shortages"
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
    Remittances and agricultural development
        Page 35
        Page 36
        Page 37
        Page 38
    Some special considerations of rural-urban migration
        Page 39
        Page 40
    Migration and rural poverty
        Page 41
    Migration, land tenure, village poverty
        Page 42
        Page 43
    Conclusions and policy recommendations
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
    Chapter 2: Agricultural mechanization in the Near East: Its output and employment impacts and related policy issues
        Page 65
        Page 66
    Table of contents
        Page 67
        Page 68
        Page 69
        Page 70
    Historical background to the mechanization of Near East agriculture
        Page 71
        Page 72
        Page 73
    The mechanization of water adduction
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
    Mechanization of crop production
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
    Mechanization of post-harvest operations
        Page 88
    The relative impact of agricultural mechanization on male and female agricultural labour
        Page 89
        Page 90
    Interrelations between land tenure, mechanization, and labour supply
        Page 91
        Page 92
    The influence of government pricing policy on agricultural mechanization in the Near East region
        Page 93
    The influence of multinational corporations and foreign aid agencies on the mechanization process
        Page 94
        Page 95
    The institutional back-up for agricultural mechanization
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
    Chapter 3: Rural women and the changing socio-economic conditions in the Near East
        Page 117
        Page 118
    Table of contents
        Page 119
        Page 120
        Page 121
        Page 122
    Progress during the last ten years
        Page 123
        Page 124
        Page 125
        Page 126
    The continuing crisis in agriculture
        Page 127
        Page 128
    Women farmers and important farmers
        Page 129
        Page 130
        Page 131
        Page 132
        Page 133
    Constraints to increasing women's agricultural production
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
    Conclusions and recommendations
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
        Page 206
    Chapter 4: Organization and management of agricultural services for small farmers in the Near East
        Page 207
        Page 208
    Table of contents
        Page 209
        Page 210
        Page 211 (MULTIPLE)
    The small farmer
        Page 212
        Page 213
        Page 214
        Page 215
        Page 216
    Institutional set-up for providing inputs and agricultural services to small farmers
        Page 217
        Page 218
        Page 219
        Page 220
        Page 221
    Production inputs and services needed by small farmers
        Page 222
        Page 223
        Page 224
        Page 225
        Page 226
        Page 227
        Page 228
        Page 229
        Page 230
        Page 231
        Page 232
    Agricultural extension workers
        Page 233
        Page 234
        Page 235
        Page 236
        Page 237
        Page 238
        Page 239
    Farmers' organizations
        Page 240
        Page 241
        Page 242
        Page 243
        Page 244
        Page 245
        Page 246
    Conclusions and suggestions
        Page 247
        Page 248
        Page 249
        Page 250
        Page 251
    Back Cover
        Page 252
Full Text

r .

World Conference on Agrarian Reform and Rural Development




Food and Agriculture Organization of the United Nations

Rome 1989






by Alan Richards

by Diana Hunt

by Virginia DeLancey and Elweya Elwy

by Salah Wazzan



The four papers in this publication were initially prepared for
discussion at the Round Table of Experts on Selected Issues in Rural
Development in the Near East which was held in Amman, Jordan, from
14-18 May 1989 in commemoration of the Tenth Anniversary of the World
Conference on Agrarian Reform and Rural Development (WCARRD). The
purpose of the Round Table was to elicit the experts' opinions on
these topics of priority concern in the region which were to be
discussed subsequently by the Government Consultation for the Near
East on the Follow-up to WCARRD, held in Rabat, Morocco, from 3-7 July
1989. The experts' opinions and recommendations were accordingly
presented to the Government Consultation to further substantiate and
enrich the discussions of these important issues.

Since these papers have evoked considerable interest, it was
decided to issue them as a separate publication in order to ensure
wider circulation and discussion of the issues. The views expressed
in the papers are those of the authors and do not necessarily reflect
FAO's position. The reports of the Round Table and Government
Consultation are also available in FAO's WCARRD Tenth Anniversary




Alan Richards
Department of Economics
University of California
Santa Cruz


Table of Contents




A. Causes 7
B. International Migration: to the EC 8
C. International Migration: to the Oil-Exporting
Countries 8
D. Rural-Urban Migration 13
E. Rural-Rural Migration 16
F. Selectivity of Migration 17


A. Analytical Framework and Hypotheses 18
B. Evidence of Sector-Wide Labour Shortages 19
C. Evidence of Shortages of Specific Types of
Labour: Extent of Substitutability 22
D. Production Consequences of Increased Labour Costs 29
E. Mechanization: A Response to Labour Shortages? 32


A. Macro-Economic Impacts 36
B. Micro-Economic Consequences: How are Remittances
Spent? 36
C. Micro-Economic Consequences of Remittance Spending 38










Labour mobility from rural areas has profoundly affected
agricultural development in the Near East. Migration from one rural
area to another, from rural areas to cities, and from rural areas to
other countries has had diverse effects on agricultural labour supply,
on investment and technological change, and on the pattern and rate of
growth of demand for food. This paper will provide an overview of
hypotheses and policy issues concerning the impact upon agricultural
development of international, rural-urban, and rural-rural migration.

There have been two major waves of international migration, as
well as a steady flow of rural to urban migration within nations. The
first wave consisted of the emigration of rural North Africans and
Turks to the countries of the European Community (EC). In the Algerian
case, these flows are several generations old, dating back to at least
World War I. Such migration peaked during the European boom of the
early 1970s. The second major wave was intra-regional migration of the
decade of the oil boom and beyond. At its peak, the number of migrant
workers abroad may have numbered between four and seven million
persons, a majority of whom probably came from rural areas. Perhaps
ten million Middle Easterners participated in emigration between 1973
and 1988, a considerably larger flow than that of the Turkish/Maghrebi
migration to the EC. The numbers who have been indirectly affected by
migration is larger still.

Rural-urban migration has been even more massive. Millions of
rural people have moved to the cities during the past generation.
Many others participate in rural-rural migration, while still others
have ceased working in agriculture to take up rural non-agricultural
jobs. All of these forms of migration have constituted a fundamental
force in shaping agricultural development in the region.

Migration has had both positive and negative effects on
agricultural development. The process has been very complicated, and
despite its importance, it remains relatively understudied.
Accordingly, this paper is mainly an attempt to formulate hypotheses
for further study. Even the total volume of migration is unknown; the
effects are still more uncertain. The following paragraphs summarize
the hypotheses which are tentatively advanced in this paper.

* This paper draws heavily on an earlier FAO paper on the same
subject presented at the FAO 19th Regional Conference for the Near
East, Muscat, March 1988.


The massive international and rural-urban migration of the 1970s
and early 1980s took place so rapidly that temporary disequilibria and
distortions emerged in farm labour markets. Although in some
countries the total agricultural labour force has actually declined,
in most countries "labour shortages" have been localized, seasonal
shortages of adult male labour. The question then becomes how well and
with what consequences other types of labour can substitute for men's.
Women have substituted for men in many, but not in all, cases. Farm
labour markets may also be geographically segmented: workers in
near-by villages may be imperfect substitutes for those within
villages. Commercial farms may suffer more from the labour exodus than
do family farms, while the farm labour force appears to be ageing in
many countries as young people abandon farming. To the extent that
workers from outside the area are good substitutes for departed
locals, rural-rural migation may help to solve the problems of "labour
shortage" which international and rural-urban migration may create.

The consequences of such labour market structure and performance
for agricultural production depend on whether output prices rise when
labour costs increase and/or whether there are alternative
labour-saving technologies available. We hypothesize that labour
migration will have greatest effect on cereal production, farms in
marginal rain-fed zones, and on labour intensive operations.

Farm mechanization may have been stimulated by labour migration
from some countries. However, the overall pattern of mechanization
suggests that machines are substituting more for animal power than for
human labour: most mechanization in the region so far has been of
power-intensive tasks. However, even here, labour scarcity may play
some role, and combine harvesters are spreading most rapidly in those
countries with declining agricultural labour forces.

Remittances have been very large in many countries, and have
become central to balance of payments equilibrium in some countries.
(Of course, such effects do not occur with rural-urban or rural-rural
migration.) We focus here on their micro-economic impacts, which are
qualitatively similar for all types of migration. Again, although we
have quite a few studies, we are still far from a thorough
understanding of the implications of remittances for agricultural
development. They seem to be spent primarily on housing, land, and
human capital. Their use for direct agricultural investments is less
common, although there are many exceptions. Remittances will be
directly spent on increasing agricultural production only if such
investments are profitable for farmers. Remittances may also raise
reservation wages, increase the demand for rural, non-farm employment,
and contribute to increases in land prices.

Rural-urban migration has some particular effects on
agricultural and food systems. First, it has significant implications
for demand patterns: urban people often consume different commodities
than their rural cousins. This, in turn, may have implications for
the balance of trade, if, for example, land suitable for bread wheat
is more scarce than that for producing seminola, as is often true in

-7 -

the Maghreb. Second, rural-urban migration also contributes to
the food subsidy bill: the bulk of subsidies go to urban areas. Such
subsidies may have significant consequences for supply by diverting
public investment funds. Third, such migration may cause high-quality
farm land to be lost to urban housing.

Migration of all types almost certainly reduces rural poverty.
Its impact on income distribution is much less clear because although
poor people migrate, the very poorest are usually unable to do so.
Although on balance we believe that migration is probably equalizing,
much more study of this phenomenon is necessary before any firm
conclusions may be drawn.

The impact of migration on land tenure, asset distribution, and
village organization is still more opaque. There may be some
incentive to return to share-cropping if labour supervision becomes
more difficult as a result of emigration. Some argue that migration
"polarizes" the distribution of land, as poorer migrants sell out to
wealthier farmers who remain. Others, however, contend that
remittances help to preserve small farms and that the additional,
non-farm sources of rural income reduce the importance of land in
village social stratification patterns. Resolving this important
debate would require a series of methodologically similar studies for
different countries and regions.


A. Causes

Although there are many political refugees in the region
(Palestinians, Kurds, etc.), most of Middle Eastern migration is for
economic reasons. Most migrants seek to make money, and to send at
least a substantial fraction of their savings back to their areas of
origin. Most international migrants expect to return home some day, an
expectation shared by their employers and host governments. By
contrast, rural-urban migration is more often permanent, although here
the patterns may be very complex, involving seasonal migration and
other patterns. The same is true for rural-rural migration.

The proximate cause of migration has been the large gap in wages
between sending and receiving areas. For example, an unskilled
Egyptian in the mid-1970s could earn thirty times more money when
working at a Saudi construction site than he could on an Egyptian
farm; even today, the differential in monthly earnings is
conservatively estimated at four-to-one. Turkish and North African
wages were less than 25% of German and French wages. Urban incomes in
the region are usually between two to three times higher than rural
incomes. We shall first examine emigration from the Maghreb and Turkey
to the EC, then migration to the oil-exporting states, then turn to
rural-to-urban migration, and finally briefly glance at rural-rural
migration in the region.

- 8-

B. International Migration: to the EC

Migration from the Maghreb and Turkey to Western Europe is the
older of the two main types of international migration. It was
stimulated by the accelerating demand for labour in Western Europe in
a period of rapid growth and by the reluctance of increasingly
prosperous Europeans to perform the dirty, difficult, and often
dangerous jobs which any industrial society requires. Accordingly,
European governments turned to foreign sources of labour. Since the
gaps in wages and income were very large, and since there were often
serious problems finding adequate employment at home, many Maghrebi
and Turkish residents eagerly sought the opportunity to migrate.

Algerian migration to France is by far the oldest of all of the
labour flows considered here. Although some Algerians began working
in France as early as the 1870s, the major increases in flows (and
stocks of Algerian workers in France) were the result of World War I
(when the numbers rose from about 30,000 to over 150,000) and the
post-World War II economic boom. The official number of Algerian
arrivals in France rose from some 67,000 in 1947 to over 200,000 in
1955. These numbers continued to rise in the first decade of
independence; in view of the hostility and discrimination suffered by
many migrants to France, the Algerian government suspended migration
in 1973. Nevertheless, the flows continued: by the early 1980s,
estimates of the number of Algerians abroad, nearly all of whom were
in France, range from 500,000 to one million (Bennoune, 1988; Nelson,
1985; Birks and Sinclair, 1980). Workers abroad, a rather high
proportion of whom are skilled and semi-skilled, may constitute
perhaps 10% of Algeria's total labour force.

Moroccan and Tunisian migration are of more recent origin and
are usually thought to be smaller in absolute size: in the 1980s
perhaps 200,000 Tunisians and 300,000 Moroccan workers are resident in
the EC; some estimates of the total number of Moroccans abroad are as
high as one million. Since these numbers include family members not in
the labour force, the actual number of workers is probably no more
than 33-50% of those figures. As with Algerian migration, the
principal boom occurred in the 1960s and early 1970s; thereafter,
immigration policy in the EC became much more restrictive.

Turkish migration, largely to the Federal Republic of Germany,
also boomed in the same decade. The number of Turkish workers abroad
rose from about 20,000 in 1962 to 165,000 in 1967, to 660,000 in 1972.
By 1979, it was estimated that about 900,000 Turkish workers lived in
the EC or in Austria. Both Turkey and Tunisia have tried to reorient
labour migration away from the increasingly restrictive EC market
towards the oil states of the Near East during the past decade.

C. International Migration: to the Oil-Exporting Countries

As in the case of migration to the EC, the fundamental cause was
the wage and income gap between sending and receiving countries. The
low wages in the sending country were the result of low productivity
in agriculture and the failure of earlier industrialization to absorb

- 9 -

surplus labour. Wages in the oil countries rose because the demand
for labour shot up while domestic supply response was limited by
economic, social, and political factors. On the demand side, the
explosion of oil prices in the 1970s flooded the treasuries of the oil
exporting states (1). These states then launched ambitious
development plans and investment projects. Such spending implied huge
increases in the demand for labour of all types, especially including
unskilled service and construction workers.

However, demographic and socio-political factors constrained
domestic labour supply in the oil states. First, the nations of the
Gulf and Libya have relatively small populations: only Saudi Arabia
has as many as ten million indigenous inhabitants. Like other nations
of the Third World, they have very young populations: approximately
40% of the population is less than 15 years old, and a majority is
younger than 20 years of age. Since all of these nations have made
major attempts to increase school enrollments, most of these
youngsters are not yet available for work (2). This implies that the
size of the economically active population and hence of the domestic
labour supply is smaller relative to countries with the same
population but having an older age profile. The low female
participation in the labour force further exacerbated the shortage of
domestic workers in the oil-exporting states (3). Meeting the huge
demand for labour which the oil boom stimulated from domestic sources
was simply impossible. Foreigners were indispensable.

Although there was some labour migration in the region prior to
the oil price increase of 1974, the aftermath of the Ramadan War marks
a turning point in regional migration. Previously, more than 80% of
workers going to the oil states were Arabs, mainly Egyptians, Syrians,
Yemenis and Palestinians. The major characteristics of the Gulf
labour market during this period were the relatively narrow wage
differentials between sending and receiving countries, and the high
skill level of many migrant workers. Further, Iraq and Oman, which
became major labour importers during the late 1970s and 1980s, were
net exporters of manpower before 1974.

All of this changed drastically in 1974: the absolute numbers of
Arab emigrant workers rose dramatically, with emigration from poorer
countries like Egypt and Yemen being especially prominent. An
estimate of the extent of the phenomenon is shown in Table 1.

(1) For example, revenues in Saudi Arabia rose from $29 billion in
1974 to over $36 billion in 1978, to $113 billion in 1981; in
Kuwait, oil revenues more than doubled from 1974 to 1980.
(2) In 1980, primary school enrollment for boys was universal in Iraq,
Kuwait and the United Arab Emirates, while Saudi Arabia had
increased the percentage of boys in school from 22% in 1960 to 77%
in 1980, according to UNESCO figures.
(3) Only about 5% of the total Saudi labour force was female in 1970,
a percentage which rose by only a point or so by 1980.

- 10 -

It must be emphasized that such figures are only approximations;
the problems of data quality and comparability are severe. Estimates
of the numbers of migrant workers vary greatly. For example, although
one study places the total number of migrant workers in Iraq in 1975
at 15,200, an alternative source offered estimates of 65,700. The
estimate of perhaps 1.5 million migrants in the whole region in 1975
should probably be treated as a lower bound.

Further oil price increases in 1979 approximately doubled
government revenues by 1980 (1). This stimulated still more ambitious
development plans, yet more lavish projects, and even more generous
social welfare programmes, raising the demand for foreign labour still
higher. By 1980, some authorities estimate that an additional 700,000
workers had sought work in Saudi Arabia and the other Gulf States
alone, while the number of migrant workers in Iraq had increased more
than ten times to about 750,000 (See Tables 2 and 3). The number of
expatriate workers in Libya had risen from just under 50,000 in 1973
to over 400,000 in 1980, by which time they constituted about
one-third of that country's labour force (Sherbiny, 1984).

As for the earlier phase, there are sharply contrasting
estimates of the total number of workers. Estimates of the total
number of Egyptians working abroad in 1980 range from 500,000 to 1.5
million. The ILO estimates the total number of workers in Saudi Arabia
in 1980 at one million, only two-thirds of the figure presented in
Table 2. Sherbiny estimates the total number of expatriate workers in
Iraq in 1980 at 500,000, of whom 340,000 were reportedly Egyptians.
Birks and Sinclair estimate the total number of Arab migrant workers
in Iraq at 125,500, in Saudi Arabia at just over one million, and in
Libya at 545,500. However, their estimate for the total number of
migrant workers in the GCC states is very close to the number reported
by Owen (1,960,920 and 2,076,640, respectively). If we use a figure
of approximately two million for the GCC States, add perhaps one-half
million for Libya, the major question mark becomes Iraq, with
estimates as low as 125,500 (Birks and Sinclair) and as high as
750,000 (Quarterly Review of the Intergovernmental Committee for
Migration, Jan. 1986). It is especially difficult to estimate the
number of non-Iraqi workers in Iraq because the government does not
require passports from other Arab workers, and as Sherbiny says,
"after 1977 only qualitative information is available" (1984, p. 36).
For want of a better procedure, let us simply take the mid-point of

(1) Revenues changed as follows:

Country Revenue in 1978 Revenue in 1980

Saudi Arabia $39,700 $104,200
Kuwait $ 9,500 $ 18,300
Iraq $11,000 $ 26,500
Libya $ 9,300 $ 23,200

Only Iran suffered a decline (from $20,900 million to $11,600
million) due to the disruption engendered by the revolution and
onset of war in that country.

- 11 -

the two extreme estimates: 400,000. Then, when adding in another
several hundred thousand workers in Oman, Jordan, and the Yemen Arab
Republic, we have a total emigrant work force in the Region in 1980 of
at least three million. That is, the number of migrant workers in the
region roughly doubled in five years.

In addition to the large increase in the total number of
workers, two other trends stand out during this period. First, the
share of Arab migrant workers in the total expatriate work force
declined from about 43% in 1975 to about 37% in 1980 as first South
Asians (Indians, Pakistanis, Sri Lankans, and Bangladeshis) and then
East Asians (especially Thais, Filipinos, and Koreans) poured into the
region. For the South Asians, the wage gap was even greater than for
some of the poorer Arab countries. East Asians often came as part of
"turn-key" construction contracts, where firms such as Hyundai
furnished plans, materials, management--and labour. These workers,
housed in company-provided barracks, had probably the least contact
with the local population. Second, the balance of the work force
began to shift away from the large preponderance of construction
workers towards labour to run the new facilities, and towards service
workers. By 1985, some believed that a majority of foreigners were
employed in services, and the share of Arab workers to the total had
declined to as little as one-third (Economist, 12/27/88, p. 34).

The fall in oil prices of the mid-1980s inaugurated a new and
uncertain period for intra-regional labour migration. The decline in
government revenues curtailed numerous development projects in the
region, reducing the demand for foreign workers in some countries.
Meanwhile the composition of demand continued to shift towards more
skilled workers and away from unskilled labour, with the important
exception of services. This trend has recently accelerated: the value
of construction contracts in the region fell by 25% from 1986 to 1987
(South, Sept., 1987, p.65). Non-Arab Asians continued to increase
their prominence; one source estimates that by 1985, South Asians
accounted for 43% and East Asians for 20%, of the totalexpatriate
workforce in the Gulf (Economist, 8/27/88, p. 34).

An estimate of the migration phenomenon in 1985 is given in
Table 3. As usual, a wide variety of estimates of the magnitude of
labour flows exists. The numbers in Table 3 are conservative; although
the pace of immigration may have slackened, large scale repatriation
of workers does not seem to have occurred, despite the conclusion of
many construction contracts. In particular, there is little sign of
decreased immigration into Saudi Arabia, the largest importer of
labour. Indeed the stock of workers in the Kingdom rose by about one
million from 1980 to 1985. This is not surprising; the Kingdom's Third
Development Plan for this period allocated even larger funds for
development projects than did the Second Plan. Estimates of the number
of expatriate workers in Saudi Arabia in 1985 range from 2.5 million
to 4 million (latter figure in Economist, 8/27/88, p.34). A comparison
of Tables 2 and 3 with Table 5 suggests that the number of workers in
Kuwait increased, but only slightly, from about 380,000 in 1980 to
perhaps 430,000 in 1985; the number of new work permits in Kuwait fell
over 59% from 1983 to 1985 (Economist, 9/6/86, p. 67). Countries such

- 12 -

as Qatar and the United Arab Emirates likewise showed only modest
increases in immigrant labour. ESCWA concluded that the total number
of workers in the GCC states in 1985 was just under 4 million.

Other sources, however, argue that there was little deceleration
of immigration during the first half of the 1980s. For example, some
have estimated the number of Egyptians in Saudi Arabia in 1985 at
800,000, not 500,000 (Mohammad Ibrahim Taha al-Saqa, Al Ahram Al
Iqtisadi, 870, 9/16/85). This same source, however, gives the number
of Egyptians working in Iraq as 1,250,000 in 1985, a number which is
widely believed to have been exaggerated for diplomatic reasons. A
common figure for the number of Egyptians abroad is 3 million.
However, perhaps the most reliable estimate of the number of Egyptians
abroad in 1985 is that of Nader Fergany, who has conducted the only
nation-wide survey of the phenonomen; he believes that the number of
Egyptians abroad has been consistently overestimated and was probably
about 1.2 million in 1985. The total number in the GCC states ranges
from approximately 4 million to 7 million, with the former figure
probably being less inaccurate.

As for previous periods, the number of workers in Iraq remains a
puzzle. Sherbiny suggests that their numbers increased from perhaps
500,000 in 1980 to 900,000 in 1985. Fergany, however, believes that a
combination of the difficult war-time conditions in Iraq, combined
with restrictions on the repatriation of wages, made that country a
less attractive destination, by the mid-1980s. An alternative source
suggests that the number of workers in Iraq declined to 248,000 in
1985. It seems quite clear that the fourth phase of migration saw a
deceleration, but probably not an actual decline, in the number of
migrant workers in the region. As a rough estimate, there were perhaps
5 million migrant workers in 1985.

However, actual decline may have set in for the 1985-1988
period. Fergany finds that Egypt is no longer a net exporter of
labour: returnees outnumber emigrants. The same also appears to be
true for Pakistan (Kazi, 1988). The share of returning migrants who
have been engaged in the agricultural sector is low; given the small
scale and capital intensity of Gulf agriculture, this is hardly
surprising. The real issue is where these returnees seek employment
upon their return. Some micro survey research, along with a
nation-wide survey, suggest that many returning Egyptians go back to
their villages. Many returnees to rural Pakistan find employment
quickly, although, like many agricultural workers, they also
experience underemployment (Kazi, 1988). The most serious impact of
return migration may be on the young, new entrants to the labour
force, who face the double problem of increased competition from
migrants and the reduced demand for labour due to the decline in

It should be emphasized that all the tables and data cited above
refer to the stock of workers at any one moment of time. The total
number of workers who ever participated in work abroad is much larger
than the stock in any particular year. Several sources (e.g., Fergany)
estimate that the average length of stay abroad for an Egyptian worker

- 13 -

is approximately two years. Tutwiler asserts that two to three years
is the usual length of stay abroad for Yemeni migrants. Let us use two
years as a rough estimate for the "turn-over" period for workers. If
we believe that the total number of migrant workers in the region was
about 1.5 million in 1975, 3 million in 1980, and 5 million in 1985,
then the total number who ever migrated may be (very roughly)
estimated at approximately fifteen to seventeen million workers during
those ten years; plausibly another five million or so have migrated
during the period 1985-1988 (although there may have been net returns,
this is perhaps compensated for by taking a three-year, rather than a
two year, turnover period). Of course, workers did "repeat": go,
return home, and go again. Nevertheless, it is not implausible to
guess that as many as twenty million people may have participated
directly in the regional international migration phenomenon. In the
Yemen Arab Republic most men have worked in Saudi Arabia at least
once. A recent, as yet unpublished, survey of over 1,000 rural
Egyptian families found that one-third of all males had worked abroad,
mainly in Iraq (Richard Adams, personal communication, 11/87). One
direct estimate places the number of Egyptians who ever migrated
during 1973-1985 at 3.5 million, just under three times the number
abroad in 1985; by this calculation, more than one-quarter of all
Egyptian workers have participated in migration for employment abroad.

To sum up, since the early 1970s, the number of foreign workers
in the oil-exporting countries has grown extremely rapidly and very
large numbers of people have sought and found work outside of their
countries of origin. While the numbers have grown, two distinct
characteristics have emerged. First, there has been a shift in demand
towards more skilled workers as development projects have come to an
end and maintenance has become more important. On the other hand, the
growth of demand for (largely unskilled) service labour remains
strong. However, especially in personal services but also in aggregate
the proportion of workers originating from Arab countries has declined
while that from Asian nations has increased several fold. Since most
immigrants from the rural areas are unskilled, the shifting structure
of demand for labour implies that the massive growth of opportunities
for rural migrants which occurred in the 1970s is unlikely to be
repeated in the 1990s and beyond.

D. Rural-Urban Migration

Rural-to-urban migration has been considerably larger than
either of the two major streams of international labour migration in
the region's modern history. It is plausible to argue that at least
half of the (very rapid) growth of the region's cities has been due to
such migration.

A very crude calculation would be to assume that the rate of
natural increase (fertility minus mortality rates) are the same in
urban and rural areas. Then the rate of rural to urban migration could
be approximated by the difference between the rate of growth of cities
and the overall rate of population growth. Such calculations are shown
in Table 4.

- 14 -

Several generalizations may be made, keeping in mind the very
crude quality of the data and assumptions. About one-third to one-half
of the growth of cities in Iran, Iraq, Morocco, Saudi Arabia, and
Turkey is due to rural to urban migration (a figure similar for other
areas of the developing world). In several other countries, the role
of migration in urbanization was even larger: from 1965 to 1980, such
population movements accounted for more than half of the total
urbanization in seven countries (Jordan, Lebanon, Libya, Oman, Sudan,
Tunisia and the Yemen Arab Republic).

Our methodology seriously underestimates the importance of
rural-urban migration for urban growth. First, the assumption that the
rate of population growth is the same in rural and urban areas is
dubious. We would expect both fertility and mortality rates to be
lower in urban areas: incomes and female literacy are higher in
cities, while infant mortality is lower. These considerations suggest
that urban fertility should be lower than rural. However, since
mortality is also lower in urban areas, it is unclear whether the
overall rates of natural increase are actually lower in cities than in
the villages of the region.

For some countries, there is direct evidence which suggests that
urban fertility and mortality rates are lower than those in rural
areas. For example, the Turkish rural population is growing at a
natural rate (i.e., excluding migration) some 40% faster than the
Turkish urban population (Kuran, 1980). Egyptian rural fertility
exceeds urban: in 1975-76 for example, the total fertility rate was
3.9 in Cairo and Alexandria, 5.0 in urban lower Egypt, 6.0 in rural
lower Egypt, and 6.8 in rural upper Egypt (National Research Council,
1986, p. 3). However, because rural mortality rates are also higher,
the rates of population growth in rural and urban Egypt may not be
significantly different.

More importantly, the high rates of natural increase in the
cities are not independent of rural-urban migration. Such migration is
often by whole families, and is especially prevalent among the young.
The transfer of population of child-bearing age to the cities
obviously raises the rate of urban natural increase. Many children
born in the cities have parents who have recently arrived from the
countryside. When this factor is taken into account, it seems clear
that rural to urban migration is the single most important explanation
for the rapid growth of urban areas for most countries of the region.
In addition to direct impacts on agriculture of emigration from the
countryside (considered below), urban growth raises the demand for
foodstuffs and often has important implications for national policy
toward agriculture, such as increasing governments' commitment to food
subsidy programmes.

In any case, the absolute numbers of rural migrants to cities
has been very large during the past generation. For example, in the
1960s over 90,000 Moroccans migrated to towns. In the mid-1970s in
Turkey, over 650,000 rural residents arrived in the cities every year;
by the 1980s, 800,000 arrived every year (Danielson and Keles, 1980).
From 1968 to 1977 just under one million rural Iraqis moved to the

- 15 -

cities, especially to Baghdad and Basra. The numbers of rural-urban
migrants in the region greatly exceed those of international migrants,
large as the latter have been. For example, the total Turkish
emigration to the EC in the peak years of 1968-1973 was exceeded by
that country's rural-urban migration of just one year of the 1970s.

Economic motives induce such migration. The most widely accepted
economic model of migration by Todaro posits that the decision to
migrate is based upon the difference in the levels of income which
rural migrants expect to earn in the city, compared with what they can
earn in the countryside. Some more recent arguments have held that
migration may be part of a family strategy of income maintenance; one
brother goes to town to seek work, while others remain behind (Stark,
1984). However, the evidence in the region suggests that most rural to
urban migration is of entire families: e.g., Iraq (Sluglett and
Sluglett, 1987), Iran (Kazemi, 1980), Turkey (Danielson and Keles,
1980), Egypt (Hopkins, 1983), and Morocco (World Bank, 1981). Still
more complicated, computable general equilibrium models have been
formulated (Kelley and Williamson, 1984). These stress the role of the
terms of trade between agriculture and industry in explaining

Urban wages and incomes are substantially (i.e., 1.5 to 3.0
times) higher than those in rural areas (El Ghonemy, 1984). Of course,
a potential migrant has to consider the risk of failing to find a job
in the cities. This risk is usually measured by the urban unemployment
rates. There are considerable problems, of course, in measuring
unemployment in developing countries. Official, measured employment
rates are often quite high in the region: 16% (Turkey, 1985), 25%
(Tunisia, 1983), 20% (Iran, 1982); 16% (Algeria, 1983); 8% (Jordan,
1983) (Kopits, 1987; Nelson, 1979; Johnson, 1988; Nelson, 1985; World
Bank, 1986). Egypt's measured unemployment rate, on the other hand,
was about 5% in 1980 (Hansen, 1985). Even if we take the higher
numbers at face value, and even if we assume that rural workers have
no difficulty finding jobs in the countryside, migrating to the city
is still a good investment: using the highest measured unemployment
rates still implies that so long as urban incomes were even as little
as one-third higher than rural earnings, migrating to the city would

In fact, the incentives to migrate to the cities are stronger
than the above story would indicate. None of the countries have
unemployment insurance schemes: many officially unemployed persons are
very likely actually earning some (meager) incomes in casual
employment. There is also some evidence which suggests that rural
migrants often find jobs fairly quickly: more than 70% of migrants
surveyed by a World Bank study found work in less than three months
(World Bank, 1981, p. 247). Further, despite the higher living costs
in urban areas, it is very unlikely that these cost differentials are
as large as the differences in nominal wages. Housing may be more
costly in the cities, but the extensive food subsidy systems
maintained by many regional governments are usually more effective in
urban than in rural areas. Since food usually constitutes over half of
the total expenditure of poor people in the region, food subsidies

- 16 -

significantly reduce the rural-urban cost-of-living differences.
Finally, other benefits of urban life, like better educational and
health facilities, further increase the gap in "real income" between
urban and rural areas. The incentives favouring rural to urban
migration are extremely powerful in the region.

E. Rural-Rural Migration

Nearly every country in the region experiences at least some
intra-rural migration. Such movements are dictated by differences in
cropping calendars in different ecological zones and by, as usual,
differences in relative income and wage levels. A very common pattern
is the migration of workers from mountainous and rain-fed areas into
areas of more intensive cultivation (often irrigated) for harvest
labour. Such a pattern is common in Iran, Morocco, Sudan, and Turkey,
among many others. Two broad patterns may be distinguished: seasonal
migration in which migrants return home after the end of the peak
labour demand season and more permanent migration in which the workers
are originally drawn by the high seasonal labour demand and wages but
then stay on in the area subsequently. Of course, most labour flows
are actually a mixture of these two "ideal types". In some cases this
migration can be quite large: for example, it is estimated that over
half of all of the harvest labourers employed on the mechanized
farming schemes around Gedaref in Eastern Sudan come from Darfur and
Kordofan. Before the outbreak of civil war in Sudan, many residents
of Bahr al-Ghazal and other southern areas of the country also
participated in such migration for work.

A second type of rural-rural migration is that of seasonal
migration of people who combine livestock husbandry with agriculture.
Probably a majority of the residents of Darfur and especially Kordofan
in Sudan move about during the year. These movements are imposed by
the agro-ecology of the monsoon rains: herdsmen move north as the
rains advance and south as they retreat. Cattle herders in Sudan must
seek a balance between available pasture and insect parasites of
cattle: the advancing monsoon brings both, so herders must continually
move. Traditional animal husbandry in these areas requires rural-rural
migration. Accordingly, most rural inhabitants of these areas
participate in such movements.

A final type of "migration" is the exit of rural workers from
agriculture to rural, non-farm employment. Since such employment
changes may not involve spatial relocation, its inclusion in a
discussion of migration may be questioned. We include it however
because of its importance in changing the supply of labour to
agriculture and for its relevance to debates on "labour shortages",
discussed below.

In many countries of the region, off-farm rural activities have
become increasingly important. For example, a USDA study of Syria
noted that the best explanation of the rate of decrease in the farm
labour force across Muhafazat was the rate of increase in off-farm
wage rates. Returns to off-farm labour in the Yemen Arab Republic are
much higher than to farm work (Tutwiler, 1989). A survey conducted in

- 17 -

Egypt in 1977 found that approximately 50% of village incomes came
from non-agricultural activities. Such diversity of incomes was not
only characteristic of the sample as a whole and of individual
villages, but of households as well: as many households (48%) received
income from non-agricultural activities as from farming (Radwan and
Lee, 1985). In Morocco, some 40-50% of total rural employment is now
outside of agriculture (USAID, 1986), while in Algeria only 40% of
rural incomes came from agriculture (Adler, 1980). Some 50% of
Jordanian farmers are not dependent on farming as their primary source
of income (USDA, 1986). In Tunisia, 50% of farm owners also had
non-farm jobs; 300,000 to 400,000 rural people were employed in
non-farm jobs, compared to a farm labour force of just over one
million (World Bank, 1982). Off-farm employment may have reduced the
supply of young adult males to agriculture, and although the
alternative jobs may not be directly displacing farm labour, they are
changing domestic farm labour markets considerably. (We shall see
below that this force, however, may itself be partly influenced by
international labour migration because of the pattern of spending of

F. Selectivity of Migration

Who are these migrants? There are major differences in migrant
characteristics according to the type of migration. Rural-to-urban,
and rural-rural transhumance of pastoralists often involve entire
families. This is much less true of seasonal harvest migrants and of
international migration. Both international and rural-urban migration
is disproportionately an activity of young people. Despite differences
in the composition of the migrant workforce among sending countries,
the vast majority of international migrants are young adult males. For
example, one study of Sudanese migrants found that over 90% were men,
and 60% were between 18 and 30 years old (Berar-Awad, 1984). A survey
in Egypt found that the average age was 32 years (Fergany, 1988).
Similar patterns have been observed in the Yemen Arab Republic and

By contrast, the educational and occupational composition of
migrants varies markedly across countries. For example, emigrants from
Sudan and Jordan are largely skilled. The ILO survey cited above found
that only 11% of Sudanese migrants were illiterate, compared to a
national illiteracy figure of over 75%. About 63% of migrants in the
ILO sample possessed some skills (Berar-Awad, 1984). Relatively
skilled workers tend to predominate among non-Arab Asian labour,
although there are many low-skilled service workers. Algerian and
Turkish migrants to the EC also have a high percentage of skilled
workers. On the other hand, emigrants from the Yemen Arab Republic are
overwhelmingly unskilled and rural. The World Bank estimates that in
1982 between two-thirds and three-fourths of the half-million Yemenis
abroad in that year were from rural areas (World Bank, 1986) (1). In

(1) As usual, there are alternative, widely different, estimates of
the number of Yemeni migrants: Birks and Sinclair give a figure
of less than 350,000 in 1980, while the Yemen Arab Republic's
Central Planning Organization believes that 1,400,000 are out of
the country.

- 18 -

particular, there is evidence that emigrants come disproportionately
from large, traditional, extended families, living in rain-fed farming
areas and often owning little or no land (Tutwiler, 1989). Still other
countries occupy an intermediate position: although many skilled
Egyptians have emigrated, it is likely that a majority of emigrants
from that country have relatively low skill levels. Fergany's survey
found that a majority of migrants came from rural areas and that only
about 10% of emigrants have a university education. However, he also
notes that the percentage of emigrants with an intermediate education
has increased over time (1). It is quite clear that rural males, the
backbone of the agricultural labour force, have participated
extensively in the migration process. Our analysis of the impact of
migration on agriculture will therefore start with its implications
for agricultural labour supply.


A. Analytical Framework and Hypotheses

Many observers of labour migration from rural areas in the
region believe that such movements have created "labour shortages".
Any discussion of the causes, consequences, and appropriate policy
responses to labour emigration from agriculture requires some (perhaps
implicit) model of how labour markets work. A simple, highly abstract
view would be that different types of agricultural labour are perfect
substitutes. Such a perspective on farm labour markets leads us to
ask: "Is there a sector-wide shortage of labour"? A rather more
realistic view of labour markets would be that different types of
labour are highly imperfect substitutes, especially in the short-run.
This model suggests different questions. Since migrants are often
young adult males, we ask: 1) Is women's labour a good substitute for
men's? 2) Can hired labour substitute for family labour? 3) Are
workers from outside a village good substitutes for those within a
village? and 4) Are older workers good substitutes for younger ones?

In the following section of this paper we shall examine some
evidence on each of these questions. It should be emphasized that
there are many unresolved issues here; there is much that we do not
know. Accordingly, the following discussion is offered mainly as a
"research agenda" rather than a set of conclusions. At the current
state of knowledge, it may be as important to pose the correct
questions as to offer the right answers.

Nevertheless, we would offer the following tentative conclusions
and hypotheses for discussion. The evidence suggests that "labour
shortages" are primarily temporally and spatially localized.

(1) "The data on educational status show a clear selectivity of
emigrants. The unqualified were less likely to emigrate and the
highly qualified were more likely to emigrate. This selectivity
has increased over time but not at the top of the educational
scale". (Fergany, 1988, p. 17).

- 19 -

They are mainly a problem of shortages of adult male labour in seasons
of peak demand for certain crops and in specific locations. Shortages
of hired labour appear to be more acute than of family labour,
although as we shall see, certain areas of Syria and the Yemen Arab
Republic may be exceptions here. Further, these shortages are likely
to be "temporary disequilibria" rather than "long run structural"
problems, given the growth of the rural labour force in most countries
of the region. The evidence suggests that the experience of the 1970s
was historically anomalous: the shortages of that decade may be
replaced, in many countries, by "surpluses" in the late 1980s and
1990s. Labour markets appear to be segmented, but women do seem to be
substituting for men in many farm tasks.

B. Evidence of Sector-Wide Labour Shortages

Several types of evidence are available here. Table 5 shows the
actual and projected trends in the absolute size of the agricultural
labour force in selected countries of the region. For most countries,
the farm work force grew during the 1970s, and is expected to continue
to grow into the 1990s and beyond. Despite the scale of both internal
and foreign migration, the absolute numbers of agricultural workers
increased in Egypt, Iran, Saudi Arabia, Sudan, Turkey, the Yemen Arab
Republic, and the People's Democratic Republic of Yemen between 1970
and 1985. In the Yemen Arab Republic the agricultural labour force
declined between 1970 and 1980, but then resumed its earlier increase.
According to FAO data, there has been a decline in the agricultural
labour force in Iraq, Jordan, Lebanon, Libya, and Syria. Some
alternative sources assert that the farm labour force in Algeria has
also declined, from about 1.3 million in 1966 to just over 1 million
in 1977 (Bennoune, 1988, p. 211). Likewise, several analysts believe
that the farm work force in Egypt declined from 1975 to 1985, by some
accounts by as much as 10% (Hansen, 1985; Commander, 1987).
Agricultural labour is very difficult to count because of the
seasonality of work; as we shall see below, the number of women
working in agriculture is also often systematically and significantly

The rates of growth of the labour force certainly exceed
estimates of the rates of growth of demand for immigrant labour in the
oil-exporting countries or in the industrialized nations. In 1984
(i.e., before the sharp decline in oil prices) the World Bank
estimated that the annual demand for immigrant labour of all types in
Saudi Arabia would grow at 2.9% between 1985 and 1990. During the same
period, FAO estimates that the annual growth rates of the labour force
in Jordan would be 4.3%, in Egypt, 2.9%, in Syria, 3.6%, and in the
Yemen Arab Republic, 3.1%. It is unlikely that the growth of demand
for labour in the oil-exporting countries will be able to absorb as
large a proportion of the region's labour force in the late 1980s and
1990s as it did during the 1970s and early 1980s. Indeed, we have seen
that there is some evidence for some countries (e.g., Egypt) that net
emigration may already be negative. The rate of growth of domestic
industrial job creation falls well below the rate of growth of the
labour force for most countries. In some countries (e.g., Egypt and
Algeria), government employment continues to expand. Since much of

- 20 -

this employment is unproductive, however, the sustainability of such
policies in the face of large external debts and diminishing sources
of foreign exchange is very dubious.

If the cultivated area and/or cropping intensity were rising
quickly enough, it would be possible for the rate of growth of demand
for farm labour to outstrip the rate of increase of supply. The ratio
of the arable land area to the agricultural labour force is an
indicator of the relative scarcities of land and labour. If it showed
a dramatic increase, we would suspect that there was indeed a "labour
shortage", in the sense of increasing scarcity of labour relative to
land. However, at the sectoral level, there is little evidence in most
countries to support this conjecture. Table 6 shows the trends in the
ratio of arable land area to the agricultural labour force. Although
the experience varies widely among countries, there has usually been
either little change in this ratio or a decrease in arable land per
agricultural worker. However, in Algeria, Iraq, Jordan and Syria we
find evidence from this ratio of "agricultural labour shortages".

A better indicator of labour shortages would be trends in real
agricultural wages. Unfortunately, most data on this critical variable
are spotty and unreliable. Such data as we have been able to assemble
are shown in Table 7. Once again there seems to be a wide range of
national experience: Egyptian real farm wages increased substantially,
almost doubling from the early 1970s to 1980s, and then increasing by
another 50% from 1980 to 1986; Egyptian real farm wages were 300%
higher in 1986 than in 1973. Nominal agricultural wages in the Yemen
Arab Republic soared during the past fifteen years, rising about 1200%
from 1973 to 1980 (Tutwiler, 1989; Carapico, 1985). Unfortunately,
however, there are very different estimates of inflation for this
period, ranging from the World Bank's low of 10% to Tutwiler's 40%.
Depending on the deflator used, real Yemeni farm wages may have
increased between 150% and 1000%. Agricultural wages in Pakistan
increased from 1977 to 1983 (Kazi, 1988). By contrast, wages in
Turkish and Moroccan agriculture have either stagnated or declined.
These figures are consistent with information on rural unemployment;
it is estimated that in 1986 perhaps one in five Turkish farm workers
suffered from unemployment (IMF, 1987).

For several countries, we have studies which attempt to assess
the balance of supply and demand for farm labour both at the present
time and in the future. The USDA conducted such a study in Syria in
1980 (USDA, 1980). A comparison of labour requirements with
availability for 22 crops, by month, and for each Muhafazah in the
country revealed that the average monthly required number of workers
for the country as a whole was only about 53% of the farm labour
force. Sector wide labour shortages did not appear even in Syria, one
of the countries where both the absolute number of farm workers and
the ratio of workers to land area declined.

The World Bank conducted a similar study for the Yemen Arab
Republic. It found that the average person-day equivalent requirements
for agriculture amounted to about 60 person days for every 275 person
days available. The World Bank then plausibly argued that no

- 21 -

agricultural sectoral labour shortage exists in the Yemen Arab
Republic. This is quite significant, since a larger percentage of the
rural labour force of Yemen Arab Republic emigrated for work abroad
than from any other country in the region. It is estimated that about
500,000 rural Yemeni men worked abroad in 1982, compared to a
agricultural labour force of just over one million. Thus, even the
massive emigration from the Yemen Arab Republic does not appear to
have created sector-wide labour shortages.

It should also be noted that forces other than emigration affect
the number of available farm workers and real agricultural wages. Some
of these, e.g. remittances, are part of the migration phenomenon and
are taken up below. Others, however, are less directly linked to
migration. In particular, two other forces have been reducing the
supply of young men to agriculture: spreading education and increasing
demand for rural non-farm workers. Although illiteracy remains
lamentably widespread in many rural areas of the region, nearly all
governments have made major efforts to increase the enrollment rates
of boys. Primary male enrollments are now essentially universal in
Algeria, Iran, Iraq, Jordan, Lebanon, Syria, Tunisia and Turkey
(UNESCO). Throughout the world, young men with even a rudimentary
education seek to leave farming for alternative employment.

A long tradition among economists holds that not only are all
workers perfect substitutes, but also that there are so many of them
that the supply of labour is perfectly elastic. Such a model predicts
smooth labour market adjustment and also stagnant real wages in
agriculture. As we have seen, there is some evidence from some
countries to support the second of these implications; at the same
time, some studies have cast doubt on the first argument. Research in
India (Bardhan, 1979) has shown that even in very poor regions of that
country (e.g., West Bengal) the supply of adult male labour was quite
inelastic with respect to the wage rate. Using data from a 1978
agricultural labour survey in Sharkiyya Governorate in Egypt,
Richards, Martin, and Nagaar (1983) found that the supply function of
adult male hired labour was inelastic with respect to the wage rate.
It would be very helpful to have similar studies from other countries
in the region.

The result of inelastic male labour supply with respect to the
wage rate is plausible for several reasons. First, a decrease in the
wage rate is unlikely to reduce the supply of agricultural labour,
since farm workers are typically very poor and cannot afford to work
any less. Second, an increase in the wage rate is unlikely to raise
their supply of labour significantly; they are already willing to work
full-time at the going wage, again largely because of poverty. If
inelastic agricultural labour supply functions are shifting back and
to the right (as workers emigrate and/or find alternative, non-farm
rural employment) while the demand for labour is shifting out because
of rapidly growing populations and incomes, then we should observe
increasing agricultural real wages. Although this sequence of events
fits the Egyptian case, it does not appear to be true for other
countries. Further research on adult male labour supply functions is
badly needed.

- 22 -

It is worth noting, however, that although the elasticity of
labour supply in agriculture may be less than one, the elasticity of
emigration from the countryside with respect to rural-urban, or
rural-foreign wage differentials may be considerably larger than one.
Further, even if individual labour supply functions in agriculture are
inelastic, continued stagnation of farm wages in some countries
obviously indicates that the rate of growth of supply exceeds that of
demand. In this case, the long-run, market labour supply function may
be highly elastic, in the spirit of the famous model of W. Arthur

To summarize, available evidence does not support the notion
that there are sector wide labour shortages as a result of emigration.
This, however, does not mean that emigration and other changes in farm
labour markets have not affected the course of agricultural
development in the region. It simply implies that a more nuanced,
disaggregated approach is required. The real world is not the
frictionless plane of simple economic theory.

C. Evidence of Shortages of Specific Types of Labour:
Extent of Substitutability

Regional farm labour markets may be characterized by highly
imperfect information, significant search costs, numerous policy
biases, and imperfect substitutability among different types of
labour. Emigration from such a world would have important consequen-
ces, especially in the short run, even if there are no sector-wide
labour shortages. If other types of labour are relatively poor
substitutes for young adult males, then their departure could engender
farm production bottlenecks. The issues here concern the extent of
substitutability and the elasticity of different types of labour
supply with respect to the wage rate. If adult males leave, and adult
male wages rise in a locality, to what extent can farmers find other
workers who are willing to work at these higher wages? What barriers
exist to employing such workers? What are the consequences of
expanding their employment?

1. Women's labour in agriculture

The emigration of young adult males has caused the feminization
of the agricultural labour force in many countries. The extent to
which women participate in farm work varies widely from country to
country and within regions of the same country. FAO data on female
participation in the farm work force is shown in Table 10. These are
probably under-estimates. Censuses and labour force surveys usually
ask respondents for their "principal occupation"; women usually
respond "housework", "caring for children", etc., even though, in
fact, they do a great deal of farm labour. It would be helpful if
national data collection agencies remedied this problem. One approach
is to try to determine how much of the total farm labour input comes
from women. For example, the 1978 labour force survey in Sharkiyya
found that women accounted for 15% of all hours devoted to crop
production; another survey in the same area found that women performed
about 40% of the work devoted to raising livestock (Fitch and Soliman,

- 23 -

1983). Both surveys found that over half (approximately 60%) of all
family labour was devoted to livestock production. This suggests that
approximately 25% of family farm labour is performed by women;
contrast this with the official estimate for 1980 of only 3.7% (Table
8). Over one-third of all family farm labour in Jordan is performed by
women (Aydin, 1989). Some observers of the Yemen Arab Republic and
Syria report that in contrast to the numbers in Table 8, female
farm-workers now outnumber males in those countries (Tutwiler, 1989;
FAO, 1986). Further realistic and comprehensive study of farm labour
markets in the region requires greatly improved data collection on
women's agricultural activities.

Female agricultural work is not limited to family labour. Women
also, and increasingly, are engaged in hired labour. Surveys in the
Egyptian Delta have found that approximately one-third of hired labour
hours come from women. They constitute 72% of the paid labour force
in Iran, over 25% in Jordan, over 40% in Morocco and nearly 40% in
Syria (FAO, 1986). This evidence suggests that one response to the
emigration of male labour has been the substitution of female labour.
But we need many more detailed studies on two critical questions: 1)
the degree of substitutability of women's labour for men's, and 2) the
elasticity of women's labour supply with respect to the wage rate.

Women have traditionally specialized in certain farm tasks,
whether as family or as hired labourers. For family labour, women
often have primary responsibility for the production of food to be
consumed by the family. One manifestation of this "farming as
extension of household responsibility" is the common pattern of
women's specializing in food crops, while men concentrate on cash
crops. In the Yemen Arab Republic, for example, women have
traditionally done most of the work in cereal production, while the
cultivation of qat and coffee is done largely by men. Women often tend
small garden plots, and do most of the work of crop processing, such
as threshing, winnowing, and storing grain. They are often responsible
for work which is done on a daily basis, such as livestock work, while
men specialize in more seasonal tasks like land preparation and other
work done jointly with animals, as well as maintaining farm capital,
such as repair of irrigation networks and terraces. Men and women
often join together for harvesting small grains and cotton.

These specializations can help to explain some aspects of farm
labour markets. For example, in Sharkiyya Governorate in Egypt there
are two peak seasons for agricultural labour, one in May-June when the
harvesting and threshing of wheat coincides with land preparation for
summer maize, and the cotton harvest season of September-October. Farm
wages for adult male labour rose some 18% above the yearly average in
the first season, but actually fell some 6% below average during the
second peak season. The most plausible explanation for this phenomenon
is that the (much cheaper) labour of women (and children) during the
cotton harvest holds down male wages in the second period, while there
were relatively few substitutes for male labour for land preparation
activities (Richards, Martin, and Nagaar, 1983). The extent to which
women's labour substitutes for men's is a crucial determinant of the
impact of male labour emigration on wages.

- 24 -

Is this "traditional" gender division of labour changing under
the impact of labour migration? There is evidence which suggests that
the answer is "yes". The data in Table 8 show that in many countries,
female labour participation has increased. This may simply reflect
improved data collection, but qualitative evidence also suggests
increasing substitution of women's labour for men's. Observers of the
Yemen Arab Republic report that "it is not uncommon to find women
doing men's work when no man is available" (Tutwiler, 1989). A survey
of over 200 Egyptian rural women found that nearly half of all those
interviewed in the Delta engaged in ploughing, traditionally a male
job in that (and in most other) country. Women's labour has also been
substituting for men's labour in threshing in Middle Egypt. It seems
very likely that the departure of adult males has stimulated the
substitution of female for male labour on a wide scale in the region.

Even if women can perform all farm tasks, the result is likely
to be that women are even more overworked than previously. This may be
detrimental to the women's health, and possibly to the health of their
children. In general, however, the extent to which women participate
in farm labour is very much a function of poverty. Women from
well-to-do farm families throughout the region typically do not
participate in farm labour. There is clearly a strong preference in
the region for women to eschew farm work; only if a family cannot
afford to hire men, or has insufficient male family labour, or cannot
pay for or obtain machine services, or lacks alternative sources of
off-farm income, do we find widespread female participation in farm
tasks (Aydin,1989; Richards and Martin, 1983).

What are the implications for the elasticity of women's labour
supply? Most studies of Third World farm labour markets suggest that
the supply of poor adult males is highly inelastic: poor people cannot
afford to reduce their labour supply if the wage rate falls, and they
are usually unable to offer more labour if the wage rate rises because
they are already fully available for work. It is plausible that poor
rural women are in much the same position. This argument implies that
poor, hired women's labour supply will be relatively inelastic.

On the other hand, other considerations suggest that women's
labour is considerably more responsive to wage rate changes than is
men's labour. If the wage rate falls, women may work less, simply
because there are always alternative activities (child rearing, food
preparation, etc.) which women continue to do. Studies from the Indian
sub-continent suggest that women's labour supply is more elastic than
men's with respect to market wages. Of course, this is an empirical
question, and we simply do not know the answer for the Near East
region. Rigorous studies of the conditions and responsiveness of
women's labour supply in agriculture are badly needed.

A further complication is that because of the strong family
preference for women not to work for pay, the emigration of males may
reduce female farm participation. When remittances enter a village,
families are able to subsist without resorting to female farm labour.
However, the pattern here is highly complex. In some cases women
continue to perform certain farm tasks, especially those related to

- 25 -

producing food for the family, while the remittances are used for
other purposes. (The patterns of remittance use in rural areas are
treated in more detail below.)

Does the increasing percentage of farm work conducted by women
constrain agricultural production? We do not know: little or no
quantitative data are available on this issue. As before, we offer
some hypotheses which we believe should be explored. Social customs
combined with high illiteracy among rural women may pose certain
difficulties for them as farmers. First, in countries where subsidized
inputs are allocated through government agencies (e.g., cooperatives),
women may not have the same access to these inputs as men. Second,
illiteracy impedes technological progress, and in all major countries
of origin of migrant workers, the large majority of rural women are
illiterate. Third, the large majority of extension agents are men, as
are the (often more effective) agents of private input supply
companies. Social customs often make it difficult for these agents to
reach women farmers. Fourth, if women have limited experience in farm
management, they may find it difficult to take over this task.

The importance of these considerations varies considerably from
region to region within the same country, and indeed, from family to
family. In the absence of concrete evidence to the contrary, we are
sceptical that women's increased role in agriculture constitutes an
important constraint on production.

Literacy is (slowly) spreading among younger rural women, there
are numerous cases of successful farm management by women, male
relatives often provide assistance, and there are much lower social
barriers to male extension agents working with groups of women.

The available evidence suggests that women's labour has been
substituting for male family and hired labour in many countries of the
region. The precise forms and consequences of this substitution remain
uncertain: we know very little about the determinants of women's
supply of labour to agricultural labour, whether on family farms or to
the hired labour market.

2. Commercial farms: is hired labour a good substitute
for family labour?

Since "labour shortages" are primarily peak-season scarcities of
adult male hired labour, they could affect larger and commercial
agriculturalists more than small family farmers. Although in many
countries even small farmers hire labour at harvest time, a much
larger percentage of farm work is usually carried out by hired workers
on large commercial farms than on small subsistence farms. Even in the
Egyptian Delta, where even the smallest farms hire labour, farms above
ten feddans hire at least 75% of all labour hours (Richards, Martin,
and Nagaar, 1983). The differences between small family farms and
larger commercial operators are much more striking in many other
countries: for example, in the Eastern Province of Sudan, farms above
100 feddans hired all labour for weeding and harvesting, compared with
less than 10% for farms below ten feddans (ILO, 1983). As landless

- 26 -

agricultural workers are often among the first to seek construction
jobs or other work in the cities, larger farmers must meet their
labour requirements increasingly from the pool of smaller farmers. But
if they grow the same crops as their less wealthy neighbours, the
small farmers may, for security reasons, use their family labour first
on their own fields, so larger farmers must offer even higher wages to
attract wages. Egyptian evidence suggests that larger farmers do
indeed pay higher wages on average than small ones. It is plausible to
hypothesize that "labour shortages" are mainly a problem for
wealthier, commercial farmers.

Small farmers can often meet labour shortages in ways which are
not usually available to larger, commercial operators. For example, a
study of Minufiyya Governorate in Egypt found that, as the evidence
given above would suggest, both large and small farmers hired farm
labour. However, when emigration reduced the supply of hired workers,
small farmers responded by using family labour more intensively, by
substituting family female labour for tasks formerly carried out by
men, and by entering into a variety of complicated "labour sharing"
arrangements in which households exchange labour time with each other
(Taylor-Awny, 1984). None of these arrangements is usually open to
larger commercial farmers.

Labour shortages have also had this differential effect in the
Yemen Arab Republic. Although subsistence cultivation in ecologically
marginal areas has been affected, the World Bank believes that the
development of commercial farming has been especially inhibited by
high wages and the lack of appropriate technologies for raising labour
productivity in many areas. Similarly, in Tunisia, large irrigated
farms suffer the most from the outflow of labour in that country.
Reports mention the "difficulty in finding labour" as one of the
causes of the relatively low level of capacity utilization of large-
scale irrigation schemes there (USAID, 1986). However, these farmers
usually obtain a wide variety of subsidies on other inputs.

Commercial farmers faced with these problems have several
options, depending on local circumstances. First, they can resort to
share-cropping arrangements, leasing their farms short-term to smaller
farmers. This seems to be the principal response in Minufiyya
Governorate in Egypt. Second, they can encourage the importation of
labour from abroad, the response of the commercial farmers of the
Jordan Valley. Third, they can mechanize farm operations which face
particularly severe labour constraints. The role of labour shortages
in stimulating farm mechanization in the region is considered below.
Finally, they may be able to hire additional women workers, as
discussed above.

3. Village-to-village differences

The impact of labour emigration in certain areas may be
magnified by geographical labour market segmentation. Farm labour
markets are often poorly integrated with each other; there may be
shortages in one town and surpluses in another. One study conducted in
Sharkiyya Governorate in Egypt from 1978-81 found evidence of

- 27 -

involuntary unemployment: agricultural workers without jobs were
willing to work for wages below the going wage in the village! This
was particularly true during the off-season; during the peak season,
labour markets appeared to "clear": wages equilibrated supply and
demand of labour (Richards, Martin, and Nagaar, 1983). However, a
later study, conducted in 1984 in neighboring Gharbiyya Governorate
found not only generally lower overall unemployment rates (a change
mainly ascribed to the expansion of rural, off-farm employment), but
also evidence that the farm labour market did not clear even at peak
season (Commander, 1987). It would be very interesting to have similar
studies from other countries.

We do not really know why this phenomenon occurs. However,
several (as yet untested) hypotheses may be offered. First, whether or
not labour markets actually clear in the formal sense of economic
theory, there is little doubt that in many countries labour markets
tighten in the peak seasons. That is, higher wages must be paid
(often in the face of uncertainty of final output prices) and/or it
may be difficult to locate reliable workers. During the off-season,
farmers are willing to pay workers higher wages than they "need to" at
that time (i.e., there are other workers willing to work for less)
because the favoured workers will promise to work for the farmers
during the peak season, when it is hard to find labour (Hansen, 1985).
Such arrangements, similar to the "internal labour markets" common in
industrial societies, imply that labour markets do not automatically
clear in every season.

Second, labour markets in the region are often quite complex and
personalized. Complicated transactions involving exchanges of labour,
credit, animals, machinery, and other factors of production usually
require the parties to have good personal knowledge of each other.
Supervision costs are often high in agriculture. Personal knowledge
and trust help to reduce such costs. This may be particularly
important if increasing numbers of farmers also have other occupations
and therefore less time to devote to locating and supervising labour.
(We shall see below that there is considerable evidence that this is
increasingly true in the region). Migration is very often highly
localized; one village may have substantial outmigration while a
neighboring one may have much less. If farm labour markets are
personalized, then the surplus of labour in one village is a highly
imperfect substitute for the shortage of labour in another. We are
unaware of any studies which have attempted to test this hypothesis in
the region.

The increasing role of women in hired labour may compound the
problems of labour market segmentation. In the Tihama region of the
Yemen Arab Republic, women constitute a large percentage of the hired
labour force, and observers have noted that they are less
geographically mobile than men because they combine hired labour
(largely seasonal work, especially for harvesting) with many household

On the other hand, the heavy reliance on rural-rural migrants
for farm work in many regions suggests that "farm labour segmentation"

- 28 -

arguments may be easily exaggerated. One hypothesis would be that
farm tasks which are relatively easily supervised would be attractive
for rural-rural migrants. This might fit certain types of harvest
labour. Indeed, rural-rural migrants have long helped with harvest
labour: the tarahil workers from Upper Egypt commonly travelled to the
Delta to clean canals and to help harvest cotton, a crop which was not
grown in their areas of origin. A "division of labour" between
year-round and seasonal labour, supplied by socially and geographical-
ly distinct groups, has been observed for many different countries.
For example, in the two major areas of commercial farming in the
Sudan, the Gezira scheme and the mechanized farms producing sorghum in
the heavy clay soils of Central Sudan, migrant labour supplies a
significant percentage of harvest labour (Aricanli, 1985; Ibnouf,
1985; ILO, 1983). Rural-rural migration makes a fundamental
contribution to farm production in Sudan. It could be argued that
supervision costs may make migrants less attractive to jobs like
weeding. However, much of the weeding of the Turkish sugar beet crop
is done by migrants. The issue of the substitutability of "outsiders"
for local labour in agriculture deserves more careful study than it
has so far received. We believe that the key is likely to be
supervision costs.

4. Age issues

The exit of young adults from the countryside has led to the
ageing of the farm labour force in many countries. In Egypt for
example, in 1978 about 19% of the rural labour force was over fifty
years old, compared to 16% of the urban labour force. In Tunisia, it
is reported that the majority of adult male farmers are now over 55
years old. In Algeria, the percentage of the agricultural labour force
which was older than 59 years rose from 23% in 1968 to 33% in 1973
(Bennoune, 1988, p. 209). By the mid 1980s, most Algerian farm workers
were over 50 years old, and more than 100,000 were expected to retire
by the end of the decade (USDA, 1986). Similar phenomena are reported
in the People's Democratic Republic of Yemen. However, not all
countries where a high percentage of the farm work force has emigrated
has had this experience. In Jordan in 1983, for example, nearly 47% of
the total farm labour force was between the ages of 35 and 64 (48% of
the male labour force), and another 39% was between 15 and 34 years
old (Aydin, 1989).

Some of the changes may be "statistical artifacts", based on
shifting definitions of "farmer". The trend toward part-time farming
and multiple jobs characterizes many countries of the region (see
above discussion on off-farm jobs). In 1983 in Jordan just over 33,000
farm holders listed farming as their main occupation, while another
25,500 farmers devoted their primary attention to other jobs (Aydin,
1989). Further, if "farmer" means the "farm holder", then clearly most
owners of family farms, like owners of any major inherited asset, will
tend to be older. However, this author would judge that the ageing of
the farm labour force is a real phenomenon.

One must ask whether the ageing of the farm labour force
matters; does it adversely affect agricultural production? It is

- 29 -

worth remembering that in 1982 42% of all U.S. farmers were over 55,
and that the proportion of farmers over 55 has remained roughly
constant since 1954 (Boxley, 1987, pp. 24-25); a similar process has
occurred in Japanese agriculture. In neither of these countries has
the ageing of the farm work force reduced agricultural output. In the
Near East, however, a much higher percentage of older people are
illiterate. It would not be surprising if this retarded the pace of
technological change in agriculture. Studies which tried to assess
variation in technologies by age of farmer could help to clarify this

In the region, as elsewhere in the world, retaining skilled and
educated young people in farming is a real problem. Only more
mechanized, and therefore more interesting, less onerous, and higher
paying farm jobs can persuade young men to remain in the countryside.
More mechanized techniques may be necessary if older farmers are to
maintain levels of output. For example, it is much easier for an older
man to plough land with a tractor than behind oxen. We will examine
the role of labour market changes in inducing farm mechanization

D. Production Consequences of Increased Labour Costs

In a world of segmented labour markets, labour costs may rise
even if no sector-wide "labour shortage" exists. What consequences
would we expect this to have for agricultural production? In general,
the exit of labour from agriculture can plausibly reduce output
significantly only if output prices cannot rise sufficiently to cover
the increased cost of production or if possibilities for technological
substitution do not exist. If prices can rise, then the increase in
costs will be passed along to consumers. The extent to which
production will fall will depend on the price elasticity of demand.
Since this is relatively low for most foods, we would expect
relatively small declines in production. Further, increasing incomes
and population growth are certainly shifting the demand schedule to
the right, thereby increasing output.

Even if prices fail to rise, output might not fall if there
exist technological substitution possibilities. If it is possible to
shift to alternative technologies with higher output per worker, total
costs might remain constant, despite the increase in labour costs. In
this case, also, there would be no major production consequences of a
"labour shortage".

Such an analytical framework suggests three hypotheses: the
principal impact of migration-induced increases in labour costs will
be found in:

1. Cereal crop production.
2. Marginal rain-fed zones.
3. Labour-intensive operations.

Let us examine these in turn.

- 30 -

1. Cereal production

Cereal production should be most seriously affected because of
the very well developed international market for cereals. This has
contributed to the increasing reliance of governments in the region on
grain imports. Low production costs (often due to OECD government
subsidies) of foreign producers create low import prices, which, in
turn, constrain the extent to which higher domestic labour costs can
be translated into higher domestic prices. During the past decade,
real cereal prices have continued their historical downward trend.
Very strong rates of growth of demand due to population growth and the
expansion of incomes from the oil boom and political commitments to
ensure an adequate supply of cereals to the urban population have
encouraged many governments of the region to rely heavily on food
imports. It should be noted here that rural-to-urban migration
contributes to such policies by increasing the relative and absolute
size of urban populations.

Government taxation of agriculture through price policies can
also have a similar result. For commercial producers, the increased
price of labour cannot be passed on to consumers; costs rise, but
prices do not. Some of these farmers may leave farming as a result.
For subsistence farmers, the increased opportunities for off-farm
employment may combine with easier access to purchased food to reduce
crop production.

Although this hypothesis deserves to be studied in much greater
detail, there exists some evidence to support it. Of major product
groups, only cereals show a rate of growth of supply below that of
population for the Near East region (See Table 9). For some countries,
cereal output has even stagnated: Algeria, Egypt, Jordan, Tunisia and
Yemen Arab Republic. This is in striking contrast to the performance
of vegetables, fruits, and livestock products, all of which enjoy some
"natural protection" from international competition due to
transportation costs and quality (e.g., freshness) considerations and
none of which is taxed. This is quite significant in assessing the
relative role of emigration, since vegetables and fruits in
particular, but also livestock products are much more labour-intensive
than cereals. If reductions in labour supply were the principal factor
in reallocating agricultural production, we would expect a relative
decline in labour intensive crops and an increase in labour saving
ones. But as Table 9 shows, we observe precisely the opposite. In
Algeria, for example, value added in cereals stagnated between 1974
and 1986; during the same period, however, value added in vegetable
production grew at 7.4% per year and that of fresh fruit expanded at
4.3% per year. In Jordan, the harvesting of lentils and the production
of cereal crops in rain fed areas have been hampered by. labour
emigration. However, in that same country, commercial farmers in the
Jordan Valley have had few, if any, difficulties in importing adequate
amounts of Egyptian labour for the very labour intensive operations
required for greenhouse horticultural production.

These trends are also highly visible in the Yemen Arab Republic.
Although FAO data show little change in the cultivated area, other

- 31 -

authorities claim that the absolute cropped area declined (Enger,
1986) by 32% from 1975 to 1980 and by 17% during the next five years.
Most authorities believe that most of this was the "abandonment of
rainfed grain land, mostly marginally productive terraces in the
highlands" (Tutwiler, 1989). FAO data show a decline of about 20% in
the total cereal area in the Yemen Arab Republic from 1980 to 1985.
Production of terraced sorghum and other cereal crops has, plausibly,
been adversely affected by labour emigration. However, this is also
most certainly a function of import competition: other crops such as
vegetables and qat have seen considerable increases in production.
Land cultivated with potatoes, vegetables, and fruits increased from
35,600 ha in 1975 to 48,500 ha in 1980 (Tutwiler, 1989). Qat
production increased more than 200% from 1975 to 1985. Although it is
often asserted that qat is considerably (say, 25%) less labour
intensive than grain crops like sorghum, this is true only of the
mature plant. Land preparation for qat "requires approximately four to
five times the labour required for sorghum. Young shrubs often need to
be hand-watered if the first year's precipitation is less than 800 mm"
(Tutwiler, 1989, p. 11). The point is that even if products require
considerable labour, if they face little or no import competition in
an environment of expanding demand and therefore rising prices and
profitability, the "labour constraint" does not impede expansion of

However, it is possible that the above reasoning understates the
role of labour emigration by overstating the impact of international
competition. A recent FAO study of price policies in the region found
little correlation across countries between low real nominal
protection coefficients (NPC's) and rates of growth of wheat
production. Further, mounting concerns with food security and
excessive import dependence led some countries to modify their pricing
policies by the early 1980s. For example, the Syrian government paid
wheat farmers about 30% more than the international price in the early
1980s (FAO, 1984), and in 1982 Yemen Arab Republic cereal farmers
received prices which averaged 360% of those prevailing on
international markets (World Bank, May, 1986).

2. Marginal rain-fed areas

The second hypothesis is that production in marginal ecological
zones should be especially affected by emigration. There are three
reasons for this. First, we might expect the emigration of labour to
be especially heavy from these areas: the very poverty of many
ecologically marginal areas encourages their young men to emigrate in
disproportionate numbers. Second, such areas usually primarily produce
cereal crops; accordingly, they will suffer from the same
international price competition as other cereal producers. Third, it
is often difficult for farmers in such areas to adopt alternative
cost-saving technological packages, for both agronomic and economic

In mountainous areas, it may be very difficult to adopt
labour-saving mechanical technology. Terraces may be too narrow to
permit the use of tractors (Yemen Arab Republic), roads may be so poor

- 32 -

that it is difficult to get tractors and spare parts into an area
(Western Sudan), etc. Economically, the highly variable rainfall of
such areas dramatically increases the risks of any investment,
including investment in labour-saving technology. Much of the Near
East receives less than twenty inches of rain per year, a level of
rainfall which makes non-irrigated agriculture extremely risky.
Precipitation also fluctuates from year to year, and is usually
seasonally concentrated. Even within the "rainy season", precipitation
can be highly irregular. To take only one example, during 1985-86
Tunisian rainfall was 150% above normal in March, but little rain fell
during the crucial planting months of October to December. This
variability dramatically increases farmers' risks, and reduces the
expected return from private agricultural investments.

This hypothesis deserves further exploration. In particular, it
is possible that "vicious circles" may have emerged in such areas: low
incomes stimulate migration, migration raises costs, higher costs
reduce production, which lowers incomes still further, stimulating
more migration, and so on. Some offsets, such as migrants' remittances
may partially or entirely counteract such forces, but we need more
detailed studies of such regions to be sure.

3. Labour-intensive operations

If labour migration raises labour costs, those crop operations
which use large amounts of labour should be especially affected. We
are unaware of any detailed studies of this issue. However, some
indirect evidence may be gained by studying the pattern of farm
mechanization in the region. If labour migration is indeed
dramatically raising the costs of farm labour, we would expect
extensive substitution of machines for workers in the most
labour-intensive farm operations such as harvesting. We now turn to a
brief examination of the pattern of mechanization in the region.

E. Mechanization: A Response to Labour Shortages?

It is widely believed that a major consequence of labour
emigration has been the rapid diffusion of farm mechanization in the
region. The mechanization of agriculture has proceeded more rapidly in
the Middle East and North Africa than in any of the other major areas
of the developing world during the past 15 years (Binswanger et al.,
1987). Since mechanization is usually capital-using and labour-saving,
the increased incomes due to the oil boom combined with labour
emigration are held to have increased the price of labour relative to
capital, causing factor substitution. In this view, farmers mechanize
because they cannot find the large numbers of workers which they would
need if traditional techniques were still used. However, the available
evidence suggests that the mechanization phenomenon in the region is
considerably more complex than this.

Tables 10-12 show that mechanization is spreading rapidly in the
region. Much, if not most, of this mechanization has taken the form of
tractorization. Although tractors are highly versatile, they are
mainly used for primary tillage, transport, and power for irrigation

- 33 -

pumps and portable threshers. The use of combine harvesters has spread
in some countries, but they are much less common than tractors. More
sophisticated machines such as seed drills, fertilizer applicators,
and fruit and vegetable harvesters are relatively rare.

This is consistent with the worldwide pattern of farm
mechanization. Binswanger et al. (1987) has suggested a very useful
typology of farm mechanization by distinguishing between "power
intensive" and "control intensive" operations. The former include
such tasks as water-lifting, primary tillage, threshing, grinding,
milling, crushing, and transport. These tedious tasks require much
physical energy, and farmers have used work animals to assist them
here for many centuries. In contrast, control-intensive operations
"require primarily the control functions of the human mind or
judgement" (Binswanger et al., 1987). Such operations include
sifting, winnowing, pest control, and the harvesting of cotton,
fruits, and vegetables. Grain harvesting and secondary tillage and
interculture are intermediate operations. Binswanger argues that the
typical pattern of diffusion of mechanization follows the sequence:
power-intensive tasks, intermediate jobs, and control-intensive
operations. Although the first stage of mechanization may save some
human labour, labour scarcity is not the key to this process; the
critical economic consideration is the relative cost of machine and
animal power. But as mechanization proceeds into the second and third
phase, labour-saving becomes increasingly important.

In many countries of the region, mechanization of power-
intensive operations has proceeded quite far. Nearly all arable
rainfed land in the Mediterranean countries is now ploughed by
tractor. In the Settat region of Morocco, for example, over 80% of the
land is tilled at least once a year by tractor (USAID, 1986). Most
land was ploughed by tractor in Syria at the beginning of this decade
(USDA, 1980). Over 90% of cotton, maize, rice, wheat, birsim, and
vegetable land preparation in Egypt is done by tractor (Commander,
1987, p. 244). Threshing machines are also widely used throughout the
country. Virtually all farm land is prepared by tractor in Jordan
(Aydin, 1989). On large (i.e., over 500 feddans) commercial sorghum
farms in the heavy clay belt of the Sudan, land preparation is
entirely mechanized; the large majority is mechanized on all farms
above 10 feddans (ILO, 1983). In Turkey the ratio of tractors to the
cultivated land area in 1984/85 was 49 ha per tractor, comparable with
the 40 ha per tractor in the United States (FAO, 1986). Some experts
believe that Turkey has too many tractors per hectare, resulting in
under-utilization of the tractor park! (World Bank, 1983).

The use of combine harvesters is proceeding more slowly. The
rate of growth of the numbers of tractors has exceeded that of
combines except in Jordan, Sudan, and the People's Democratic Republic
of Yemen. In the first two cases, the growth rates are somewhat
deceptive because the starting base point is so low. Furthermore, the
Sudanese government heavily subsidizes the use of combines. A recent
study of farming in the Damazine area found that adoption of combines
raises farm incomes by 30%, but that 68% of the net income of fully
mechanized farms (i.e., tractors and combines) was attributable to

- 34 -

government subsidies (Ibnouf, 1985) (1). Using a linear programming
framework, Ibnouf found that even if one assumed that the shadow price
of capital (real rate of interest) was zero, if fully mechanized farms
were charged input prices which reflected the actual scarcities in the
economy, they would lose money (earn negative profits). By contrast,
partially mechanized farms (tractors only) remained socially
profitable even with a real rate of interest of nearly 20%.

Apart from the anomalous Sudanese case, the diffusion of combine
harvesters is furthest advanced in countries whose agricultural labour
force has actually declined, such as Algeria, Jordan, and Iraq, or is
relatively stable, e.g., Turkey and Syria. (See Tables 5 and 11).
However, the prevalence of combine harvesters also seems to reflect
the difficulties which large commercial farmers have in supervising
large numbers of hand harvesters. This is consistent with the earlier
argument that labour shortages are more of a problem for these kinds
of farmers. In some regions of Morocco, despite stagnant or falling
real agricultural wages, combine harvesters have spread rapidly on
large farms. USAID (1986) reports that combine harvesting cost only
25% as much as hand harvesting in that country. It is likely, however,
that this calculation applies more to large farms than to small, as
argued in the previous section. Even in Jordan, whose labour force has
declined markedly, in 1985, less than 40% of small-scale
owner-operators harvested by combine, compared to 100% of larger
commercial farmers (Aydin, 1989) (2). It is clear that mechanization
of harvesting is far less advanced than tractorization in most
countries. Finally, few countries of the region have proceeded very
far with the mechanization of control-intensive operations, which
continue to be largely performed by hand labour.

This pattern of mechanization is consistent with the hypothesis
that machines are substituting mainly for animal power rather than for
human labour (e.g., Egypt, Syria, Turkey). A major reason for this is
the very rapid rate of growth of demand for livestock products, which
in turn is the result of both growth in per capital incomes and changes
in local food preferences. From 1973 to 1980 the consumption of meat
and milk in the region grew annually at 4.6%, well above the popula-
tion growth rate, due to the regional preference for animal products,
whose income elasticity exceeds one in most countries (Sherbiny,
1984). Average incomes have risen rapidly in most countries of the
region during the past 15 years, and this has accelerated the growth
of demand for animal products. Consequently the opportunity cost of
using animals for work has increased (working animals produce less
milk and meat than animals kept in stalls), providing incentives to
mechanize power-intensive operations like land preparation.

(1) By contrast, he calculated that 19% of the net income of partially
mechanized farmers (i.e., tractors only) was due to subsidies.
(2) By contrast, over 90% of such smaller farmers prepare their land
using tractors.

- 35 -

This does not mean that migration has not contributed to the
observed pattern of mechanization. Because many farm operations
require much physical force (e.g., plowing behind oxen), social
customs have often allocated these tasks mainly to adult men, who
constitute the overwhelming majority of migrants. Migration also
contributes to mechanization by providing many rural families with
additional income from remittances. When a portion of these higher
incomes are spent on meat, this contributes to the incentives for
other farmers to mechanize plowing, freeing their animals to produce
more milk, cheese, and meat.

Mechanization has also been encouraged by government policy.
Most countries of the region directly subsidize tractorization,
usually by offering loans at less than the economy-wide opportunity
cost of capital and cheap diesel fuel. The Agricultural Bank of
Sudan, for example, charges 14% on tractor loans; with a rate of
inflation of about 30%, this yields a real interest rate of 16%
(Ibnouf, 1985). The private real cost of capital on loans for tractor
purchase in Egypt has also been well below zero in Egypt. Diesel fuel
prices in the early 1980s in Egypt were only 10% of international
prices. Imported tractors are either directly subsidized, as in Saudi
Arabia and Libya, or are indirectly encouraged by overvalued real
exchange rates. For example, in 1982 the Moroccan government began to
promote farm mechanization by removing all duties and taxes on
agricultural equipment and by expanding credit for such purchases.
The real cost of a tractor fell by 30% and not surprisingly, the
numbers of tractors rose from 23,000 in 1978 to 40,000 in 1986 (USAID,
1986). These direct subsidies are complemented by the paucity of
research into improved animal-drawn implements (e.g., in Turkey, World
Bank, 1983). These policies, combined with growing incomes, more
abundant capital, accelerating demand for animal products, as well as
the departure of large numbers of adult males, have contributed to the
very rapid mechanization of power-intensive operations throughout the

In summary, farm mechanization in the Middle East and North
Africa has so far been mainly of the "power-intensive" tasks.
Mechanization of "control-intensive" operations has made relatively
little progress, while there are substantial differences across
countries in the extent of mechanization of grain harvesting. These
latter differences appear to be primarily explainable by labour market
changes, with the exception of Sudan, where government subsidies, not
labour shortages, drive the adoption of combines.


Remittances are fundamental to the emigration process. The
impacts of these remittances on agricultural development may be
divided into macro- and micro-economic impacts. The former apply only
to international migration.

- 36 -

A. Macro-Economic Impacts

The most important macro-economic effect is the impact on the
balance of payments. Official remittances are the principal source of
foreign exchange in the Yemen Arab Republic; they exceed the value of
commodity exports in Egypt and Jordan, and constitute about 40% of the
value of exports in Sudan, 25% in Morocco, and over 15% in Turkey.
However, much more money enters these countries through unofficial
channels (Choucri, 1986). These foreign exchange inflows may increase
investment and the demand for food, especially those with a high
income elasticity of demand (e.g., fruits, vegetables, and livestock
products), creating opportunities for some farmers. However,
remittances may simply be spent on imported goods, especially consumer
durables. To the extent that remittances enter the economy outside of
the banking system, the government may lose control of credit and the
money supply. This may compound inflationary pressures. However,
spending remittances on imported goods may reduce inflation by
expanding domestic supply and by increasing competition in local
output markets. A study of Pakistan found an inverse correlation
between changes in the price level and in remittances (Kazi, 1988).

B. Micro-Economic Consequences: How Are Remittances Spent?

The principal area of debate on the micro-economic effect of
remittances concerns the division of the funds between consumption and
investment, and the type of investments which are selected. Emigrants
and their families spend their money "rationally". A substantial
proportion of remittances are certainly spent on consumption. Such
spending improves the welfare of the migrants' families.

Most remittances not spent on direct consumption are used to buy
housing, land, and education. Relatively little appears to be spent on
direct agricultural investment, although there seems to be
considerable variation across countries and areas within countries

Remittances are often spent on housing and land. One survey in
Egypt found that over one-fifth of all migrants spent their money in
this way (El-Dib, et. al., 1984); even poorer emigrants from Cairo try
to buy plots on the edge of the city. About one-sixth of Pakistani
remittances are spent on real estate (Kazi, 1988). Parallel patterns
have been found in Morocco, Sudan, Tunisia, and Turkey. Other priority
items of expenditure include furniture and simple household articles
(some 58% in one Egyptian case study Egyptian Ministry of Labour,
1984). Better housing improves the quality of life of rural people and
contributes directly to meeting basic human needs. To the extent that
poorly ventilated, poorly heated or excessively hot, vermin-ridden
housing is a threat to human health, the spending of remittances on
better housing may be viewed as an investment in human capital. More
direct investment in human capital, in children's (especially sons')
education is also a prominent use of remittances in many countries.

To the extent to which remittances are directly spent on
non-real estate, agricultural investment varies widely from country to

- 37 -

country, and indeed from village to village. Many believe that
relatively few remittances are spent on agricultural machinery or on
other productivity enhancing investments. However, it is notable that
in the Tihama region of the Yemen Arab Republic, where large landlords
have owned diesel irrigation pumps since the 1950s, now many smaller
farmers have been able to acquire them as the result of earnings in
Saudi Arabia (Tutwiler, 1989). A study of Turkish migrants in a
relatively fertile area of that country found that by 1975 one-third
of all tractors in the district of Bogazliyan had been purchased by
emigrants (compared with none in 1966) (1). A comparison of villages
with many emigrants with those with few suggested that there was a
correlation between tractorization and emigration, largely explained
by the purchase of tractors by migrants (2). By contrast, few studies
conducted in Egypt have shown any significant spending of remittances
on non-real estate agricultural investment; Jordanian migrants also
seem to eschew investment in farming (Aydin, 1989).

We hypothesize that differences in farm policy and in production
possibilities are the main determinants of the use of remittances for
direct investment in agriculture. Migrants' failure to invest in
highly risky, ecologically marginal rain-fed farming areas is hardly
surprising. Their preference for non-agricultural investments (e.g.,
transportation, small workshops, and retail distribution) suggests
that the private returns to these activities exceed those to
agricultural ones. Agricultural investments will only occur if they
are profitable, if available technology is appropriate for local
conditions and if prices of outputs are not kept low by competition
from imports and/or government taxation. Studies in the Yemen Arab
Republic have found that the basic determinant of whether remittances
were spent on farm machinery was topography: in valley villages, where
machinery could be used, remittances were used to buy tractors (or in
the Tihama, where irrigation was feasible, migrants bought irrigation
pumps). In mountain villages with narrow terraces on steep slopes,
tractors and other machines were impractical, and were not purchased.
The Turkish example cited above is a further case in point: the study
area was in a relatively fertile area of Western Turkey, and Turkey
pursued policies which were quite favorable to agriculture during the
1970s. Remittances will be spent on agricultural investments if those
investments are viable. The task of government policy is, therefore,
to raise the profitability of such investments and to identify other
areas of viable investments.

(1) In one case, the authors reported a case of a Turkish worker who
had used his savings to buy a Mercedes (the sort of thing analysts
often decry). He then sold the Mercedes in Istanbul and used the
money to buy a combine harvester (Abadan-Unat, et. al., 1976).
(2) It should be remembered, however, that investment in tractors does
not normally increase land productivity directly. The same study
also notes little difference in the crop yields of migrant vs.
non-migrant land, and between mechanized and non-mechanized farms.

- 38 -

One challenge for policy makers is to determine how to channel
remittances into needed public and semi-public goods and services,
especially in rural areas. It is difficult to entice private funds
into irrigation works, roads, public health facilities, schools, etc.
because of the "free rider" problem. One promising attempt to solve
this problem is in the Rural Development Associations of Yemen Arab
Republic, in which zakat (Islamic tithe), levied on all income, is
allocated to wells, schools, and roads.

The fact that a large proportion of remittances are not
transferred through the national banking system reduces the percentage
of remittances used for productivity enhancing investment. Weak and
unstable financial systems may have discouraged people from using
banks. Returned migrants often keep their savings literally under the
mattress; although they save, they cannot be expected to create their
own investment opportunities. Overvalued foreign exchange rates
further dissuade international migrants from forwarding their savings
through official channels. Policy makers can, therefore, increase the
use of remittances for investment by strengthening and stabilizing the
national banking system and by maintaining realistic foreign exchange
rates. Recent exchange rate reforms in Egypt provide an example of
what policy reform can accomplish: the realignment of exchange rates
in the summer of 1987 led to an increase in bank deposits of perhaps
$750 million. Many of these funds were remittances of Egyptian workers
abroad. Kazi estimates that about 70% of Pakistani remittances enter
the country through the banking system; it may not be accidental that
a relatively high 18% of remittances are used for non-real estate
investment there. Sound banking systems, and realistic foreign
exchange rates, would enhance the contribution of remittances to

C. Micro-Economic Consequences of Remittance Spending

It is crucial to note that the spending of remittances may have
very important indirect effects on agricultural development. First,
remittances to families may raise the reservation wage of labour.
Second, they increase the demand for rural, non-farm goods, which in
turn raises employment in this sector. Third, remittances may push up
land prices.

By providing additional income to relatively poor families,
remittances raise the reservation wage of hired labour. There is
evidence of this phenomenon in Egypt and the Yemen Arab Republic, and
certain areas of the Sudan. Remittances may also change the perceived,
subjective "rates of return" to family labour. This is likely to be
especially true in areas with low productivity due to natural
conditions (i.e., low and fluctuating rainfall), on farms for which
there are few technological possibilities for substituting machinery
for labour, and on farms producing crops which fetch low prices.
Finally, both the reservation wage for hired labour, and the
subjective rate of return for labour on family farms may be affected
by the investment of remittances in off-farm activities, investments
which increase the demand for (and returns to) rural, non-farm labour.

- 39 -

Second, the spending of remittances on non-agricultural rural
activities creates numerous employment opportunities which compete
with agriculture. The employment multiplier effects may be quite
large, because favoured uses of remittances, such as housing and
furniture, are very labour intensive. The rapid growth of off-farm
employment, which we have seen has had important consequences for farm
labour markets, has been fueled by remittances in many parts of the

Remittances are frequently invested in land. Since the supply of
land nearly everywhere in the region is highly inelastic, the
increased demand has pushed land prices sharply upwards. Of course,
not all of the increase in land prices can be attributed to
remittances. The increase in land prices seems to have occurred
primarily in the highly inflationary environment during the decade of
the oil boom. Investors usually move into real estate in inflationary
periods. Land is often also an attractive investment for social status
reasons, or, most importantly perhaps, because of the lack of
profitable alternative investments (e.g., in the Yemen Arab Republic -
Swanson, 1979). More recently, in many countries of the region, land
prices have softened considerably. For example, the average purchase
price of irrigated land in Jordan fell by about 40% between 1983 and
1986 (FAO data tape). Land prices in Egypt have also declined, partly
because of new, more stringent government regulations against building
on agricultural land.

The main consequence of remittance-fueled land price increases
for agriculture is likely to be felt at the urban-rural geographical
margin: the process of converting prime agricultural land into urban
real estate. Urbanization, itself largely the result of rural-urban
migration, contributes to this phenomenon, as do increased incomes. We
need detailed studies to try to determine the extent to which the loss
of high quality farm land in the region has been abetted by


Rural-urban migration has some special consequences for
agriculture, in addition to its contribution to changes in rural
labour markets as examined above. First, rural and urban food
consumption patterns often differ, partly as a result of differences
in income, but also because of different technologies of food
preparation and food availabilities. Rural-urban migration may
therefore shift the pattern of demand among food commodities. In
particular, such migration has contributed to the shift in demand
toward bread and wheat, and away from other cereals such as sorghum
and seminola. If wheat production faces ecological and/or economic
constraints, such a shift may increase the need to import wheat,
exacerbating problems of the balance of trade. The greater ease of
distributing imported wheat in urban areas as contrasted to
domestically produced cereals (See, e.g. FAO, 1984) creates a further
danger for rural-urban migration to worsen the trade balance.

- 40 -

Second, rural-urban migration may further entrench costly food
subsidy systems. Most subsidized food is distributed in urban areas,
although in some countries (e.g., Egypt) subsidies also reach the
countryside. It is possible that food subsidies and rural-urban
migration are mutually reinforcing. On the one hand, migration
increases the size of the subsidy bill directly. Such migration also
makes modification of existing subsidy systems more difficult. The
more urbanites who are covered by subsidies, the larger the vested
interest which will oppose any modification in the system. On the
other hand, some have argued that the presence of food subsidies
promotes rural-urban migration by increasing the rural-urban real wage
gap (Parker and Coyle). This is certainly plausible, but we need to
know more about the magnitude of the impact of urban food subsidies on
the rural-urban real income gap.

What are the implications of food subsidies for agricultural
development in the region. Of course, this issue has received much
attention (see, e.g. FAO, 1988); we shall only very briefly review
the debate here. First, it is often argued that subsidy systems are
"too expensive". Of course, we must ask, "relative to what?" If we
compare food subsidy expenditures with total public spending, we find
that food subsidies are a relatively small (usually less than 10% and
often considerably smaller e.g., 1.7% in Jordan in 1981) component
of national budgets (FAO, 1988). On the other hand, food subsidies as
a percentage of budgetary deficits are, of course, much larger, e.g.,
30% in Egypt in 1987. Such a comparison is unfair, however, since it
implicitly assumes that food subsidy expenditures are a "total loss"
to the nation. This is far from true; numerous studies have
documented the importance of such subsidies for maintaining the
nutritional levels of the poor. The problem, of course, is that
current systems are a rather inefficient mechanism for attaining this
very important goal. And, as noted above, rural-urban migration may
complicate the implementation of necessary reforms.

The long-run consequences of food subsidies may be more
important than their contribution to annual budgetary deficits.
Subsidy funds have a high opportunity cost, typically taking the form
of foregone investment in the agricultural sector. It has been shown
for Egypt that an increase in food subsidies has been associated with
a decline in agricultural investment (Scobie, 1985). Subsidies are
often a significant (25-33) percent of planned agricultural
expenditure. This plausibly reduces the rate of agricultural growth,
a problem which will be made worse if governments finance subsidies
through lowered producer prices.

In this "worst case" scenario, we have rural-urban migration
raising the demand for food, reducing agricultural labour supplies,
contributing to weakening government investment in agriculture, and
reinforcing disincentive pricing policies.

Finally, rural-urban migration may reduce the supply of
high-quality farm land. It is estimated that Egypt lost 60,000 acres
of land per year to city and village growth during the late 1970s and
early 1980s. The problem may be especially acute around those cities

- 41 -

which were built in (scarce) well-watered areas, as, for example, the
Oasis of Damascus. High quality farm land is very scarce in many
countries; its loss to cities is very detrimental to meeting food
security goals. This is a very worrisome consequence of urban growth,
at least half of which may be attributed to rural-urban migration.
Most countries have laws and regulations which attempt to limit
encroachment on arable land; the problems seem to be the difficulty of
enforcing existing legislation.


There is considerable controversy on the question of how
emigration has affected the distribution of income. Poor people often
migrate, but the very poorest usually cannot; analysts stressing the
former aspect usually argue that migration is equalizing, while those
stressing the latter typically hold that migration widens income gaps.
Unfortunately, there are relatively few systematic studies which can
help us to provide answers.

Migration offers a clear avenue of upward mobility for those
able to emigrate and contributes to the survival of many small family
farms. Access to migration opportunities has become an important
variable in class differentiation, alongside more conventional factors
like ownership of land and access to education. This makes the class
stucture more complicated, and almost certainly, more fluid than

A recent study by Fergany in Egypt compares the Gini
coefficients of income distribution of migrants and non-migrants
before (1973) and after (1984) migration. He finds that the Gini
coefficient fell for non-migrants but rose for migrants. He concludes
that "labour migration during the period (1973-1984) militated in the
direction of increasing income inequality in Egypt" (p. 13).

There seems to be a difficulty with this inference: it assumes
that the increased equality of "stay-at-homes" is independent of
emigration. That is possible, but it is certainly not obvious:
plausibly, the exit of some workers and the spending of remittances on
rural housing etc. dramatically increased the demand for unskilled
labour, increasing the wages and incomes of those at the bottom of the
income scale. Further, from a national perspective, the fact that a
disproportionate number of emigrants are from rural areas should
itself increase equality at a national level, since average urban
incomes exceed those in rural areas by 200 to 300%.

The impact of emigration on poverty is clearer than its effect
on income distribution or the class structure. Emigration and
remittances have reduced rural poverty, the worst poverty in the
region. In Egypt, for example, emigration dramatically reduced the
ranks of landless labourers, historically the poorest of the poor in
that poor country. Although roughly two of every five rural Egyptian
families were landless in 1961, that percentage had declined to 15% by
1975 (Commander, 1987); by the mid-1980s, it was difficult to find

- 42 -

landless farm workers in many villages in the Delta. An ILO survey
found that although 23% of all Egyptian villagers received
remittances, 75% of the poorest 30% of villagers received them. For
the absolutely poorest people, who are often disabled, single persons,
remittances constituted more than 75% of their (extremely meager)
incomes. For the third poorest income decile, 40% of their income
came from remittances. The poorest fourth and fifth deciles obtained
more than 30% of their incomes from agricultural wage labour, whose
wages more than doubled in the last decade, thanks to emigration and
the expansion of remittance-fueled non-farm rural employment (Radwan
and Lee, 1985; Richards and Martin, 1983).

Similarly, emigration and remittances have ameliorated the
serious rural poverty in the Yemen Arab Republic. Nearly one-third of
all rural incomes now come from remittances (World Bank, 1986).
Researchers report improved food and nutrition and better housing
there. In some areas, sharecropping contracts now commonly allocate a
higher percentage of the crops to tenants. This constitutes evidence
that poorer groups have benefitted not only absolutely, but also
relatively. Emigrants disproportionately come from poor, disadvantaged
regions of the country, especially relatively remote, low rainfall
areas of the Central Plateau (Tutwiler, 1989). Similar "geographical
selectivity" is found elsewhere: Western Sudanese, Upper Egyptians,
and Southern Tunisians all participate extensively in emigration (1).

However, migrants are usually not the very poorest rural
inhabitants, simply because migration requires some money to finance
the journey abroad. Labour contractors and family networks can often
bring these costs down to within the reach of quite poor Middle
Easterners. The very poorest people may benefit directly (e.g.,
remittances in Egypt) or indirectly (increased job opportunities) from
the emigration of others. Nevertheless, migration is unfortunately
neither large nor stable enough to alter radically the serious
problems of rural poverty in the region.


There exists a debate on the impact of migration on land tenure.
The competing hypotheses resemble those in the controversy over the
impact of migration on income distribution. On the one hand, there
are those who believe that migration polarizes the agrarian structure.
The departure of adult males makes it harder for their families to
maintain farm production. Sons are unwilling to remain in

(1) This is precisely what economic theory would predict, of course:
the decision to migrate is a direct function of the difference in
the expected wage abroad and at home. Poor, isolated, and dry
areas have low agricultural productivity and low wages: we would
expect them to have a high propensity to migrate. It is perhaps
worth noting that the same phenomenon is easily observed elsewhere
in the world: the arid centre-north of Mexico provides the
majority of rural migrants to the U.S.

- 43 -

the countryside, and when their fathers are no longer able to farm,
the family sells its land. The buyers are the more successful
farmers. The spread of mechanization, which is variously treated as
independent of or linked to emigration from the rural areas, increases
the optimal farm size, further contributing to polarization.

On the other hand, many analysts point out that there is little
evidence to support the argument that average farm size is growing in
the region. (Indeed, the available evidence suggests the contrary.)
Further, it is often argued that remittances shore up otherwise
unviable farms, enabling small farms to survive longer (see the
evidence in Radwan and Lee, 1985). Similarly, most small farm
families engage in hiring out some family members' labour for at least
part of the year; any increase in wage rates caused by the emigration
of others would raise the incomes of those remaining behind. The
increased demand for rural, non-farm labour, also stimulated by
remittances, has a similar effect; as sources of income diversify, it
becomes less imperative for poor farmers to sell small land parcels to
survive. The increased demand for products such as livestock in which
small farmers may have a comparative advantage also impedes
polarization. Finally, if migration merely draws off some of the rural
surplus labour, then continued population growth, coupled with Islamic
inheritance law, will probably continue the historical trend toward
farm fragmentation. At a minimum, in such a "limited migration"
context, it would be difficult to argue that migration promotes

This author believes that the available evidence supports the
second hypothesis. We would expect considerable regional and local
variation here, however. Further research on these competing
arguments would be very helpful.

The impact of migration on tenancy arrangements is equally
unclear. Some observers believe that the decline in labour might
foster sharecropping, because the increasing scarcity of labour raises
search costs. Increasingly scarce labour should also change the terms
of sharecropping contracts, with labour receiving a higher share. On
the other hand, spreading mechanization may reduce the incentives for
sharecropping arrangements. To the extent that landlords respond to
increased demand for output by supplying greater quantities of modern
inputs, their share would rise; there is some evidence of this
phenomenon in the Yemen Arab Republic (IFAD, 1988). Given the
theoretical controversies over the causes and consequences of
sharecropping, and the paucity of empirical studies in the region, it
is not surprising that no firm conclusions may be drawn on the impact
of labour emigration on sharecropping arrangements. This paper adopts
an agnostic attitude on the question: in the absence of further work,
we suspect that emigration plays a relatively minor role in changes in
tenancy arrangements.

Finally, we may ask what impact labour emigration may have on
village social organization. Again, few firm conclusions are
warranted, but some hypotheses for future investigation may be
advanced. First, the relative importance of land ownership in social

- 44 -

stratification will fall with increased migration. This is because
alternative sources of income (remittances, returns of labour and
capital in non-farm, rural enterprises) will become increasingly
important. Second, the power of heads of households in extended
families should weaken, as young men obtain sources of income
independent of their fathers' farms. Third, as explored above, there
are likely to be major changes in the economic and social role of
women in the village. There may be an increase in de facto
female-headed households. Finally, migration will foster a whole
series of subtle changes in village attitudes by bringing villagers in
contact with information, commodities, and ideas from a wider world.


The three types of migration surveyed in this paper have had
numerous, diverse consequences for agricultural development in the
region. Principal impacts include: changes in labour markets, shifts
in demand patterns, and implications for investment. The diversity of
the impacts among the types of migration and across countries,
combined with the fact that all types of migration may have both
positive and negative impacts for agriculture and rural poverty
alleviation, suggests that policies will vary among countries.
Nevertheless, some generalizations may be made.

First, all countries need a better "data base" for making
appropriate decisions. In particular, the extent of labour-market
segmentation and therefore the consequences of labour emigration for
farm labour supply is really not well known in most countries. We
would propose that a series of studies be undertaken to test the
hypotheses offered above in Section III. The consequences of the
"feminization" and ageing of the farm labour force deserve particular
attention. The extent to which rural-rural migration can ease
localized labour shortages is also a subject worthy of close

Second, those policies which maximize the benefits of
international labour migration and which reduce the costs of
rural-urban migration are the same policies which governments should
adopt in any case if they wish to accelerate the pace of agricultural
development. Policies should not penalize the agricultural sector,
but should promote the efficient use of resources, and accelerate the
rate of growth of output and employment.

More specifically, six policies stand out.

1) Governments should maintain realistic foreign exchange rates

Overvalued real exchange rates penalize agriculture by shifting
the terms of trade against farmers. This not only weakens production
incentives, but also contributes to rural-urban migration. Further,
overvalued exchange rates discourage international migrants from
remitting their savings through the banking system.

- 45 -

2) Governments should offer farmers stable and favourable
output prices

Farm prices which are on average equal to world prices and which
fluctuate relatively little from year to year are necessary to provide
incentives for farm investment and to reduce the difficulties which
farmers face when labour costs rise. Of course, such policies also
improve agriculture's terms of trade, and therefore help to slow down
rural-urban migration. Fortunately, many countries of the region have
recognized the importance of favourable producer's prices, and have
improved farm prices.

3) Governments should target urban food subsidies to ensure
that they improve the nutrition of the poor at a manageable

Very large food subsidies divert funds from agricultural and
other employment-creating investment. They contribute to budgetary
deficits and foster rural-urban migration. Governments need to
continue to devise mechanisms to ensure that subsidies meet their
legitimate goal of protecting the nutrition of the poor at manageable

4) Governments should develop appropriate farm technology
packages and step up investment in rural infrastructure

Technological change is the key to successful development.
Existing technology is quite appropriate for many areas. However,
there exist regions where this is not true, for example, tractors in
very mountainous regions. Here a greater research effort into new,
appropriate technologies is needed. Stemming the rural exodus, and
encouraging private investment in farming requires complementary
public investment in roads, communications, electricity, and health,
educational, and recreational facilities in the countryside.

5) Governments should encourage the integration of local
labour markets by improving communication and information

The fundamental cause of labour market segmentation is imperfect
information. Governments can reduce this problem by improving
transportation and communication, and by gathering and publicizing
information about local labour market conditions.

6) Governments should avoid costly subsidies to inappropriate

Mechanical technology should not be subsidized. Such subsidies
are unnecessary and unwise. They are unnecessary, because if farmers
face realistic prices, they will respond to any increases in labour
costs created by emigration by buying machines on their own. Such
subsidies are unwise, because they will contribute to employment
problems. Subsidizing the mechanization of "control-intensive"
operations such as grain harvesting is especially dangerous in the
current conjuncture of austerity and heavy foreign indebtedness. The

46 -

inexorable increase of the labour force suggests that employment
creation must be at the core of development strategy.

None of these policies constitutes a response specifically and
exclusively to migration; they would be found in any sensible
development strategy, regardless of the volume of migration. But they
can all contribute to maximizing the net benefits of labour migration
in the region.


Arab Migrant Workers in the Near East Region, 1975

Country or area of origin and percentage distribution among countries of employment

Country Yemen Yemen
of employment Egypt AR Jordan PDR Syria Lebanon

No. Z No. Z No. Z No. Z No. Z No. Z

Bahrain 1 237 0.3 1 121 0.4 614 0.2 1 122 1.6 68 0.1 129 0.3
Iraq 7 000 1.8 5 000 1.9 3 000 6.0
Jordan 5 300 1.3 20 000 28.4 7 500 15.1
Kuwait 37 558 9.4 2 757 1.0 47 653 18.0 8 658 12.2 16 547 23.4 7 232 14.6
Libya 229 500 57.8 14 150 5.3 13 000 18.5 5 700 11.5
Oman 4 600 1.2 100 0.0 1 600 0.6 100 0.1 400 0.6 1 100 2.2
Qatar 2 850 0.7 1 250 0.4 6 000 2.3 1 250 1.8 750 1.1 500 1.0
Saudi Arabia 95 000 23.9 280 400 96.6 175 000 66.1 55 000 77.9 15 000 21.3 20 000 40.3
UAE 12 500 3.1 4 500 1.6 14 500 5.5 4 500 6.4 4 500 6.4 4 500 9.0
Yemen AR 2 000 0.5 200 0.1 150 0.2 -

Total: 397 545 100.0 290 128 100.0 264 717 100.0 70 630 100.0 70 415 100.0 49 661 100.0

of migrants by
country of
origin 30.7 22.4 20.4 5.5 5.4 3.8

Source: Birks,
in the
a wide

J.S. and Sinclair, C.A., "International Migration and Development
Arab Region", ILO, Geneva, 1980, pages 134-135. (The authors used
variety of official sources for their estimates.)


TABLE 1 Cont'd.

Country or area of origin and percentage distribution among countries of employment

of employment Sudan Tunisia Oman Iraq Somalia Morocco Total

No. Z No. Z No. % No. Z No. Z No.

Bahrain 400 0.9 1 383 3.6 126 0.6 6 200

Saudi Arabia
Yemen AR


7 000
15 000
1 500



38 500



-- -
3 660 9.5

1 870 4.9
.7 500 45.6
L4 000 36.4

17 999 87.3

2 000 9.7
500 2.4




2 500

1.8 143
98.2 310

Total: 45 873 100.0 38 649 100.0 38 413 100.0 20 625 100.0 6 547 100.0 2 547 100.0 1 295 750

of migrants by
country of
origin 3.5 3.0 3.0 1.6 0.5 0.2 100.0

Source: Birks,
in the
a wide

J.S. and Sinclair, C.A., "International Migration and Development
Arab Region", ILO, Geneva, 1980, pages 134-135. (The authors used
variety of official sources for their estimates.)

- -

- 49 -


Estimates of the Numbers of Arab Migrant Workers in Saudi
Arabia, and the GCC States in the early 1980s

Country of At Work in Total in GCC*
Origin Saudi Arabia States

Egypt 800 000 1 150 000
Iraq** 3 250 44 760

Jordan 140 000 227 850

Lebanon 33 200 54 850

Oman 10 000 33 450

Somalia 8 300 12 200
Sudan 55 600 65 470

Syria 24 600 67 150

Tunisia/Morocco 500 920

Yemen AR 325 000 336 145

Yemen PDR 65 000 83 845


Total: 1 465 250 2 076 640

GCC: Gulf Cooperation Council includes: Saudi Arabia, Kuwait, UAE,
Qatar, Oman, Bahrain.

** In 1980 there were approximately 750 000 foreign workers in Iraq.
("Labour Immigration in the Gulf States: Patterns, Trends and
Prospects", Quarterly Review of the Inter-governmental Committee
for Migration, January 1986).

Source: Owen, R. (1985). "Migrant Workers in the Gulf", Minority
Rights Group, London. Report no. 68.


Numbers of Foreign Workers in Gulf States by Country of Origin, 1985 1/


Total Egypt Jordan Yemen AR Yemen PDR Others
No. Z No. Z No. Z No. Z No. Z No. Z
Bahrain 85 2.3 4 0.5 4 1.5 5 0.6 2 2.0 70 3.9
Kuwait 430 11.5 137 18.4 70 26.2 24 3.0 11 11.0 188 10.3
Oman 145 3.9 7 0.9 7 2.6 3 0.4 1 1.0 127 7.0 ,
Qatar 108 2.9 16 2.2 8 3.0 8 1.0 3 3.0 73 4.0
Saudi Arabia 2 500 67.0 500 66.9 158 59.2 720 90.0 75 75.0 1 047 57.7 o
UAE 462 12.4 83 11.1 20 7.5 40 5.0 8 8.0 311 17.1

Total: 3 730 100.0 747 100.0 267 100.0 800 100.0 100 100.0 1 816 100.0
Z of Total: 100 20 7.2 21.4 2.7 48.7
1/ The number of migrant workers in other oil-exporting countries in the Region is approximately
estimated at 1.3 million.
Source: Economic and Social Commission for Western Asia (ESCWA), "Study on Impacts of Returning
Migration in Selected Countries of the ESCWA Region", DPD/86/14, UN, March 1986, page 19
(in Arabic).

- 51 -


Rural-Urban Migration as Percentage of Urban Growth
(Assumes Rural and Urban Population Growth Rates are the Same)

Country 1965-80 1980-85

Algeria 21% 11%
Egypt 17 18
Iran 42 33
Iraq 36 43
Jordan 51 7.5
Kuwait 15 14
Lebanon 65 n.a.
Libya 54 42
Morocco 40 40
Oman 56 34
Saudi Arabia 41 30
Sudan 58 44
Syria 24 35
Tunisia 50 38
Turkey 44 43
UAE 16 n.a.
Yemen AR 74 66
Yemen PDR 38 47

Lower Income LDCs 25% 33%
Lower-Middle Y LDCs 44 32
Upper-Middle Y LDCs 42 38
High Y-Oil Exporters 45 28

Source: Calculated from World Development Report,
1987, pp.266-67; 254-55.

- 52 -


Estimates and Projections of the Agricultural Labour Force in
Selected Countries of the Near East Region, 1960-2000


Country 1960 1970 1980 1985 1990 2000

Algeria 1 990 1 394 1 262 1 301 1 342 1 387
Egypt 4 364 4 765 5 158 5 526 5 902 6 786
Iran 3 220 3 547 4 026 4 082 4 199 4 331
Iraq 972 1 125 1 081 1 043 1 049 1 073
Jordan 199 162 66 63 59 51
Lebanon 199 130 106 90 86 69
Libya 200 150 137 131 127 118
Morocco 2 195 2 333 2 594 2 746 2 860 2 950
Oman 97 102 140 163 163 165
Saudi Arabia 817 1 019 1 333 1 490 1 599 1 729
Sudan 3 376 3 601 4 331 4 606 4 866 5 348
Syrian AR 685 785 707 713 747 848
Tunisia 663 559 668 648 630 549
Turkey 10 991 11 361 11 146 11 385 11 418 11 335
Yemen AR 973 1 024 1 013 1 103 1 226 1 561
Yemen PDR 201 208 199 203 206 208

Total: 31 142 32 265 33 967 35 293 36 479 38 508

"World-Wide Estimates

Agricultural and Non-Agriculti
1920-2025", Statistical Anal
Division, Economic and Social Po

and Projections of the
rural Population Segments,
ysis Service, Statistics
licy Division, Rome, 1986.

Source: FAO,

- 53 -


Ratio of Arable Land to Economically Active Population
in Agriculture, 1960-86 a7


Country 1960 1965 1970 1975 1980 1985 1986

Algeria 4.00 4.07 4.88 5.53 5.95 5.77 5.76
Egypt 0.59 0.60 0.60 0.58 0.47 0.45 0.45
Iran 4.74 4.57 4.43 4.34 3.41 3.63 3.61
Iraq 4.84 4.72 4.44 4.93 5.04 5.22 5.22
Jordan 1.68 1.88 2.28 3.21 6.03 6.66 6.77
Libya 9.85 10.90 13.50 13.98 15.18 16.26 16.45
Morocco 3.18 3.19 3.22 3.22 3.09 3.06 3.06
Saudi Arabia 0.83 0.85 0.85 0.96 0.83 0.79 0.79
Sudan 3.20 3.23 3.25 3.08 2.87 2.71 2.68
Syrian AR 9.32 9.45 7.53 7.21 8.04 7.89 7.87
Tunisia 6.41 7.04 8.01 7.93 7.60 7.60 7.29
Turkey 2.29 2.36 2.41 2.42 2.56 2.42 2.42
Yemen AR 1.27 1.22 1.28 1.39 1.33 1.22 1.20
Yemen PDR 0.57 0.59 0.63 0.72 0.78 0.82 0.82

a/ Arable land includes permanent crops.

Source: Calculated from FAO data:

- Arable land: FAO printouts on land use; Commodities

and Trade Division; ICS
and Processing System of
available as of February

:Interlinked Computer Storage
FAO Commodity Data); data

- Economically active population: FAO Production
Yearbook 1985, vol. 39, Rome, 1986.


Indices of Real Wages in Agriculture in Selected Countries of the Near East Region

Turkey Morocco Egypt

Index Index Index Index Index of
Year No.1 a/ No.2 b/ Year No.1 c/ No.2 d/ Year real wages

1977 100 100 1975 92 87 1960 100
1978 106 106 1977 83 77 1974 110
1979 105 111 1979 89 85 1975 154
1980 84 92 1980 89 86 1976 174
1981 85 91 1981 95 90 1977 192
1982 85 89 1982 99 92 1978 208
1983 94 97 1983 112 105 1979 236
1984 84 82 1985 102 95 1980 208
------------------------------------------------------------------------------------- 1
a/ Agricultural daily wage deflated by current price index (CPI) for low-income
residents in Ankara, Turkey.

b/ Agricultural daily wage deflated by food price index (FPI) for low-income
consumers in Ankara, Turkey.

c/ Agricultural minimum wage deflated by general CPI.

d/ Agricultural minimum wage deflated by general index of food prices.

Sources: Turkey: From FAO data tapes.

Morocco: Calculated from World Bank, "Staff Appraisal Report: Kingdom of
Morocco, 6th Agricultural Credit Project", Report No. 6094 MOR,
May 28, 1986.

Egypt: Amin, G. and Awani, E.T., "Migration of Egyptian Labour: A Critical
Review of Literature on Egyptian Labour Migration", Centre for
Research on International Development,
Memo No. 108-A, 1986 (in Arabic).

- 55 -


Female Participation in the Agricultural Labour Force
in Selected Countries of the Near East Region, 1960-1985

Percentage of Farm Labour which is female

Country 1960 1970 1980 1985

Iran 5.0 9.4 20.8 22.1
Iraq 2.0 5.3 41.0 41.0
Jordan 3.9 4.1 0.9 1.1
Lebanon 17.5 22.7 32.2 37.7
Libya 1.7 7.2 16.0 17.7
Morocco 2.8 10.4 14.1 16.0
Oman 3.1 2.9 2.9 2.6
Saudi Arabia 3.8 3.8 3.2 3.2
Sudan 23.3 23.8 24.1 26.4
Syrian AR 9.2 13.4 27.4 27.5
Tunisia 1.5 5.5 19.9 20.8
Turkey 49.9 48.9 51.6 54.3
Yemen AR 4.6 5.5 9.4 10.2
Yemen PDR 11.2 13.9 13.0 13.6

Source: FAO, "World-wide Estimates and
Agricultural and non-Agricultural
1950-2025", Statistical Analysis

Projections of the
Population Segments,
Service, Statistics

Division, Economic and Social Policy Division, FAO,
Rome, 1986.

- 56 -


Rates of Growth for Different Product Groups
Near East, 1975-86*

Rate of Growth of
Product Group Production (%/yr)

Cereals 2.3%

Fruit 2.8%

Vegetables & Melons 3.7%

Meat 4.5%

Eggs 8.0%


"Near East" excludes Morocco, Algeria & Tunisia.

Source: Calculated from FAO Production Yearbooks


Ratio of Arable Land per Tractor (ha/tractor)

--------------- --------------------------------------------------
Country 1960 1965 1970 1975 1980 1985
Algeria 248 226 148 171 171 143
Egypt 219 156 124 63 54
Iran 1 359 984 781 404 236 153
Iraq 1 028 537 366 268 245 179
Jordan 882 635 477 365 305 274
Libya 885 776 694 155 148 76
Morocco 648 633 611 376 327 278
Saudi Arabia 2 854 1 873 1 423 1 387 931 722
Sudan 3 029 1 765 1 267 1 387 1 129 732
Syria 1 064 803 647 367 206 149
Tunisia 373 349 206 201 176 180
Turkey 536 343 260 115 65 49
Yemen AR 2 980 1 334 675 628
Yemen PDR 272 171 142 150 124 152

Source: Calculated from FAO Production Yearbooks

- 57 -


Ratio of Arable Land per Harvester (ha/machine)

Country 1960 1965 1970 1975 1980 1985

Algeria 1 409 1 356 1 163 1 846 1 821 1 503
Iran 15 358 11 907 9 547 7 454 5 713 5 113
Iraq 3 757 1 908 1 287 1 088 1 434 1 981
Jordan 22 634 13 684 9 460 7 184 5 750 4 660
Morocco 1 710 2 213 3 001 2 843 2 501 2 636
Saudi Arabia 6 714 6 082 5 487 4 007 2 792 2 181
Sudan 51 500 27 173 18 800 13 511 10 797 10 639
Syria 4 744 4 403 4 190 3 011 2 410 2 014
Tunisia 1 615 1 518 1 451 2 074 1 869 1 838
Turkey 4 133 3 578 3 244 2 405 2 083 2 030
Yemen PDR 52 500 30 750 24 833 11 833 11 142 11 133

Data on Libya and Yemen Arab Republic are unavailable.

Source: Calculated from

FAO Production Yearbooks


Growth Rates of Tractor Park
and Combine Harvesters in Selected Countries,



Yemen AR
Yemen PDR





* 1970-85.

Source: Calculated from FAO Production Yearbooks.

- - - - --- --- ----------- -- -- --" -

- 59 -


Anbadan-Unat, Nermin et. al. Migration and Development: A Study of the
(1976) Effects of International Migration on Bogazliyan
District. Ankara: Ajans-Turk Press.

Adler, Stephen "Swallow's Children: Emigration and Development in
(1980) Algeria, 1980". ILO, WEP-2-26/WP 46.

Amin, Galal and Elizabeth Awny International Migration of Egyptian
(1985) Labour: A Review of the State of the Art. Ottawa,
Canada: IDRC, May.

Aricanli, Tosun "Agricultural Labour Responses to Population Growth
(1985) and Agrotechnical Intervention: The Gezira Scheme", in
Frederick Shor ter and Huyda Zurayk, eds., Population
Factors in Development Planning in the Middle East. New
York: The Population Council.

Aydin, Zulkuf "Agricultural Labour and Technological Change in
(1988) Jordan". In Tully, ed.

Bardhan, Pranab "Employment and Wages in a Poor Agrarian Economy".
(1979) American Economic Review 69, 1 (March).

Beaumont, Peter and Keith McLachlan, eds. Agricultural Development in
(1985) the Middle East. New York: John Wiley and Sons.

Bennoune, Mahfoud The Making of Contemporary Algeria. Cambridge:
(1988) Cambridge University Press.

Berar-Awad, Azita Employment in the Sudan: An Overview of Selected
(1984) Issues. Geneva: ILO (September).

Binswanger, Hans Agricultural Mechanization: Issues and Policies.
(1987) Washington, D.C.: World Bank.

Birks, J. Stace and Alan Richards "Migration and Agriculture in the
(1984) Middle East: Two Case Studies: Oman and Egypt". In
Philip L. Martin, ed. Migrant Labour in Agriculture: An
International Comparison. Berkeley, California: The
Giannini Foundation of the University of California.

Birks, J. Stace and Clive A. Sinclair Arab Manpower. London: Croom
(1980a) Helm.

Birks, J. Stace and Clive A. Sinclair International Migration and
(1980b) Development in the Arab Region. Geneva: ILO.

Birks, J. Stace and Clive A. Sinclair "The Socio-economic Determinants
(1982) of Intra-Regional Migration". In UNECWA, pp. 733-752.

Boxley, Robert "'Aging' Farmers: Not a Crisis". Choices, First
(1987) Quarter, pp. 24-25.

- 60 -

Carapico, Sheila "Yemeni Agriculture in Transition", in Beaumont and
(1985) McLachlan, eds. pp. 241-254.

Choucri, Nazli "The Hidden Economy: A New View of Remittances in the
(1986) Arab World". World Development, 14, 6 (June),
pp. 697-712.

Commander, Simon The State and Agricultural Development in Egypt Since
(1987) 1973. London: Ithaca Press.

Danielson, Michael N. and Rusen Keles "Urbanization and Income
(1980) Distribution in Turkey", in Ergun Ozbudun and Aydin
Ulusan, eds, The Political Economy of Income Distribu-
tion in Turkey. New York: Holmes and Meier, pp. 269-310.

El-Dib, M.A.M., S.M. Ismail, and Osman Gad "Economic Motivations and
(1984) Impacts of External Migration of Agricultural Workers in
an Egyptian Village". Population Studies, 11 (68), pp.
27-46. (Arabic).

El-Ghonemy, M.R. Rural Poverty in the Near East Region. Rome: FAO.

Enger, W. "Yemen Arab Republic Agricultural Assessment, Update,
(1986) 1985". San'a: USAID.

FAO The State of Food and Agriculture. Rome: FAO.

FAO "The Role of Women in Food and Agricultural Production
(1986) in the Near East Region". Paper prepared for Regional
Conference on the Near East.

FAO "Agricultural Price Policies in the Near East: Lessons
(1988) of Experience" Rome: FAO (October).

Fergany, Nader "Some Aspects of Return Migration in Egypt". Mimeo.

Fitch, James B. and Ibrahim A. Soliman "Livestock and Small Farmer
(1983) Labour Supply". In Richards and Martin, eds., pp. 45-78.

Hansen, Bent "The Egyptian Labour Market: An Overview". Washington,
(1985) D.C.: World Bank Discussion Paper, No. DRD 160.

Hopkins, Nicholas Agrarian Transformation in Egypt. Boulder, Colorado
(1987) and London: Westview Press.

Ibnouf, Mohammad Ahmed Osman An Economic Analysis of Mechanized Food
(1985) Production in the Central Plains of the Sudan. Ph.D.
Dissertation, Michigan State University. Ann Arbor,
Michigan: University Microfilms.

- 61 -

Ibrahim, Fouad Ecological Imbalance in the Republic of Sudan With
(1984) Reference to Desertification in Darfur. Bayreuth:
Druckhaus Bayreuth Verlagsgesellschaft mbH.

IFAD Report of IFAD's Special Programming Mission to Yemen
(1988) Arab Republic. Rome: IFAD, Peport No. 0106-YA (July).

ILO Labour Markets in the Sudan. Geneva: ILO.

Johnson, Paul "The Impact of Islamic Revolution on the State/Private
(1988) Sector Symbiosis". New York: Social Science Research
Council, Mimeo.

Kazemi, Farhat Poverty and Revolution in Iran: The Migrant Poor, Urban
(1980) Marginality and Politics. New York: New York University

Kazi, Shahaz "Domestic Impact of Remittances and Overseas Migration:
(1988) Pakistan". ARTEP (ILO/UNDP), Working Paper No. 7,
RAS/85/009 (April).

Kelley, Allen C. and Jeffrey G. Williamson What Drives Third World
(1984) City Growth? Princeton, N.J.: Princeton University

Khafagi, Fatma "Socio-Economic Impact of Emigration from a Giza
(1983) Village". In Richards and Martin, eds. pp. 135-157.

Kirwan, F.X. "The Impact of labour Migration on the Jordanian
(1981) Economy". International Migration Review, 15, 4,
pp. 671-695.

Kopits, George Structural Reform, Stabilization and Growth in Turkey.
(1987) Washington, D.C.: IMF.

National Research Council Population Growth and Economic Development:
(1986) Policy Questions. Washington, D.C.: National Academy

Nelson, Harold D. ed. Algeria: A Country Study. Washington, D.C.:
(1985) American University Press.

Nelson, Harold D. ed. Tunisia: A Country Study. Washington, D.C.:
(1985) American University Press.

Parker, John B. and James R. Coyle Urbanization and Agricultural
(1981) Policy in Egypt. USDA, ERS, Foreign Agricultural
Economic Report No. 169 (September).

Radwan, Samir and Eddy Lee Agrarian Change and Rural Poverty: A Case
(1985) Study of Egypt. Geneva: ILO.

- 62 -

Richards, Alan And Philip L. Martin, eds. Migration, Mechanization
(1983) and Agricultural Labour Markets in Egypt. Boulder,
Colorado and Cairo: Westview Press and the American
University in Cairo Press.

Richards, Alan And Philip L. Martin and Rifaat Nagaar "Labour
(1983) Shortages in Egyptian Agriculture". In Richards and
Martin, eds., pp. 21-44.

Saket, Bassam K. "Promoting the Productive Use of Remittances", in
(1982) ECWA, pp. 1093-1127.

Scobie, Grant M. Food Subsidies and the Government Budget in Egypt.
(1985) Washington D.C. International Food Policy Research

el Sherbiny, A.A. Food Security in the Arab Near East. London:
(1980) Pergamon Press.

Sherbiny, Naiem "Expatriate labour in Arab Oil-Producing Countries".
(1984) Finance and Development. December, pp. 34-37.

Sluglett, Marion Farouk and Peter Sluglett Iraq Since 1958: From
(1987) Revolution to Dictatorship. London and New York: KPI.

Stark, Oded "A Note on labour Migration Functions". Journal of
(1984) Development Studies.

Swanson, Jon C. Emigration and Economic Development: The Case of the
(1979) Yemen Arab Republic. Boulder, Colorado and London:
Westview Press.

Taylor-Awny, Elizabeth "From Peasant to Proletarian? A Study of
(1984) Egyptian Peasant Household Production", Cairo, Mimeo.

Tully, Dennis, ed. Agricultural Labour and Technological Change in
(1989) Dry-Farming Areas of the Middle East and North Africa.
Boulder, Colorado and London: Westview Press for ICARDA.

Tutwiler, Richard "Agricultural Labour and Technological Change in
(1989) the Yemen Arab Republic", in Tully, ed.

UNECWA International Migration in the Arab World. Beirut.

USAID "Morocco: Agricultural Sector Review". Rabat.

USDA Syria: Agricultural Sector Assessment. Washington, D.C.
(1986) and Damascus: USDA in cooperation with USAID and the
State Planning Commission, Syrian Arab Republic.


World Bank

World Bank

World Bank

World Bank

63 -

Middle East and North Africa: Situation and Outlook
Report. Washington, D.C.: USDA, ERS, RS-86-1, (April).

Morocco: Economic and Social Development Report.
Washington, D.C.: The World Bank.

Tunisia: Agricultural Sector Survey. Washington, D.C.:
The World Bank.

Turkey: A Country Economic Study. Washington, D.C.:The
World Bank.

Yemen Arab Republic: Agricultural Sector Survey.
Washington, D.C.: The World Bank.




Diana Hunt
School of African and Asian Studies
University of Sussex

- 67 -

Table of Contents




A. Expansion of Mechanization 71
B. Dominant Influences and Motives in the
Mechanization Process 72


A. Examples of Regions of Significant Expansion
of Water Adduction 74
B. The Output and Employment Impact of Investment
in Mechanized Water Adduction 75
1. The output impact of mechanized
water adduction 75
2. The employment impact of mechanized
water adduction 78
C. Drainage Problems and Soil Salinization 79


A. Recent Expansion of Mechanized Crop Production
in the Near East 79
B. Output Impacts of the Use of Machine Power
in Crop Production 80
1. Expansion of cultivated area 81
2. Yield effects of mechanization on
previously cultivated land 81
3. Possible adverse yield effects of
production mechanization 83
C. Employment Impact of the Use of Machine
Power in Crop Production 84
1. Extension of aggregate cultivated area 84
2. The mechanization of crop production
on land already cultivated 85
D. Small-scale Tractors: Their Viability and
Employment Impact 87
E. The Impact of Mechanization of Crop Production
on the Structure of Agricultural Employment 88




- 68 -


A. Distribution of Land Rights and the Decision
to Invest in Motorized Equipment 91
B. The Distribution of Land Rights and the
Impact of Mechanization on Agricultural Labour 92




A. Credit and Distribution 96
B. Servicing 96
C. Research and Training 97
1. Research 97
2. Training 97
D. The Need for Farmer Government Dialogue on
Mechanization Issues 98







In recent decades Near East countries have experienced a notable
expansion in agricultural mechanization. This expansion has not been
smooth: individual countries have experienced spurts and lags in the
expansion of their agricultural machine park. Nor has the expansion
been uniform. Varying degrees of mechanization have been achieved both
between regions and between farms within regions. Yet the overriding
trend is clear: towards an increase in the use of motor-powered
equipment relative to both land and labour.

The expansion of mechanization has given rise to a number of
policy issues. These issues can be classified in various ways, but
four broad headings embrace most of them: these relate to technical,
economic and social aspects of mechanization and, fourthly, to the
role of foreign equipment suppliers and foreign aid in this process.
Among the more notable features of mechanization which are of
significance throughout the region are the changes which it is
generating with respect to the use of land, labour and foreign
exchange, the level and composition of farm output, the distribution
of land and income and the need for development of supporting

Farm mechanization is not a simple 'either/or' issue. Rather,
the central issue concerns the appropriate form and extent of
mechanization.(2) Various factors underly the need to adopt this more
complex perspective. The scope for mechanization is determined partly
by the relevant physical conditions. Secondly, farm mechanization is
also potentially highly divisible in that mechanization of one
operation does not necessarily entail that of others. Thirdly,
motorized power sources come in a variety of forms with respect to
such characteristics as function, size, range of equipment, and
capital and operating costs. Fourthly, farms, too, vary in size and
structure. Finally, the appropriate mechanization is also a function
both of the cost and efficiency of the available technical
alternatives, and of mechanization's output impact. Since these
factors vary both spatially and over time, so too the appropriate form
and extent of farm mechanization varies both spatially and

(1) Mechanization is here interpreted to mean motorization: i.e. the
application of all types of mechanical motors or engines,
regardless of energy source, to activities associated with
agriculture (Gifford, 1980:4). In this paper, however, the focus
is placed on motors powered by petrol, diesel or electricity.
(2) Cf. Bartsch, 1977, Gifford 1980, and, for an excellent
illustration, Agarwal, 1981; see also Hopkins, 1984:187.

- 70 -

While the rural sectors of Near East countries have certain
notable characteristics in common semi-arid climate, a significant
role for irrigation, low overall population density more detailed
examination reveals considerable diversity between their agricultural
sectors. These vary in terms of size (spatial, demographic, economic),
population density on cultivable land, agricultural labour force
trends, land tenure structures, proportion of irrigated farm land,
topography, soils, climate and further irrigation potential. The
national economies also vary in terms of pertinent factors such as
share of agriculture in GDP, GDP per capital, and foreign exchange
availabilities. It is thus hardly likely that the same pattern of farm
mechanization will be appropriate in all countries.(1)

Three factors complicate attempts to promote appropriate farm
mechanization strategies in Near East countries. Firstly, there is a
chronic lack of relevant data. Secondly, what is appropriate to the
private farmer may not appear equally appropriate when viewed from the
perspective of society as a whole. Thirdly, what is appropriate from
the latter perspective is open to more than one interpretation.

In principle, these ambiguities can be resolved. Meanwhile,
however, their existence, and the reasons for them, serve to pinpoint
the breadth of policy issues that are relevant to the mechanization
debate. These concern all those aspects of government economic
intervention that affect the farmer's decision environment, including
agricultural pricing, land tenure, the focus of agro-engineering
research, training and credit provision. Policy in some of these
fields is not always set with a view to the determination of
production technique, although in other cases it is. Where it is.not,
it may influence farmers' technical decisions in unintended ways.

However, the existence also of deliberate policy interventions
regarding farm mechanization brings us to the second element of
ambiguity noted above. Farm mechanization may entail trade-offs
between social objectives. It does not necessarily do so. However, by
way of illustration, it may expand marketed farm output but generate
increased structural unemployment. Again, it may expand output and/or
reduce drudgery, but absorb increasing quantities of possibly scarce
foreign exchange. Likewise, mechanization may be highly effective in
meeting a target of national food self-sufficiency, but only at unit
production costs far above world market prices. Where such trade-offs
exist there cannot be consensus on the appropriate degree of
mechanization without prior consensus on the relative importance of
the variables involved in the trade-off.

This paper does not aim to resolve such issues. Rather, having
highlighted the problem, it seeks, through a review of the available
evidence, to explore more fully some of the matters raised in this
introduction. While it would clearly have been desirable to generate

(1) The possibilities range from prolonged adherence to zero
mechanization in certain conditions (see Varisco, 1982, for an
example) to a more or less rapid transition to highly mechanized
farming in others.

- 71 -

precise conclusions on the extent of the incidence of trade-offs such
as those just noted, any reliable estimation of their scale is
precluded by data limitations. Moreover, for reasons of space, the
first part of the paper focuses on the output and employment impacts
of farm mechanization, using primarily micro-economic evidence, where
available, to illustrate these. Other possible impacts are touched
upon, but receive less emphasis. The second part of the paper
(Sections VII-X) reviews a number of the policy issues that are
relevant to the promotion of farm mechanization.


A. Expansion of Mechanization

Agricultural mechanization in the Near East dates back to the
late nineteenth century, to the introduction of mechanical pumps for
lifting irrigation water. In the early decades of the twentieth
century the first investments in tractors occurred. By the late 1920s
there were 2,000 tractors in Turkey (Kozacioglu, 1982:7). From the
mid-1920s tractors began to be imported into Iran, and into Jordan and
Pakistan in the 1930s. In North Africa too, in the 1920s and 1930s
mechanization developed, chiefly on colonial farms (Maklouf, 1979:5).
However, outside the colonial farming areas of the Mahgreb, these
early initiatives did not spread rapidly. Rather, various factors,
including transportation constraints, the lack of service and repair
facilities, spare parts and fuel and, in some cases, the distribution
of land rights, restrained the spread of mechanization (Kozacioglu,
op. cit.; Okazaki, 1985).

World War II provided a new market stimulus that led to the
start of mechanized farming in the Sudan and Syria(l), and to the
establishment of mechanized state farms in Turkey(2). However, by 1948
there were still only 1,800 tractors in Turkey, and in 1947 about 500
in Pakistan (Ansari and Raza, 1984:20(3)), while in each of the other
countries of the region there were probably substantially less. It is
essentially since World War II that the mechanization of Near East
agriculture has got underway.

Table IA summarizes the size of the tractor park that had been
attained in each of the Near East countries by 1961, and its
subsequent expansion to 1986, while Table IB provides the same
information for the park of combine-harvesters. Both the pace and
timing of expansion have varied between countries (Table IC).
Meanwhile, within individual countries one finds farms that are close
to fully mechanized, others where selected operations only are
mechanized, and others where farmers still rely entirely on
traditional power sources. Even in Turkey, a country with one of the

(1) Affan, 1984: Manners and Segafi-Nejad, 1985:264.
(2) Mostly turned over to the private sector after 1948 (apart from
seed farms).
(3) See also Hussain, 1985.

- 72 -

highest tractor:arable land ratios in the region, this diversity still
exists, albeit in this case with the added complication that some
industrial crop growing regions appear to be facing overmechanization
judged in terms of the ratio of machine capacity to cropped area
(Balkir, 1984:59).(1)

Overall, this number of tractors in the region rose from 81,000
in 1961 to 1.2 million in 1986.

B. Dominant Influences and Motives in the Mechanization Process

Various factors have contributed to the mechanization process,
including the policies of national governments, the promotional roles
of foreign suppliers and of foreign aid agencies and, in some
countries, a decline in the agricultural labour force. At the farm
level, these have been reinforced by a desire to expand income, reduce
risk and reduce drudgery.

The main policy instruments that have influenced mechanization
were mentioned above (p. 70). While sometimes the influence of these
has been incidental rather than deliberate(2), often it has been the
latter, reflecting a positive desire to promote mechanization with one
or more of five main underlying motives:

a) to raise output by expanding total arable area and/or
irrigated area;

b) to facilitate an increase in cropping intensity;

c) to raise farm yields through improved seedbed preparation and
more precise use of recurrent inputs (especially seed and

d) to sustain farm output in the face of a declining farm labour

e) to fully modernize production methods to compete with the
agricultural systems of the industrially advanced countries.

The extent to which the first three technical objectives have
been met is discussed in Sections III.B. and IV.B. below. Here,
however, it is necessary to comment on the other two, for both reflect
more questionable motives for mechanization, either in some countries
(d), or in all countries (e).

Regarding the need for agricultural mechanization to replace a
declining agricultural labour force, Table II shows that agricultural
labour force trends in the Near East region are far from uniform. In
some countries, for example Libya, the recorded agricultural
population has been declining for several decades, in others it has

(1) This last point applies even more strongly in Cyprus (see Table
(2) As when high prices for dairy products and meat encourage a
reduction in the use of animal draught power.

- 73 -

more recently peaked, but in the majority, including Pakistan, Egypt
and Sudan, it is still rising. Moreover, in countries recording a
downward trend, this has not always been indicative of labour
scarcity, even at peak seasons. For example, Tunisia in the early
1960s had considerable agricultural unemployment (some 30 percent of
the labour force) as a result of previous land appropriation and
agricultural mechanization by colonial and other farmers (Maklouf,
1979). Even at peak periods there appears to have been some slack.
More recently, there has been circumstantial evidence that in the two
Yemens there may also have been some slack, in the form of under-
employed labour, given the extent of the ability of both countries to
maintain agricultural output during the periods of high

Currently available official estimates of the 1980 agricultural
labour force in some Near East countries may not yet fully reflect the
spurt in out-migration from these to the main oil exporting countries
of the region that occurred in the mid-1970s (due to delays in
census-processing). Yet by now, the late 1980s, this flow is reported
not simply to have slowed down, but to be moving into reverse gear,
with an influx of returning migrants in some countries. While some
rural migrants obtained urban sector jobs in their countries of
destination (for example in construction) they will not necessarily
find full-time non-farm occupations on their return; some may look for
agricultural employment. A combination of returning migrants plus a
continuing upward long-run trend in the rural population may in coming
years generate problems of excess labour supply in agriculture in some
countries of the region, especially those which already have a high
labour:cultivable land ratio.

Motive (e) for promoting mechanization provides a less strong
justification than any of the others, due to the incompleteness of the
underlying economic argument. The international price competitiveness
of products certainly depends partly on their production costs, but
neither the relative unit costs of production factors nor their
productivity are uniform across countries. In developing countries,
imported machinery tends to be more expensive relative to labour than
in the industrially advanced countries, for a variety of reasons:
imported farm machinery is often bulky and has quite high per unit
transport costs; the local distribution mark-up is high; labour is
often (not always) relatively abundant, and the prevailing real wage
lower, in developing countries than in those more industrially
advanced; and foreign exchange, particularly in the non-oil producing
countries, tends to be scarce, which further raises the cost of
imported production factors relative to locally available substitutes.
Consequently, the most cost competitive production method is not
invariably that which is most mechanized.

(1) See FAO, 1987:84. Only in Yemen Democratic Republic in 1978 and in
1982 was there a notable fall in output. On the existence of
labour slack, see also Commander and Burgess, forthcoming, Azzam,
forthcoming and Bourinet, 1975.

- 74 -

The questions attached to these last two motives for mechaniza-
tion serve to underscore the more general point that the appropriate
pace of introduction of mechanical technology cannot be generalized.

Meanwhile, however, mechanization has been actively promoted by
foreign equipment suppliers anxious to expand sales, and has also
periodically received notable support from foreign aid agencies, such
aid being based on a more or less subjective assessment of local
conditions and needs. In the case of bilateral aid a definite element
of national self-interest has been involved: in particular a desire to
expand foreign machinery exports.

While governments have promoted mechanization for a variety of
reasons, so, too, farmers have invested in the use of motorized power
for a variety of purposes and with various objectives, as Sections
III-V illustrate. In these sections three types of farm mechanization
- in water adduction, crop production and crop processing are
analysed separately in order to clarify their output and employment
impacts. Of these three broad categories of mechanization none
automatically entails the other. Indeed, within crop production alone
there are a range of individual processes, each of which may or may
not be mechanized.


A. Examples of Regions of Significant Expansion of Water Adduction

Traditionally, irrigation in the Near East has been based upon
seasonal flooding, gravity distribution and water adduction by human
and animal power. However, in recent decades the mechanization of
water adduction has expanded substantially. Provision of irrigation
water has also been expanded through the construction of major dams
and associated gravity distribution to large command areas (including
in Algeria, Morocco, Tunisia, Egypt, Syria, Iraq and Pakistan).
Mechanized water adduction from rivers and wells is being used partly
as a supplement to other water distribution systems and partly as the
sole basis for irrigation.

The countries in the region that have experienced the largest
expansion of motorized pumping include Pakistan, where, according to
one relatively cautious estimate, the number of private tubewells
increased from 32,500 in 1965 to 179,000 in 1980 (Bergmann,
1984:56)(1); Turkey, where the number of farm pumps (centrifugal and
motor) rose from 52,000 in 1963 to 416,000 in 1982 (Balkir, 1984:291);
Syria, where substantial investment in river pump irrigation created
some 250,000 pump irrigated hectares by the early 1950s, and where
privately owned pumps rose from 21,000 in 1960 to 47,000 in 1980
(Manners and Segafi-Nejad, 1985:263), the Sudan, where 0.5 million ha
were already punp-irrigated by 1963 with further expansion since (El
Mangouri, 1983:74), and Egypt. There has also been a notable expansion
of mechanized water adduction in recent years in Saudi Arabia and some

(1) According to O'Mara in Pakistan 'the number of private tubewells
increased from less than 5,000 in 1960 to about 200,000 in 1980'
(O'Mara, 1984:39).

- 75 -

of the other smaller Gulf states. However, expansion has not been
confined to these countries, and has occurred throughout the region.

B. The Output and Employment Impact of Investment in
Mechanized Water Adduction

Investment in mechanized water adduction may either be under-
taken to replace animal draught power (as in Egypt) or to expand total
farm water availability (as in the rest of the region). Such
investments have been geared to both raising and stabilizing farm
output and incomes. These goals have been pursued through using
irrigation water to:

expand the cultivable area;

increase the scope for producing more moisture-sensitive and
moisture-demanding crops and crop varieties;

increase cropping intensity;

reduce swampiness and soil salinity by lowering the water-

reduce variations in plant moisture availability due to
uncertain rainfall;

raise livestock yields (of dairy produce and meat) by
releasing animals from water lifting.

For example, early investment in pump irrigation in Iraq was
undertaken to increase the scope for summer cropping(l). The early
1960s public sector tubewell investment in Pakistan was undertaken to
reduce swampiness and soil salinity, but subsequent private sector
investment was in order to reduce risk, facilitate adoption of
high-yielding varieties and raise cropping-intensity. In Saudi Arabia
and some of the smaller Gulf states investment has been primarily to
expand cultivable area. In Egypt, on the other hand, the mechanization
of water adduction appears to have been undertaken partly in order to
raise overall farm profitability by releasing animals from draught
work, given both the potential returns to the production of animal
products and the limited availability of land for fodder production.

Quantitative data on the output and employment impacts of
mechanized water adduction in Near East countries are limited.
However, the following are among the more noteworthy findings to have
emerged from studies undertaken so far.

1. The output impact of mechanized water adduction

Factors affecting the output impact of mechanized water pumping

what previous method of water adduction,if any,was being used;

the extent of new land brought under cultivation;

(1) Theobald and Jarad, 1982.

- 76 -

the cropping intensity on newly cultivated land;

the actual production methods (biochemical and technical) that
are employed on newly irrigated land;

the extent of increased cropping intensity on already
cultivated land, and any change in cropping pattern and/or
production methods on such land;

any changes, positive or negative, in soil salinity and
swampiness and, hence, in crop yields and crop choice.

In Egypt where mechanization of water lifting has been chiefly a
straight substitute for animal draught power, the majority view is
that the output impact has been an increase in livestock yields as
energy released from draught work is converted into milk and higher
live calf production (see e.g. Richards and Martin, 1983). However,
the adverse yield effects of using milK cows and buffaloes for draught
purposes depend on the amount and manner of their use and on whether
the animals are fed properly. An alternative view is that in
water-lifting adverse yield effects are insignificant. In the
continuing absence of conclusive data the issue remains controversial
(Director of Institute of Animal Production, Giza, personal
communication, Februrary 1989).

In other countries, however, mechanization has facilitated an
expansion of irrigated area and, hence, crop production. Among a
sample of farmers in the irrigation command area of the Pakistan
Punjab, for instance, Bergmann found that after investing in their own
tubewells farmers reduced fallow land by one-third, and raised their
irrigated acreage from 47 percent to 68 percent of the total farm
area. Most farmers (84 percent) also increased their cropping
intensity, while the share of high-yielding crops, especially wheat
and IRRI rice also expanded. In the Punjab, where the water table is
high, tubewell investment also permitted some farmers to take steps
against swampiness.

The rapid expansion of farm output in Saudi Arabia since the
mid-1970s (food output trebled between 1976 and 1986 (1)) bears
further witness to the potential output impact of mechanized water
adduction. There irrigation investment has permitted a major expansion
of cultivated area (Joffe, 1985).

The substantial expansion of privately owned tubewells within
the Near East region is indicative that farmers regard them as
privately profitable. Since in most cases output and labour use are
both being increased as a result of these investments (only in Egypt
are there net labour savings), the profit must come primarily from the
increased output. Indeed, in a review of a range of possible
investments in Moroccan agriculture on both large and small farms,
Cleaver (1980) found that the highest financial, economic and social
rates of return accrued to small farm investment in a well and a pump:
37 percent, 60 percent and 50 percent respectively (see Table III;

(1) FAO, 1987:84.

- 77

interestingly the projected economic rate of return is equalled by
only one other investment: large farmer tractor purchase)(l).

In Morocco, at least, mechanized water adduction holds its own
when appraised using appropriate shadow prices for relevant inputs and
outputs. In contrast, Egypt's experience in mechanized water
adduction, which may be exceptional given the limited output impact,
does not show strong economic returns. As Table IV shows, in the early
1980s in Sharqiyya Governorate, potential private returns to tubewell
investment were high but economic returns were in almost all cases
negative. (The latter were calculated first using prevailing prices
but taking into account external costs and returns, and then,
secondly, using border prices.) Irrigation pumps entail significant
costs to the Egyptian exchequer because they consume heavily sub-
sidized diesel fuel. The value of the extra livestock output plus
labour savings that the pumps generate does not fully offset this

There are also countries where the farm output impact of motor-
powered pumps is larger than in Egypt, but where the economic returns
to mechanized water adduction would be shown to be negative if
calculated using the same principles. One is Saudi Arabia, where in
the early 1980s the price offered to wheat farmers was five times that
prevailing on the world market (Joffe, 1985). Major policy issues, and
associated value judgements, pertain to the establishment of such
price differentials: in particular, an evaluation of the social value
of national food self-sufficiency, on the one hand(3), and of the
scarcity value of foreign exchange on the other.

The net returns to irrigation are also a function of the crops
grown. A number of Near East countries are now developing greenhouse
production of horticultural crops. Controlled water distribution in
greenhouse conditions is more economical of water than orthodox field-
irrigation, and this offers a promising supplementary line of
irrigation expansion in the future. Other promising lines of
development, particularly for countries facing severe water
constraints and/or salinization problems, are improved methods of
field distribution: drip and sprinkle irrigation. Cyprus and Jordan
have both made advances with these methods on relatively small farms,
while Saudi Arabia and Libya have both introduced large-scale centre
pivot systems.(4)

As compared with the mechanization of crop production and of
post-harvest operations, investment in mechanized water adduction is

(1) The economic return was calculated using shadow prices for inputs
and outputs. The social return was estimated using the same shadow
prices and, in addition, an income weighting to reflect
distributional objectives.
(2) Dyer and Gotsch, 1983.
(3) On the importance of this objective for Arab countries see
Abusmeileh, 1987.
(4) On the latter, see Keller and Plocher, 1984.

- 78 -

characterized by a greater probability of a positive output impact.
However, the impact can be negative even in the short run if increased
irrigation is not linked to adequate investment in soil drainage and
is not properly managed; if these conditions are not met, the results
may be waterlogging and increasing soil salinization. Such problems
have been experienced in, for example, both southern Iraq and the
Pakistan Punjab. In the medium-term too, negative yield effects may be
generated if excessive use of sub-surface water lowers the water table
and dries wells. There is a risk that this may become a serious
problem in Saudi Arabia, the two Yemens and some of the smaller Gulf

In coastal regions excessive pumping can also cause saline water
intrusion into the aquifier (as has occurred, for example, in

2. The employment impact of mechanized water adduction

In water lifting the employment impact of mechanizing water
adduction may be positive, neutral or negative, depending on whether
and by how much the investment extends irrigated area. In general,
though, this direct effect, whether positive or negative, is likely to
be relatively slight, as illustrated in Table V, which gives the
labour requirement per feddan for maize irrigation in Egypt. On the
other hand, the indirect employment effect of mechanized water
adduction that arises when there is an extension of irrigated area is
positive and can be substantial, as illustrated in Table VI. Table VI
is based on an amalgam of data sources (chiefly from research stations
and farm studies) for the Pakistan Punjab. It shows that, assuming no
mechanization of production processes, on land already cultivated
tubewell irrigation can generate an increase in labour use of up to
37.5 days per hectare per year, if followed by adoption of the 'green
revolution' bio-chemical technology and a doubling of cropping
intensity. With the mechanization of threshing the labour requirement
falls by 3.25 days. However, the use of a seeder plus improved animal
draught implements for land preparation raises the requirement again
slightly: by 1.25 days. On previously uncultivated land the positive
employment effect would be correspondingly greater.

Cleaver's estimate of the employment impact of small(2) farm
well and pump investment in Morocco is still higher per hectare
irrigated. On 0.75 ha he predicts an annual increase of 24 days of
family labour plus 50 days of seasonal labour, assuming that the
irrigated land is used for potato production (Cleaver, 1980:4). The
aggregate employment impact would depend on the scale of expansion of
small-scale irrigation. Cleaver concluded that in Morocco there was
scope for creating some 17,000 man-years of employment by this means.

A variety of factors including climate, cropping intensity,
cropping pattern, bio-chemical technology and crop production
techniques affect the extent of the positive employment impact of

(1) Carruthers and Stoner, 1981:68.
(2) 6 hectare.

- 79 -

investments in motorized water adduction. As between the Pakistan and
Moroccan examples the different impacts can be explained largely by
the second of these.

Bergmann's survey of farms in the Pakistan Punjab confirms that
labour use rises following investment in tubewells. He also reports a
consequent change in the composition of the farm labour force. With
the more even spread of labour requirements over the year that follows
from increased cropping intensity, and with increased returns to
labour arising from increased crop yields, there is a greater full-
time involvement of family members in farming (an increase of
15 percent was observed on a sample of 186 farms), together with some
increase in permanently hired workers (+9 per cent)(l). There was also
an increase in the part-time involvement of other family members
(usually the second or third sons of the owner). The trend with
respect to casual hired workers, however, was less clear-cut, largely
due to the in practice widespread mechanization of crop production
operations, especially ploughing, and also harvesting and threshing,
following investment in tubewells.(2)

C. Drainage Problems and Soil Salinization

Despite the foregoing, it should be remembered that, unless
carefully controlled, the expanded use of irrigation water may also
affect yields and, hence, employment negatively. To ensure that the
potential gains from mechanized water adduction are realized, and are
sustained, effective policy measures are required. In Near East coun-
tries there is a need for regular monitoring of the sub-surface water
table, for control, where necessary, of the rate of water extraction
and for provision to ensure good management of irrigation water.

A more unusual, but important, example of the adverse impact of
mechanized water adduction comes from the Yemen Arab Republic.
Expanded well and pump irrigation in the coastal plain is reported to
have created labour shortages in the highlands. The more remote
terraced farms have been abandoned, with serious implications for
drainage patterns, and for flooding and soil erosion in the middle
stretches of the valleys. Groundwater regeneration may also be
affected, as more water flows wastefully to the sea (Kopp,1985:47).(3)


A. Recent Expansion of Mechanized Crop Production
in the Near East

The most notable feature of mechanized crop production in the
Near East is unquestionably the tractorization of land preparation.

(1) Bergmann, 1984:147.
(2) Bergmann, op.cit.: 147, 8.
(3) Tutwiler, forthcoming, notes that planting qat on abandoned
terraces can help to reduce these problems as the tree's root
system holds tenuous highland soil and encourages moisture

- 80 -

Between 1961 and 1986 the estimated number of tractors in use in the
region rose from 81,000 to 1.2 million, and the estimated ratio of
tractors to arable land rose from 1:1,000 to 1:80 a twelve-fold
increase (for individual countries see Tables IA and ID). The rate of
growth of the tractor park in individual countries by five-year
periods is summarized in Table IC.

This substantially expanded stock of machinery has been used
primarily for land preparation. However, to varying degrees the same
tractors are also used for other purposes, notably for transportation,
to power stationary threshing machines, and for seed application. Some
weeding is also done by tractor. In addition, there has also been an
expansion of mechanized harvesting in the region (Table IB).

Reasons for the expansion of mechanized crop production include:

i) to expand cultivated area;

ii) to improve seed-bed preparation;

iii) to increase timeliness of operations;

iv) rising labour-costs relative to machine power;

v) to ease labour management on larger farms;

vi) to release a)-fodder land for other crops or b) livestock for
other productive uses;

vii) to reduce drudgery.

However, the extent of production mechanization varies con-
siderably. There are many farms that are only partially mechanized (to
varying degrees), and many too which, for diverse reasons, continue to
rely on human and animal draught power alone. Even in Turkey, in 1980
(when the tractor park was 435,000) some 45 percent of farm holdings
over 0.5 ha did not report the use of tractors.(1)

B. Output Impacts of the Use of Machine Power in Crop Production

Broadly speaking it is possible to identify four different
contexts in which selected crop production processes have been

to farm previously uncultivated land, either rainfed (e.g.
central Sudan, and parts of Turkey, Syria and Iran) or
irrigated (e.g. Saudi Arabia);

to replace human energy and animal draught power on previously
cultivated land, either rainfed (e.g. parts of the Maghreb,
Turkey, Cyprus) or irrigated (e.g. Egypt, parts of Pakistan

In the first two cases, the predominant output impact has been
positive, although there are also instances in which use of

(1) FAO Report on the 1980 World Census of Agriculture, Census
Bulletin No. 13.

- 81 -

mechanization to extend cultivated area can have a negative output
impact in the medium-term (see below). In the second two cases,
however, although the agricultural engineering evidence for potential
yield gains is strong, these are often not achieved in practice.

1. Expansion of cultivated area

Notable examples of the use of mechanized land preparation to
extend cultivated area include:

i) expansion of mechanized production in the Gorgan region of
Iran in the 1950s, and much of the expansion of cultivated
area in north-east Syria and part of that in Turkey in the
same period;

ii) the more recent expansion of mechanized cultivation in central
Sudan (where the clay soils are also very difficult to work
using animal draught power or hand tools); and

iii) the recent dramatic expansion of wheat production in Saudi

Mechanization has not, however, invariably been a precondition
for large-scale extension of cultivated area. Thus, according to one
report, in Turkey between 1948 and 1956 the area cultivated by draught
animals rose by 40 percent, accounting for 60 percent of the expansion
of cultivated area in- this major period of area expansion in Turkish
agriculture (Kozacioglu, 1982:25, 6). Nonetheless, in sparsely
populated regions such as those cited, mechanization of land
preparation in particular, and in some cases, harvesting also, has
facilitated important extensions of cultivated area.

The output impact of such extensions depends on the scale of
area extension, the climate, and the farming practices used. Dura
output in the Sudan more than doubled between 1969/70 and 1975/76 as a
result of expanded land use arising from mechanization.(1) However, on
mechanized rainfed land there are also considerable seasonal varia-
tions in output caused by the climate (e.g., Jordan, Sudan, Syria).

2. Yield effects of mechanization on previously cultivated land

Especially on heavy soils, as in much of the grain producing
areas of North Africa, farmers have been encouraged to mechanize in
order to raise yields through improved seed-bed preparation. On these
soils deeper ploughing than can be achieved by animal draught power,
and a higher number of passages over the soil, can both improve the
seed-bed and increase moisture penetration. However, in much of the
eastern Mediterranean, where the soils are either sandy or rich loams,
the interaction of soils, climate and the timing of agricultural
operations means that different ploughing techniques are needed: disc-
or mould-board ploughing pulverizes the soils and increases the risk
of soil erosion during the rains that follow autumn-ploughing. In
these conditions chisel-ploughs are preferable and the traditional
animal-drawn nail-plough did a good job(2).

(1) From 424,000 kgs to 1,070,000 kgs (Affan, 1984:45).
(2) Mr. D.W. Sanders, Senior Officer, Soil Resources, Management and
Conservation Service, FAO, personal communication, February 1989.

- 82 -

Here the potential yield effects from mechanized land
preparation may be lower. Still, it can be argued that throughout the
region mechanized land preparation using the appropriate implements
can achieve a better, more even quality, seed-bed, thereby
facilitating more precise seeding and fertilizer application(l).

Meanwhile, mechanization, especially of larger farms, can also
improve the timeliness of land preparation and harvesting, both of
which are crucial for maximizing yields in semi-arid areas, and may
also assist in raising cropping intensity on irrigated land. Research
in Morocco indicates an average 0.3 ton/ha increase in cereal yields
as a result of early sowing permitted by tractors (Cleaver, 1980:6).

However, farmers do not necessarily achieve these potential
yield increases partly due to lack of knowledge and skill and partly
through deliberate choice given their farming system and overall
resource constraints. Thus, for example, on most Egyptian farms
positive yield effects from mechanized land preparation are reported
to be manifest for only two of Egypt's main crops maize and cotton,
but not for wheat, sorghum, rice or birsim (clover). Following
interviews with farmers this has been explained by the additional
number of ploughings given to maize and cotton land (Dyer and Imam,
1985:164). A fuller analysis of the prevalent farming systems would be
necessary to explain this choice. Meanwhile, many farmers in most
parts of the region (Cyprus is one notable exception) have not yet
invested in the equipment needed to raise yields through more precise
seed and fertilizer application. In the Pakistan Punjab, for example,
Bergmann found that only 2 percent of tractor owners owned seed drills
(see Table VII).

The literature on crop production mechanization is replete with
references to the need for farmers to use their machines to achieve
improved seed-bed preparation and more precise input applications, and
also with the observation that most persistently fail to do this.
Since there may be a variety of reasons why this is so lack of
skills and knowledge, resource constraints, failure of additional
returns to justify additional costs it is pointless for extension
workers to push such proposals uniformly until better understanding
has been achieved of why, in different farming contexts, farmers do
not adopt these practices. Only where there is clear evidence of
income foregone to the farm's overall operations (through, for
example, use of the wrong land preparation implements), should
attempts to change existing practices receive high priority.

Meanwhile, whether mechanized land preparation increases yields
through improved timeliness also depends heavily on machine
availability. High rates of breakdown, and limited availability of
spare parts and of good mechanics, can lead to yield loss when farmers
delay land preparation in expectation of the arrival of a
non-functioning tractor.

(1) R. Gifford, Chief, Agricultural Engineering Service, FAO, personal
communication, January 1989.

- 83 -

In principle mechanization of crop production may help to raise
cropping intensity and, hence, cultivated area and output, through
speeding up key operations. In practice, whether motorization of
selected production processes is essential for increased cropping
intensity depends on a variety of factors, including:

the length of the crop growth cycle;

the period of adequate moisture availability for the
establishment of each crop;

the adequacy of alternative power sources for land preparation
and harvesting.(1)

The second of these is determined jointly by climate and availability
of irrigation water.

Both Bergmann, and Dyer and Imam, report that they could
identify no significant impact of motorization on cropping intensity
in the Pakistan Punjab and Egypt respectively.(2) However, Bergmann
also observes that his findings may be distorted by the fact that the
tractor owners, whose 'post-purchase' behaviour he studied, may
already have been hiring tractors to help raise cropping intensity
prior to buying them.

The release of fodder land for higher value crop production
appears to have been a subsidiary motive for mechanization among some
farmers, for example in Pakistan. On the other hand, a desire to
increase output of livestock products (given their attractive price)
has apparently been more important in leading Egyptian farmers to
substitute motorized power for the use of multi-purpose (milk and
draught) animals in land preparation as well as water adduction.

3. Possible adverse yield effects of production mechanization

As with water adduction, so in the case of use of motorized
power for land preparation, inappropriate machine use and the wrong
choice of equipment can have negative yield effects. Important
examples in the region include the promotion of soil erosion by
ploughing up and down steep slopes (for example, in parts of Algeria,
Morocco and Tunisia) and by inappropriate choice of land preparation
equipment (as in central Sudan). Evidence from North Africa also
suggests that the use of motorized power sources to crop fallow land
may also carry a risk of reduced yields, especially in non-irrigated
areas where rates of fertilizer application may be constrained by low
rainfall. In the same region, good rangelands of Rhanterium suaveolem,
a valuable species in arid and semi-arid areas, has been degraded by
cultivation followed by invasion by the unpalatable Artemisia

(1) See Bartsch, 1977:43.
(2) Bergmann, op.cit.: 129; Dyer and Imam, 1985:165 and 174.
(3) Source: D. Crespo, Grassland Group, Crops and Grassland Service,
FAO, personal communication, February, 1989. See also Crespo,
1988:3, 4.

- 84 -

C. Employment Impact of the Use of Machine Power in
Crop Production

The employment impact of mechanized crop production has been
highly varied. In cases where mechanization achieved substantial
extensions of cultivated area labour demand has risen; in cases where
already cultivated land has been mechanized this has sometimes been
undertaken to offset the effects of rising male outmigration
(especially in the mid- to late-1970s and early 1980s), but in other
cases large farm mechanization has apparently sometimes displaced
agricultural labour for which adequate alternative employment was not

1. Extension of aggregate cultivated area

Where mechanization extends the cultivated area, the employment
impact depends on the scale of extension, the crops grown and the
number of operations mechanized. For example, by the early 1980s, the
partial mechanization of some 2.5 million hectares in central Sudan
was reported to have created seasonal (3 months per year) wage employ-
ment for about 0.5 million farm workers (Affan, 1984). In this case
mechanization had been confined to land preparation and seeding.
Annual employment per feddan was 5.7 and 4.9 mandays for sorghum and
sesame respectively.(l)

The extent of the welfare gain to these migrant workers and
their households is -discussed in Affan (1981). While the rainfed
mechanized farms in central Sudan have offered a substantial
(50 percent) wage differential to their hired workers as compared with
the return obtained in the more densely settled Gezira area, various
additional costs must be incurred in obtaining this income. These
include transport, accommodation and.subsistence while away from home,
plus the opportunity cost of any farm output foregone by the labour
supplying household as a result of the worker's absence during much of
the cropping season. Affan reports that in extreme cases such
opportunity costs alone could amount to 60 percent of own-farm output.
Clearly they depend on the extent to which other household members can
compensate for the absent worker. However, as long as supplying house-
holds retain their own farms the potential (fall-back) income from the
latter enables hired workers to set a minimum supply price for their
labour that compensates them for the costs incurred. In semi-arid
farming areas regular seasonal wage-employment offers the attraction
of a more secure income: the risks of harvest and income failure are
transferred from the small farmer to the large one. In the Sudan, the
welfare gains to seasonal workers would become less secure, however,
if further expansion of mechanized farming were to lead ultimately to
displacement of the labour supplying households from their land.

Most farmers in Sudan's mechanized rainfed farming sector grow
non-combinable sorghum varieties (Harbi, 1985). As long as this
continues, and as long as sesame harvesting also remains unmechanized,
the sector will generate substantial labour demand. However, in
general farmers may respond to labour scarcity in newly cultivated
areas by fuller mechanization, if the nature of the crop permits this.
This occurred, for instance, in the Gorgan area of Iran in the 1950s

(1) Figure derived by Affan from a sample survey of mechanized farms
in S. Kordofan conducted in 1976.

- 85 -

and 1960s, where wheat farmers rapidly introduced combines (although
when the Gorgan farmers switched to cotton production a pattern of
partial mechanization re-emerged(l)). Combines are also used on the
new wheat farms in Saudi Arabia. A switch to combining has also been
recommended for the Sudan (Nour, 1985).

Meanwhile, in both Sudan and Saudi Arabia, pastoralists have
lost land rights due to the development of mechanized agriculture.

2. The mechanization of crop production on land already

Since in much of the Near East the positive yield effects of
mechanizing crop production on land already cultivated are still
limited, the main social justification for such mechanization must
presumably come from its role in supplementing inadequate labour
supplies at peak periods. Certainly, the labour-saving effects of
mechanizing peak season operations (land preparation, harvesting) are
high. However, what is less clear is the extent to which in practice
such mechanization either (a) acts as a supplement to alternative, but
inadequate energy sources, or (b) displaces labour for part of which
no adequate alternative employment exists, and/or (c) perpetuates
existing un- and underemployment.

One view is that in the Near East any adverse employment effects
have been offset by (a) rising rural labour demand due to expansion
both of irrigation and of off-farm jobs linked to mechanization and to
expanding farm output and incomes, and (b) falling rural labour supply
in some parts of the region.(2) However, even if this analysis is
broadly correct, we should not be too complacent about its
implications for continuing mechanization, for two reasons:

i) as already noted, agricultural labour supply in most countries
in the region is still rising. Further mechanization could in
some instances create future labour market imbalances, even at
peak seasons;

ii) even where the overall rural employment situation seems fairly
healthy, with no widespread and conspicuous unemployment
arising from mechanization, there may still be groups of rural
households, most probably amongst the poorest, who have lost

It is true that farm mechanization, especially if output
raising, generates expanded off-farm employment in activities such as
input supply, repair and maintenance, transport and processing.
However, it is certainly not inevitable that increased employment in
these occupations will exactly offset decreased employment in farming.
The likelihood that mechanization will create excess labour supply is
greatest where prices fail to reflect the relative opportunity costs
of labour and capital cheapening the latter relative to the former,

(1) Okazaki, 1985.
(2) See Ahmed and Amjad, 1984:175-177. For a more detailed analysis of
the relationship between labour migration and mechanization see
Richards, 1989.

- 86 -

and/or where the agrarian structure is characterized by high
population density combined with unequal distribution of land rights.

In either case an upward trend in labour supply for demographic
reasons increases the probability of excess supply occurring.

In view of the upward labour supply trends just noted, the
following paragraphs, which review illustrative evidence, chiefly from
the Near East region, on the rate at which mechanization of crop
production replaces labour, assume a particular significance.

Table VIII provides illustrative data from Egypt on the impact
of motorization of seed-bed preparation on labour demand. For tillage
alone the data show a decline from 4 to 0.5 man-days per feddan
(equivalent to a fall from approximately 9.5 to 1.2 man-days per
hectare). The labour demand impact of motorization of the same
operation in different regions is unlikely to be identical. A variety
of factors, including soil type, tractor size, and the proficiency of
machine operators will influence the outcome. Nonetheless, Ansari and
Raza report, without citing their data source, the same 8:1 ratio
found by Dyer and Imam in Egypt as applying also in the Pakistan
Punjab(l) (where agro-climatic conditions are reported to be in many
ways similar to those in Egypt(2)).

In the irrigation command area of the Pakistan Punjab Bergmann
reports a drop in total man-days per hectare on tractorized farms to
30, from the 81 days/ha expended on bullock-farms (Bergmann, op.cit.:
132 and 199). While the decline appears very large it may include
labour savings in on-farm transport (see p. 88 below). However, this
order of magnitude is in any case supported by the findings of
Cleaver's farm model for a 120 ha rainfed farm in Morocco. The model
incorporates the assumptions that following mechanization fodder land
is released for crop production and that yields rise by 0.3 ton/ha
(see p. 82 above). Nonetheless, labour requirements per hectare are
reported to fall by 42.5 days/ha if a tractor alone is purchased, and
a further 9.17 days/ha if a combine harvester is also acquired. (No
yield effect is anticipated from the introduction of combine-
harvesting.) In this model no trailer purchase is anticipated.(3)

(1) Ansari and Raza, 1984.
(2) See Gotsch, 1983:270. However, Agarwal found in the Indian Punjab
that the human labour hours required per hectare for ploughing
with bullocks were only some 540 per cent greater than for
tractorized land preparation, as compared with the 8:1 ratio found
by Dyer and Imam (see Table VIII). See Agarwal, 198 :119.
(3) See also Azzam, forthcoming, who reports that in Morocco the joint
use of a tractor and combine harvester reduces labour requirements
by 77 percent for hard and soft wheat and 74 percent for barley.
Tunisian agriculture of the 1930s also provides evidence of the
impact of mechanization on labour demand in semi-arid rainfed
agriculture. The farms concerned were based on cereal monoculture
interspersed every second year with a ploughed but un-planted
fallow. Here labour requirements per hectare fell to as little as
5 days per year following mechanization (Maklouf, 1979:6).

- 87 -

Farm mechanization also increases wage labour supply in relation
to demand when farmers who have bought tractors seek to expand their
mechanized area by displacement of tenants and land rent or purchase
from low income farm households whose own farming methods are labour-
intensive. This occurred, for example, in Iran following the 1962 land
reform, and there is also evidence of its occurrence in Pakistan. In
his survey of tractor owners in the Pakistan Punjab Bergmann reports a
5.3 percent increase in land owned, a 125 percent increase in land
leased-in, and a 53 percent decrease in land leased-out following
tractor purchase. (Most owners had acquired their tractors within the
previous five years.) The increase in land owned by 94 tractor owners
was equivalent to the total farm area of 'almost 100 of the smallest
and small holdings in the villages surveyed (Bergmann, 1984:125).

D. Small-scale Tractors: Their Viability and Employment Impact

Where motorization of crop production is based on smaller-scale
more labour-intensive equipment (single-axle tractors or small four-
wheel machines) labour displacement per hectare cultivated is likely
to be less than with larger machines. However, single-axle rotovator
cultivators are only technically suitable for a limited proportion of
the cultivated area in the Near East. They are best suited to land
preparation in irrigated rice production and to inter-row cultivation
of horticultural crops. They are, furthermore, relatively expensive
machines per hectare of work capacity as compared with larger tractors
(Gifford, 1980; Panhwar, 1984). In the Near East region they appear to
be extensively used only in the irrigated rice region of the Caspian
littoral in Iran (Okazaki, 1985).

In Cyprus two-wheeled tractors have been adopted in vine-growing
areas where four-wheeled tractors were not suitable 'due to the
topography and the lack of roads'. Since their adoption the vine
acreage has been increased (Panayiotu, forthcoming).

Small four-wheel machines are also more expensive per hectare of
work capacity than larger ones.(A given percentage fall in horse-power
generates a smaller percentage fall in capital and operating costs.)

Agro-engineering research indicates that, for tillage opera-
tions, tractors below 47 h.p. also generate a lower quality
performance than tractors above this threshold, while machines of this
size, correctly equipped, perform as well as larger tractors.
Interestingly, in Pakistan, farmers interviewed by Bergmann indicated
a preference for tractors with an average of 52 h.p. (The actual range
was from 35 to 65 h.p.)(l). Smaller tractors, however, may perform a
range of other functions perfectly adequately (e.g. seeding,
fertilizer application, spraying and transport both on and off the
farm). From an engineering point of view they are best used as a
supplement to larger tractors.(2) However, farmers facing a capital
constraint yet wanting both to mechanize and the autonomy of machine

(1) Bergmann, 1984:118.
(2) I am indebted to R. Gifford, Chief, Agricultural Engineering
Service, FAO, for the information in the preceding part of this

- 88 -

ownership may sometimes prefer to buy these, and to use them also for
tillage, as an alternative to custom-hire.

E. The Impact of the Mechanization of Crop Production on
the Structure of Agricultural Employment

The mechanization of crop production has also affected the
structure of agricultural employment. Firstly, it has raised the stock
of mechanical skills embodied in the labour force. Secondly, it has
tended to modify the ratios of family to hired, and of permanent to
casual labour employed in agriculture, although the pattern of these
changes varies between farm systems. In the irrigation command area of
the Pakistan Punjab an increase of about 10 percent in permanently
hired labour and a slight decrease in casual labour have been recorded
on tractor-owning farms, the latter despite an average extension of
21 hectares in their cultivated area.(1) In Tunisia, however,
mechanization of semi-arid rainfed cereal monoculture led to a
decrease in permanently hired workers and increased reliance on casual
labour at peak periods.(2) In Sudan, too, tractorized rainfed
agriculture makes heavy use of casual labour.

In irrigated areas the level and structure of employment on
mechanized farms often reflects the interaction of mechanization of
water adduction and crop production. Labour savings by the latter have
to varying degrees offset increased labour demand stemming from new
investments in irrigation,in improved bio-chemical technology, and, in
some cases, changes in cropping pattern towards more labour-intensive
crops. Sometimes the outcomes have been aggregate labour requirements
per hectare remarkably similar to premechanization levels, but more
evenly spread over the year.(3) However, the scope for such offsetting
changes has varied between regions, as do the opportunities for future
development. Expanding herbicide use is also lowering the labour
demand effects of investments in bio-chemical technology.

Meanwhile, in rainfed areas labour reabsorption following
mechanization of crop production must often occur largely outside
farming within those areas (in local non-farm production or through
migration to work in other farming areas, in towns or other
countries). Difficulties with reabsorption may occur not only because
of aggregate imbalances between labour supply and demand, but due to
the socio-economic circumstances of those displaced (e.g. unskilled
casual labour from low income households, possibly with low mobility).


Notable developments in the mechanization of post-harvest
operations on Near East farms have occurred in transport and
threshing. Widespread use of tractors for transport has often been
noted.(4) In Pakistan, for example, trailers are the second most

(1) Bergmann, 1984.
(2) Maklouf, 1984:6.
(3) Cf. Bartsch, 1977. It is the combined effect of these changes that
has led to the increased use of family and permanently hired
labour, and reduced use of casual labour.
(4) See e.g., Binswanger and Donovan, 1987:14.

- 89 -

numerous implement acquired by tractor owners (Table VII), but little
attempt has been made either in the Near East or elsewhere to assess
the significance of this development for agricultural output and
employment. Mechanization of threshing has received more attention.

Various output effects of mechanized threshing have been noted.
In the Pakistan Punjab, mechanization of threshing has helped to
increase cropping intensity, and has contributed to an increase in the
proportion of land under wheat (Bergmann, 1984:135). In north-east
Syria farmers find that stationary motor-driven threshers produce
straw of lower quality for livestock feed, and generate higher straw
and chaff losses than the traditional animal-drawn jerjer (a 6 to
7 percent loss as compared to 2 percent)(1) Again, however, the
motor-driven thresher has greater capacity and speed, which may enable
farmers to achieve other farm operations in a more timely manner.

Estimates of labour saved by mechanized grain threshing range
from a reduction to one-fifteenth of the previous requirement (Bailey,
1980:6) to a lower, but still substantial, proportional saving
amounting in absolute terms to 7.2 days per feddan, or approximately
17.28 days per hectare per season, in Egypt (Dyer and Imam, 1983:
177). In the early 1980s in Sharqiyya Governorate in Egypt, on all
except the smallest farms, mechanized threshing was found to be
profitable when evaluated from both a private and social perspective
(see Table IV) because it releases labour at a peak period.

In double-cropping zones, speedy harvesting may be needed to
free land for preparation for the next crop; but mechanization of
harvesting entails a high capital outlay, is only feasible for some
crops and is sometimes also impeded by topography. Mechanization of
threshing, which entails a lower capital investment in equipment that
can often be locally manufactured, may release human energy either for
harvesting or other competing operations.


Agricultural mechanization impinges on male and female labour
according to the existing division of labour and the nature of the
operations mechanized. In the Near East it is largely, but not
entirely, male labour that has been affected by mechanization.
However, to fill out the picture, this section focuses on the impact
on women's labour.

In Near East countries women's role in farming is in livestock
maintenance (where they are assisted by children)(2), in less energy
intensive field operations including planting, weeding and harvesting,
and in the post-harvest operations of threshing and winnowing.
However, there are cultural variations within the region in the extent
to which it is regarded as acceptable for women to be involved in
field tasks (see e.g. Hopkins, 1987:83).

(1) Bailey, 1980: 5, 6.
(2) See Hopkins, 1987: Chapter 6, and Palmer, 1985.

- 90 -

Regarding livestock maintenance the replacement of animal
draught power by machine-power might be expected to reduce the demand
for women's labour. However, in an era of rising per capital incomes
this effect has often been offset by the rising demand for animal
products, which has sustained the demand for women's labour in the
livestock sector (see e.g. Richards and Martin,1983; Hopkins,1987:90).

The use of motor power to increase irrigation and, hence, crop
output and cropping intensity, raises the demand for women's labour in
the tasks that they perform, unless there is a concomitant shift in
cropping pattern towards crops that are regarded as men's
responsibility. One example of the latter is the use of irrigation to
expand 'qat' production in the Yemen Arab Republic. However, Palmer
(1985) suggests that this expansion, occurring at a time of heightened
male out-migration from agriculture, may also have been associated
with increased involvement of women in the production of this crop.

In the Yemen Arab Republic, following investment in well and
pump irrigation by small-scale farmers on the coastal plain, women
work more on the family farm and less on outside work(Mundy, 1985:31).

Turning to mechanization of crop production, mechanized land
preparation does not usually replace women since this is essentially a
male task, but mechanized seeding may do so. Secondly, where tractori-
zation has sustained the cultivated area in the face of increasing
male out-migration it has also sustained the need for women's labour
in non-mechanized tasks. Thirdly, mechanization has sometimes
increased demand for female labour: in Turkey, mechanization increased
the demand for women's labour in certain non-mechanized tasks hoeing
and harvesting cotton, harvesting tobacco, citrus and hazel-nuts
(Mulayin, 1985:93). In the Sudan, women supply up to 30 percent of
hired labour on some mechanized farms in the rainfed areas, where they
work chiefly on weeding, gleaning and winnowing (Sammani, 1987:5).

The picture changes when tasks that sometimes involve women -
notably harvesting and threshing are also mechanized, although it
may still be male labour that is predominantly affected. No
quantitative evidence on the share of female labour time in these or
other tasks is available from within the Near East region. Table IX
therefore reproduces for illustrative purposes relevant data from the
Indian Punjab. These reveal that, on those farms (28 percent) that
used female labour in HYV wheat cultivation, women provided less than
10 percent of all labour time for the crop, chiefly in harvesting
(where they supplied 16 percent) but also in sowing (19 percent) and
threshing (6 percent). These data may also be reasonably
representative of the small farm sector in the Pakistan Punjab.

In the Near East, whether the displacement of women's labour
from agriculture is regarded as 'good' or 'bad' appears to depend
largely on the socio-economic status of the household. Among wealthier
rural households the withdrawal of women from 'visible' production is
often seen to be desirable for cultural/religious reasons. Women
themselves may also value positively the status and less arduous work
associated with this withdrawal. On the other hand, women from low
income households depend partly on wage labour to augment household

- 91 -

income. Loss of this income is particularly serious when the household
has no adult male present and capable of working.


The distribution of land rights can affect the scale of
decisions to mechanize,the extent to which such decisions are
responsive to the overall availability of agricultural labour, and the
subsequent pattern of change in labour use and labour supply following
initial mechanization.

A. Distribution of Land Rights and the Decision to
Invest in Motorized Equipment

Lack of secure land rights may deter fixed investment in
agriculture, but it is doubtful whether most investments in mechanized
equipment should, in this context, be regarded as fixed, other than
permanently sited irrigation pumps. Generally, within existing rural
communities it is not insecure land rights that constrain investment
in motorized equipment, but rather the low wealth and income, and
restricted access to credit, which limited access to land, however
secure the rights, often represents. Where small-scale farmers have
seen a profitable market for custom-hire, and have been able to raise
the necessary financial resources, they have invested in tractors and
occasionally combines too (see e.g. Okazaki, 1985).

For potential investors coming from outside a community,
however, secure land rights in the form of freehold tenure or secure
tenancy are generally an essential precondition for medium- to
long-term investment in agriculture. Where traditional tenure systems
do not provide secure guarantees of land rights to outsiders, state
intervention has sometimes been used to guarantee these for example,
by overriding traditional tenure systems and abrogating to the state
the right of land disposition, as has occurred in the mechanized
farming areas of the Sudan, and in Saudi Arabia (Joffe, 1985). In both
cases pastoral communities lost grazing rights.

Where mechanized farming has led to major extensions in
cultivated area this has often been associated with changes in the
socio-economic characteristics of farm owners, with substantial
investments in mechanized farming by wealthy urban-based individuals.

Agrarian reform may also promote agricultural mechanization
through the conditions attached to acquisition of land rights. Thus
the 1962 Iranian reform exempted landlords from redistribution of all
mechanized farm land and bound them to work with agricultural
machinery on the lands that they were allowed to retain (Okazaki,
1985). More egalitarian land reforms may also promote mechanization
through creation of mechanized producers' cooperatives (as in
Algeria)(1) or establishment of state- or cooperative-owned machine

(1) In Algeria, after independence, mechanized cooperatives were
created on former colon farms. In the 1970s about 6,800 new
production cooperatives were created, partly through expropriation
of land from all private farms over 50 hectares. The area in
producer cooperatives rose at this time from 35 to 50 percent of
the total farm area (Cleaver,1982:50; see also Bennoune, 1988).

- 92 -

hire stations with central coordination of land use, the land itself
being owned by small farmers (as in Egypt, Syria and Iraq). In these
cases some compulsion may be attached to the use of machine power.

In areas characterized by a high degree of farm fragmentation
and/or the presence of many small farms, the last two types of reform
can facilitate the spread of mechanization to poorer farm households
in ways which should both be equitable (between farmers) and ensure
efficient machinery use.(l) Both arrangements are, however, reported
to have given rise to problems. In Algeria, these relate not just to
machine use, but to overall motivation and efficiency, given both the
rules governing the operation of producer cooperatives and the
official prices for farm produce (Cleaver, 1982:50). In Egypt, the
problems are reported to relate to central land use planning and to
poor maintenance of publicly owned machinery. Thus:

'The small plots to be worked at any one time on each small farm
resulting from the complex rotation impose unduly high costs on custom
service operations which cannot be sustained at the low utilization
rates actually achieved without substantial subsidy.' (Cuddihy,

Cuddihy further reports that 'private custom hire services are
more efficient than the cooperatives', achieving an average of
907 hours annual usage as compared with 557 (the break-even level is
1,000 hours)(2).

Meanwhile, fragmentation remains a widespread problem in the
Near East (Abu Oaf and Georgiades, 1978; ESCWA, 1985) (3), which
raises unit costs of mechanized crop production, largely due to the
amount of unproductive time spent travelling between plots. In Cyprus,
a further problem arises from the drift from mechanized spraying on
small plots.(4)

B. The Distribution of Land Rights and the Impact of
Mechanization on Agricultural Labour

Following successful investment in motor power many farmers try
to expand their cultivated area.

Even if the initial investment is in mechanized irrigation, as
the cultivated area is extended this is likely to be followed by at

(1) The latter may of course still be labour displacing.
(2) See also Mahdi, forthcoming.
(3) While it is sometimes suggested that this is due to the operation
of Muslim land law, the problem is in fact more widespread. Some
of the highest fragmentation has been found in Cyprus. Any
egalitarian inheritance system can generate this problem. It is
further exacerbated when land potential varies widely, within
limited areas, because then each plot is likely to be subdivided
at succession.
(4) On consolidation programmes in the Near East, see Abu Oaf and
Georgiades, 1987:18, 19, and ESCWA, 1985:11-13.

- 93 -

least partial mechanization of crop production and crop processing,
particularly on larger farms. Whether, and how, expansion of the
mechanized area affects farm labour depends on a number of factors,
including: population density in mechanized areas, whether the labour
market was previously in balance, the extent of offsetting changes in
cropping pattern and bio-chemical technology, and the degree of
mechanization, which is itself partly a function of the scale of farm
size expansion. The scale of farm size expansion is in turn likely to
be influenced by the pre-existing land tenure system.

It is in densely populated areas that the mechanization of both
crop production and processing is most likely to promote an expansion
of wage labour supply in excess of demand. The likelihood of this
occurring is increased where short-term and insecure share-tenancy is
widespread, and is also enhanced by an unequal farm size distribution,
with large numbers of small, poor farmers, some of whom may be easily
induced to sell or lease out their land. Tunisia in the 1930s, Iran
during the 1960s and, more recently, Pakistan, have experienced
widespread displacement of share-croppers and a reversal in the
direction of land-leasing following the introduction of mechanization
(Maklouf, 1979; Okazaki, 1985; Bergmann, 1984).

In Turkey also it is reported that use of modern machinery
contributed to elimination of some sharecropping arrangements, thus
increasing the landless households (Balkir, 1984:125; Mulayin,

In contrast, a more even farm-size distribution, and greater
security of tenure, might be expected to generate a pattern of
mechanization more appropriate to rural labour force endowments. The
problem, however, is that if security of tenure is based on individual
freehold, it may be difficult to modify the farm size structure as
economic development occurs: while emerging labour scarcity may make
it economically efficient to amalgamate the smaller farms and to
increase mechanization, individual attachment to the land impedes

In densely populated regions and/or where population growth is
high, it may be preferable to base private land tenure on leasehold
from the state. (Farm leases might be for about 20 years, unfinished
leases being inheritable, but with limits on subdivision.) By regular
review of the land:labour situation it should be possible to adjust
farm sizes over time to either a rising or declining rural labour
force. (For example, the Sudan Government currently provides 25-year
leases for large mechanized farms in the rainfed areas, while farmers
on the major irrigation schemes in Sudan also have leasehold tenure.)


In addition to the structure of land tenure, price policy is
also an important influence on agricultural mechanization. Government
policies with respect to the foreign exchange rate, the rate of import
duty on agricultural machines, equipment parts and fuel, domestic
subsidies on these items, formal sector interest rates and farm output

- 94 -

prices: all have influenced the mechanization process and in many
cases they have tended to accelerate it either deliberately or
incidentally, as a spin-off from the pursuit of other policy goals.

Cheap agricultural credit has been widespread in the region, and
has tended to lower the price of capital equipment relative to labour.
For example, in the Maghreb in the early 1980s, real agricultural
interest rates were: Algeria -9.5 percent; Morocco +1.2 percent; and
Tunisia -1.5 percent (Cleaver, 1982:46). Other things being equal, a
negative real cost of borrowing can be expected to encourage capital
investment. In Algeria, this period was in fact associated with a
marked increase in the rate of mechanization investment (see Tables IB
and ID).

Policies with respect to the price of, and access to, foreign
exchange, and with respect to duties on imported equipment and fuel,
have been more varied, and with respect to import duties have also
sometimes been offset by other elements of domestic price policy. For
example, zero import duties on farm machinery and implements have
formed part of a package of measures to encourage agricultural
mechanization in Jordan. In Morocco, farm machinery in the early 1980s
faced import duties and indirect taxes of about 30 percent. However,
these taxes were offset for Agrarian Reform Cooperatives and Farmers'
Associations by government subsidies of about 25 percent of the
domestic price of tractors and 20 percent for combine harvesters
(Cleaver, 1980:11).

In some cases, as in Egypt, heavily subsidized fuel prices have
accentuated the tendency to mechanize. As noted earlier, in Egypt as
well, mechanization has been encouraged by the high (uncontrolled)
prices for livestock products relative to the controlled prices for
most major crops (Richards and Martin, 1983).(1) The extent of price
distortion in Egypt is reflected in the wide differences in the
financial and economic rates of return to investment in farm equipment
(see Table IV).

In view of the changing labour supply situation in some
countries, and in view of the potentially adverse efficiency and
equity consequences of inappropriately high rates of mechanization,
there is a need for a review of the overall impact of pricing policy
on farm mechanization in most Near East countries (see e.g. Cleaver,
1980; Richards and Martin, 1983).


While land tenure and pricing policies have influenced
mechanization, so too have foreign equipment suppliers and foreign aid
agencies, most notably through the provision of easy credit terms

(1) The Egyptian Government has probably not been helped by the con-
flicting advice that it has received on the interrelated issues of
mechanization and livestock product pricing. Contrast,for example,
Saab, 1967:108, 9 and ERA 2000 INC, 1979, with Cuddihy, 1983.

University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs