• TABLE OF CONTENTS
HIDE
 Front Cover
 Title Page
 Table of Contents
 Foreword
 I. Setting the scene in Turkey
 II. Farm and farmer characteristics...
 III. Farmer incentives for adoption...
 IV. Adoption and characteristics...
 V. Multivariate analysis of factors...
 VI. Summary of findings and policy...
 Copyright






Group Title: adoption of new bread wheat technology in selected regions of Turkey
Title: The adoption of new bread wheat technology in selected regions of Turkey
CITATION THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00080049/00001
 Material Information
Title: The adoption of new bread wheat technology in selected regions of Turkey
Physical Description: vi, 27 p. : ill. ;
Language: English
Creator: Demir, Nazmi
International Maize and Wheat Improvement Center
Publisher: Centro Internacional de Mejoramiento de Maíz y Trigo
Place of Publication: Mexico, D. F.
Publication Date: 1976
 Subjects
Subject: Wheat -- Turkey   ( nal )
Genre: international intergovernmental publication   ( marcgt )
bibliography   ( marcgt )
non-fiction   ( marcgt )
Spatial Coverage: Turkey
 Notes
Statement of Responsibility: Nazmi Demir; edited and abridged by CIMMYT.
Bibliography: Includes bibliographical references.
 Record Information
Bibliographic ID: UF00080049
Volume ID: VID00001
Source Institution: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 03985394

Table of Contents
    Front Cover
        Front Cover
    Title Page
        Page i
        Page ii
    Table of Contents
        Page iii
        Page iv
    Foreword
        Page v
        Page vi
    I. Setting the scene in Turkey
        Page 1
        Page 2
    II. Farm and farmer characteristics in the regions studied
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
    III. Farmer incentives for adoption of recommendations
        Page 8
        Page 9
        Page 10
        Page 11
    IV. Adoption and characteristics of the farm, farmer, and government programs
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
    V. Multivariate analysis of factors affecting adoption
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
    VI. Summary of findings and policy implications of the study
        Page 25
        Page 26
        Page 27
    Copyright
        Copyright
Full Text


















THE ADOPTION OF NEW BREAD WHEAT TECHNOLOGY IN SELECTED
REGIONS OF TURKEY
Edited and abridged by CIMMYT















Nazmi Demir
























CENTRO INTERNATIONAL DE MEJORAMIENTO DE MAIZ Y TRIGO 1976
International Maize and Wheat Improvement Center, Apartado Postal 6-641, Mexico 6, D.F. Mexico
























Correct citation: Demir, Nazmi. 1976. The adoption of new
bread wheat technology in selected regions of Turkey--Edited
and abridged by CIMMYT. Centro Internacional de Mejora-
miento de Mafz y Trigo. Mexico City. vi + 27 p.
















CONTENTS


Foreword v


I. Setting the scene in Turkey 1

Sources of production 1

Coastal spring wheat programs 2

Review of subsequent chapters 3


II. Farm and farmer characteristics in the regions studied 3

Sampling strategy 5

Regional characteristics from survey data 7


III. Farmer incentives for adoption of recommendations 8

Measures of profitability of recommended procedures 8

Measures of risks of recommended procedures 10

Summary 11


IV. Adoption and characteristics of the farm, farmer, and government programs 12

HYV's and individual variables 12

Fertilizer and individual varieties 16

Other elements of the technology 18


V. Multivariate analysis of factors affecting adoption 19

Factors affecting adoption of HYV's 20

Factors affecting application of fertilizers 23


IV. Summary of findings and policy implications of the study 25

Findings 25

Policy implications 26


























FOREWORD:ON LAUNCHING THE STUDIES AND MOTIVATING HYPOTHESES


Launching the Studies

The study described in the following chapters is one of
a series aimed at enlarging understanding of the factors
impinging on the adoption of new maize and wheat
technology. Better understanding of the elements shaping
the diffusion of new cereals technology can help govern-
ments and development assistance agencies to increase
farmer income, hence the interest in the topic. Interest
increased as controversy about effects of introducing new
technologies attracted widespread attention to the theme.
CIMMYT, with its mandate defining its role in the
development and diffusion of maize and wheat technology,
quickly assumed a participant's role in the discussions. The
concern and the interest emanating from the critical im-
portance of the theme stimulated CIMMYT to look for a
modus operandi through which patterns of adoption and
the forces shaping those patterns could be identified.
Better understanding of these relationships would influence
CIMMYT efforts to develop new technology, the orientation
of its training program, and the approach taken in counsel-
ing governments about national programs.
In order to better comprehend what influences farmer
response to new technology, CIMMYT set out to facilitate
the research on which this and the other studies of the
series are based. We decided to examine eight cases in
which maize or wheat technology had been introduced to
farmers. In identifying programs for study, we limited
consideration to those in which the technology had been
available to farmers for no less than five years and in
which no less than 100,000 hectares of land might have
been affected. Eight programs were selected for study. For
maize the focus was on Colombia, El Salvador, Kenya west
of Rift Valley, and Mexico's Plan Puebla. For wheat,
programs in India, Iran, Tunisia and Turkey were consider-
ed. CIMMYT's maize and wheat staff participated in the
selection of these programs. With their knowledge of
programs around the world it was possible to choose a
varied set of experiences-e.g. programs with and without
irrigation, w:th and without effective price guarantees,
with massive extension effort and with virtually none.


To the extent possible, each of the adoption studies was
under the supervision of an indigenous economist. In only
one case was it necessary to turn to an expatriate and
there we had the good fortune to collaborate with a re-
searcher with several years experience in the area. Each
of the collaborators shared CIMMYT's concern for farmer
response to new technology.
Beyond sharing this concern, each collaborator had an
interest in farm level research done in close cooperation
with agricultural scientists. The importance of this interest
emerges from our conviction that agricultural scientists who
are knowledgeable about a particular maize or wheat area
can contribute substantively to research on the cereals
economy of that area. Their special knowledge about the
interaction between plants and their environments is im-
portant in identifying agro-climatic zones, critical periods
for the crop, and activities which are essential to effective
cultivation. Many agricultural scientists played a prominent
role in these studies; each warrants our gratitude for his
contribution.
As the studies were completed it became apparent that
much could be said for publishing them in a standard
format. With several serving as Ph.D. dissertations and
others as less formal research pieces, a common format
could only be achieved through reworking the original
monographs. In every case but one, then, CIMMYT's
publication is an abridgement of a longer piece. The
Indian study, itself a review of the findings of several other
research efforts, is being published in its entirely with no
effort to recast it in the form of the others.
In making the abridgement we have followed certain
norms. Mathematical proofs have been eliminated, litera-
ture reviews have been included only where they relate to
points which are unique to a given study, and the discussion
of the hypotheses motivating the studies have been dropped.
This last decision arises from recognition of the substantial
commonality of these hypotheses among the studies. This
suggested that, rather than presenting essentially the same
discussion in the text of each abridgement, the hypotheses
could be treated once in an abbreviated form for all studies.
That treatment follows below.







The Hypotheses

While each of the studies examines a somewhat different
set of circumstances, all depart from the same general
assumption about farmer behavior. The assumption is
that farmers are income-seeking risk averters who are
sensitive to the nuances of the environment in which they
farm and that they are generally effective in their decision
making. For the six studies based on original survey data
and to a more limited extent for the study of Plan Puebla,
this common point of departure leads to a great deal of
similarity in the motivating hypotheses.
Given a farmer oriented by the assumptions described
above, we might expect to see relationship between the
adoption of elements of the new technology and 1) char-
acteristics of the farmer-his age, education, family size,
farming experience, off-farm work, percentage of land own-
ed, 2) characteristics of the farm-its agro-climatic region,
competition of industrial crops, relative importance of
cereals, nearness to markets, farm size, 3) characteristics of
government programs-access to credit, access to informa-
tion (though extension agent visits or visits to demonstra-
tion plots).
Some of the relationships between these variables and
the adoption of elements of the new technology are more
arguable, some less. Least arguable are hypotheses relating
adoption to education, farming experiences, percentage of
land owned, more favored climatic regions, relative import-
ance of cereals, nearness to markets, farm size, access to
credit, and access to information. With other things equal
and accepting our assumptions that farmers are income-
seeking, risk-averting, sensitive, and effective maximizers,
virtually no one would argue that any one of these relation-
ships should be negative.
Somewhat more arguable is the relation of age and family
size to adoption. Even here it is likely that only a few
would argue that these relationships might be positive.
Most arguable are the relationships linking adoption to
off-farm work and competition of industrial crops. With
respect to the former, some hold that the relationship is
positive as more off-farm work implies more income, there-
fore a greater capacity to bear risk, hence a greater willing-
ness to adopt new technologies. Others hold the converse,
arguing that more off-farm work implies less interest in the
farm, hence less willingness to put in the time and energy
associated with taking on new technologies. So too for
industrial commodities, where those who see the relation-
ship as positive allude to greater experience with improved
inputs and larger incomes while the contrary view rests on
capital restrictions and the high opportunity cost of labor.
With knowledge of the relationships among these var-
iables, researchers and policy makers can better develop and
diffuse new technologies. Some of the variables considered,
e.g. age and family size, are beyond the control of these
decision makers. Nonetheless, by incorporating them in the


analysis the effects of variables subject to their control are
more clearly discerned. Knowledge of how these variables,
e.g. agro-climatic zones and extension programs, relate to
adoption can be of critical importance in affecting the
development and diffusion of new technology.
With this rough sketch of the general argument, readers
wanting more detail about the derivation of the hypothe-
sized relationships can turn to the relevant original piece
from which this series of abridgements was drawn. In
all cases the studies feature the effects of agro-climatic
region and farm size on adoption of elements of new
technology. This emphasis is related to the earlier contro-
versy about the effects of new technology where these two
factors played prominent roles.
Before moving to the abridgement, some attention to
the phrase "elements of the new technology" is warranted.
Much has been made of the concept of a package of practi-
ces in the introduction of new technology. We've chosen
to look at this a bit differently, taking the view that the
differences in risk, expected income, and cost of each
element of the technology are large enough to outweigh
the effects of the interaction among these elements. That
is to say, perceptive and prudent decision makers might
well choose to take up only a part of the package rather
than the entire package. For the programs studied, the
two dominant elements in the package are improved seed
and fertilizer. These two were analyzed as dependent var-
iables in each of the studies. Of lesser importance are
such elements as seed treatment, date of planting, method
of planting, use of herbicides, use of pesticides, planting
density, and seed bed preparation. Nevertheless, where
any of these was recommended and where data are adequate,
these are also treated as dependent variables.
While CIMMYT has been associated with these studies
since their inception, the opinions expressed by the authors
are not necessarily endorsed by CIMMYT.

What follows

This report summarizes results of a study of wheat
technology in Turkey's coastal areas. The study is based on
a survey of over 800 wheat farmers conducted in early
1973. That survey was a part of a larger effort under the
direction of Dr. Resat Aktan and supported by Turkey's
Ministry of Agriculture, the State Institute of Statistics
and CIMMYT. Dr. Aktan is now examining the entire
survey, emphasizing the two winter wheat areas excluded
from this report where few farmers had taken up new
varieties.
The study on which this abridgement is based was
undertaken by Dr. Nazmi Demir while a visiting scientist
at CIMMYT.
Don Winkelmann


El Batan



















I. SETTING THE SCENE IN TURKEY


Turkey's marked dependence on wheat is evidenced in that
crop's role in national consumption and in production. On
the consumption side, the country's 38 million people have
an average annual consumption of some 200 kilos per person
a year, among the highest in the world. Wheat supplies
roughly 50 percent of the calories and a bit over 50 per-
cent of the protein in the diel of the average Turkish
consumer.
On the production side, 8.6 million hectares of land
were devoted to wheat in 1972. When the land in fallow-
another 8.6 million hectares is added to this, the wheat
industry accounts for some 70 percent of the country's
tillable land. Wheat's vital contribution to Turkish agri-
culture is evident. Given that agriculture employs roughly
two thirds of the population, accounts for about a third
of GNP, and supplies nearly nine tenths of the country's
exports, wheat's overall importance to the Turkish economy
is also evident.
With this notable reliance on wheat, the Turkish govern-
ment has long sought to stimulate production. Its first
goal, only occasionally achieved and then only in good
years, is to make the country self sufficient in the basic
product. Area, yields, and year-to-year fluctuations combine
with consumption to make Turkey an importer in all but
the best years. These imports range up to 8 percent of an-
nual use.
When considered in the context of a growing popula-
tion and increasing incomes, the attainment of self suf-
ficiency is seen to be difficult. Population is growing at
the rate of 2.6 percent per year. Annual per capital income
is increasing at a rate of over 2.5 percent per year. Even
assuming a relatively small relationship between changes in
income and changes in consumption it is clear that produc-
tion must increase at a more than 3 percent per year.
Closing the gap now filled with imports implies an even
greater annual increase in annual production. Contrast
this target, however, with the experience of the 15 years,
1958 to 1972, when area increased little and yields increased
by a bit less than 2 percent per year.



Sources of Production

The bulk of Turkey's wheat production is now carried on
in two quite distinct environments. The largest contributor,


both in terms of area and in terms of production, is the
winter wheat region of the Southeast, the Anatolian
Plateau, and Thrace. These regions account for about 60
percent of the total production on somewhat more than
60 percent of the wheat area exclusive of fallow. Nearly
three fourths of the fallow lies in the South East and the
Anatolian Plateau.
There is little opportunity to expand production through
expansion in area. In fact it is the hope of the government
that area in the Anatolian Plateau can be reduced by 500,000
hectares, with all of this returned to pasture.
Yields of winter wheat in Turkey are generally low, on
the order of 1.0 to 1.2 tons per hectare. Yields are also
characterized by substantial year-to-year variations because
of the substantial variation in weather. There is little
opportunity for irrigation.
One knowledgeable official emphasizes the need for new
varieties adapted to the wheat-fallow rotation, varieties
which feature efficient use of available moisture. In ad-
dition these vareities must be responsive to such inputs as
fertilizer. New varieties now available tend to respond well
to inputs but lack yield stability, e.g. Bezostaya under
adverse conditions. The land-race varieties, on the other
hand, are quite stable but are also quite unresponsive to
intensive management practices.
Until recently winter wheats were widely grown under
traditional practices. Over the last decade technological
change, in the form of tractors, combines, herbicides and
fertilizer, has entered the production scene. New varieties
have been developed but are not yet widely diffused except
in Thrace, an area which will be considered in detail later.
Tillage practices for managing fallow and wheat are being
developed and promising results are being promoted in a
new production project covering 20 provinces in the winter
wheat region. It is still too early to see the impact of these
efforts to stimulate production. Even so, it is clear that
production increases in the winter wheat area will encounter
significant hurdles.
All of this implies that no easy solutions are at hand for
stimulating rapid increases in wheat production in the major
part of the winter wheat producing regions.
Looking now at the spring wheat areas, these are found
along Turkey's Aegean and Mediterranean coasts. This
zone is characterized by mild winter temperatures and good
rainfall, ranging from 500 to 800 milimeters per year.






Some 15 to 20 percent of the country's wheat area and
on the order of 25 to 30 percent of the production are
found in these coastal regions
With their emphasis on spring bread wheat, the coastal
regions were the natural recipients of new varieties develop-
ed in Mexico and Italy. At about the same time the farmers
of Thrace, with access to higher and more stable rainfall
than the farmers of Anatolia or the Southeast, took up the
improved Russian variety Bezostaya.
In the late 1960's these areas, grouped into four regions
for purposes of the Ministry of Agriculture-Thrace, South
Marmara, Aegean, and Mediterranean regions-became the
focal point for government efforts to stimulate wheat pro-
duction. Starting in 1967 in the spring wheat regions of
South Marmara, Aegean and Mediterranean, efforts were
intense as the government concentrated on the introduction
of new varieties, fertilizers, herbicides, seed treatment,
increased seed production; and launched an extension pro-
gram aimed at growers. The intent of the government was
to attain sharp increases in production through yield in-
creases achieved by changing varieties and practices.



Coastal Spring Wheat Programs

The program in the spring wheat area was initiated in 1966
with the importation of some 60 tons of seed of the Mex-
ican variety Sonora. This seed was divided among 100
farmers and grown by them with the counsel of the Ministry
of Agriculture. Yields of 4 tons per hectare were obtain-
ed, a marked increase over the 1.5 tons from the local tall
varieties. Encouraged by these results Government decided
to import large quantities of seeds from Mexico, the U.S.
and Russia. In 1967, 22,000 tons of HYV seeds were
imported and distributed. Most of this was in the Mexican
varieties Penjamo, Lerma Rojo, and Super X with some
500 tons of Burt and Brevor from the U.S. and of Bezosta-
ya from Russia.
About 60,000 farmers with 170,000 hectares of land
took part in the 1967 program. Penjamo showed itself
quite well adapted to the Turkish scene and became the
dominant variety among the imports. By 1969, it was
estimated that over 600,000 hectares were sown to the new
improved wheats.
The Ministry of Agriculture relied heavily on its exten-
sion staff for diseminating information of the new wheat
technology. Extension services are essentially in the hands
of General Directorate of Agriculture, with headquarters
in Ankara. The General Directorate has technical agri-
culture directors located in each of the country's 67 pro-
vinces together with experts and supporting staff in various
branches of agriculture.
When the campaign was launched in 1967, few farmers or
extension agents were familiar with the new varieties or the
complementary agronomic practices. At the outset, Go-


vernment assigned 250 extension workers to work full
time for the new program. These agents were informed
through short seminars, informal meetings, and printed
materials of the essential pronts regarding the new varie
ties and were instructed to promote the complete package
to area farmers. The package included preparation of an
appropriate seed bed, seeding technique and timeliness of
seeding, seed treatment, adequate fertilizer applications,
and weed control. All avenues for diffusing information
bulletins were produced and distributed, many of them are
still in print. These indicate expected yield levels between
4 and 6 tons per hectare under recommended practices
and with average weather conditions.
During the first years, with little information yet avail-
able from research in Turkey, it was necessary to estimate
planting dates and fertilizer recommendations on the basis
of experience from other similar environments. These
experiences were buttressed by data from an extensive
FAO program. This program was started in 1961 and
ultimately was responsible for thousands of demonstration
plots throughout the country. In time, as Turkish research
institutions had the opportunity to focus on the problems
of HYV production, a full set of locally-produced recom-
mendations became available.
The Turkish recommendations featured variety and fertil-
izer but included: preparation of a good seed bed; use
of drills for seeding and fertilizer application; applica-
tion of phosphatic fertilizer at seeding time and of nitrogen
in three applications; a seeding density of 80 kilos per
hectare; seeding during the period mid-November to
late December; weed control when needed; irrigation
when needed if available. Seed treatment was also a part of
the recommended package as smut (Ustilago tritici), bunt
and rots can be a problem for Turkish farmers. It was soon
discovered, however, that P6njamo is highly resistant to
smut, reducing the need for seed treatment to protection
against bunts and rots.
For Turkish farmers market risks and uncertainties are
minimized by a price support program of wheat. This
program is managed by TMO (Toprak Mahsulleri Ofisi)
under the direction of the Council of Ministers. Each
year, before the harvest, the Council of Ministers announces
prices for each type of wheat (and for some other cereals as
well). Farmers can supply an unlimited amount of wheat
to TMO at the announced price. The agency has storage
space for about 20 percent of the normal crop and maintains
more than 300 procurement stations scattered throughout
the wheat producing regions. TMO is also responsible to
provide the infrastructure and the funds necessary to absorb
large increases in production, the sort of thing that might
occur from a marked change in technology or from particu-
larly favorable weather. This hasn't been necessary in the
recent past, however, as local demand continues to outrun
production. It should also be noted that an active private
market parallels TMO as local traders, wholesalers, millers,






and speculators compete for annual production.
Farmers also had access to seeds through Government-
managed seed production farms. Certified seed is pro-
duced on large state-owned farms or under contract by
private growers. The seed is then shipped to supply points
for purchase by farmers. Seed is priced at 125 percent of the
TMO guaranteed price of wheat. It can be bought on credit
but only 75 percent of thesvalue of the seed can be loaned
to the farmer. This credit system has suffered from serious
difficulties in the recent past. Witness the year 1969 when
loans by the agricultural banks for commercial seed retail-
ing were 340 million Turkish lira of which 62 million was
in default, 100 million was overdue and only 178 million
were current. One consequence of this was that credit for
seed was notably reduced in the early 1970's and this was
accompanied by a dramatic decline in the sale of seed,
from a high of over 250,000 tons to a low of under
50,000 tons in 1972.
From 1967 on, the Turkish farmers in the coastal areas
had access to HYV's, to some credit, to an active extension
program, to a package of recommendations, and to an
active, guaranteed market for Their product. By 1969
HYV's were estimated to be on over 600,000 hectares for


spring wheats and by 1972 the estimated area was over
900,000 hectares.



Review of Subsequent Chapters

This is a dramatic shift to new varieties, from virtually
nothing in 1966 to over 50 percent of the spring bread
wheat area 7 years later. Still, it was clear that some
areas remained with the old varieties or with the old
practices. The intent of this study, launched in early
1973 and based on data from the 1971-72 crop year, was
to see to what extent a pattern of adoption emerged.
The goals of the study, a description of the study regions,
the sample, and the data are discussed in Chapter II. Chapter
III presents survey data pertaining to the average profitabili-
ty of new varieties and fertilizers, and the risks associated
with these new inputs. The following two chapters discuss
the relationships between adoption of varieties and fertilizer
in one hand and farm characteristics, farmer characteristics,
and government programs on the other. Conclusion and
descriptions of new research are the topics of the final
chapter.


II. FARM AND FARMER CHARACTERISTICS IN THE REGIONS STUDIED


The study of adoption of the elements of new wheat
technology by Turkish farmers focuses on three spring
wheat regions and one winter wheat producing region.
The country's remaining winter wheat producing regions
are not considered because it was known that few farmers
had adopted new varieties by 1972.
Before considering the regions it is convenient to in-
troduce some terms describing wheat varieties. In what
follows HYV's are short-stemmed varieties introduced to
Turkey from Mexico, from Italy, and from Russia. The
major Mexican variety was Penjamo. Italian varieties were
Conte-Marzotta, Mara, and Libellula, and the principal
Russian variety was Bezostaya. "Other" varieties includes
all other varieties, consisting primarily of improved and
local Turkish wheats.


The Regions Studied

For identification, the three spring wheat regions are called
Mediterranean (Region 1), Aegean (Region 2), and South
Marmara (Region 3) while the winter wheat region is
Thrace (Region 4). The regions and the provinces sampled
within each are shown in the Map. These regions and re-
presentative provinces were identified with the assistance
of agricultural scientists familiar with Turkish wheat pro-
duction practices. Regions were defined so that within each
region conditions for wheat production would be essentially
homogeneous, except that hillsides and flat lands within a
region were assumed to have different characteristics. The
analysis, then, recognizes two zones within each of the three
spring wheat areas. Only one zone is recognized within the






winter wheat region, as Thrace is characterized by rolling
land.
Looking now at the regions, the Mediterranean Region
has an annual average precipitation of 934 mm with the
monthly distribution over the 5 years, 1967 to 1971, as
shown in Figure 1 .The average elevation is 20 meters. Wheat
is planted in November and harvested in May. Wheat makes
up some 30 to 60 percent of the average farmers' crop
mixture. Because HYV's can be harvested in May a second
crop often follows wheat. While the area has a well develop-
ed irrigation system, wheat is rarely irrigated as rainfall is
usually quite adequate. Spring frosts rarely occur and
there is little damage attributable to hail or to wind. The
region accounts for roughly 10 percent of Turkey's wheat
area and some 13 percent of the total production. The
provinces chosen for sampling are less representative of
the entire region than is the case for the other three


regions studied. While this restricts generalization of the
data to the eastern part of the region, that part does
account for the bulk of the region's area and production.
The Aegean Region has an annual average precipitation
of 670 mm. The monthly distribution for the 5 years,
1967 to 1972, is as shown in Figure 1. The wheat season
extends normally from the beginning of December to the
middle of June. Somewhat less than 50 percent of the
wheat is produced in the Region's main valleys and the
rest on the adjacent sloping lands and foothills, at eleva-
tions ranging from 20 to 500 meters. The higher eleva-
tions are subject to spring frosts. While the local wheats,
which mature late, are rarely influenced, frost can damage
the earlier maturing spring HYV's. In the valleys wheat
is usually grown in rotations with cotton and pulses while
on the hillsides tobacco and wheat are featured. Wheat
is harvested in late May and early June. As in the Medi-


BLACK SEA


THRACE


IRANEAN


SEA


Counties surveyed


The Thrace, South Marmara, Aegean, and Mediterranean regions of Turkey and the
surveyed counties within each region.


T'P
C)


Ill


MEDITERRANEAN






terranean Region, irrigation is available but little is practi
ced on wheat. The region accounts for some 7 percent
of the wheat area anc 9 percent of total production.
South Marmara Region averages 680 milimeters of rain-
fall annually and elevations of cultivated lands range up to
250 meters. Average monthly distribution of rainfall for
the period 1967 to 1971 is given in Figure 1. Major field
crops in addition to wheat are sugar-beets, tobacco, pulses,
sunflower, and potatoes. Wheat is harvested in July, 20 to
30 days later than in the Aegean Region. South Marmara
has 4.5 percent of Turkey's wheat area and 3 percent its
production.
Unlike the coastal regions, Thrace is a winter wheat






160 MEDITERRANEAN



120 -



80 -



40 -


0 1 1 1 1 1 1 1 1 1 1 1 1


SOUTH MARMARA
120-



80-



40-



SONDJ0FMAMJJ
SONDJFMAMJJA


MONTH


aiea. Annual aveiiig piecipitallon i; 590 millhimleis, distln
buted as shown in Figure 1 (data from 1967 to 19/1).
Wheat is the dominant crop with sunflower following in
importance. Thrace has nearly 6 percent of the country's
wheat area and contributes a bit over 7 percent to total
production.


Sampling Strategy

After identifying essentially representative privinces from
each region, counties were selected at random from each
province. Within each county villages were stratified in






AEGEAN


I TRACE


THRACE


I I I I I I I I I I I
SONDJFMAMJJA


MONTH


Fig. 1. Average monthly millimeters of precipitation by region (1967-71).








Table 1. Characteristics of sampled wheat farmers by region.

Mediterranean Aegean S. Marmara Thrace

Age 52 49 47 56
Family size 7.8 6.2 5.5 6.0
Education a 2.4 2.3 2.1 2.3
Days working off-farm/year 13.5 13.5 9.9 12.5
Plowing with tractors, % 94 45 38 82
No. agr. radio programs/month 6 3 5 7
Participating field days, etc. b % 51 10 39 38
Membership in agric. society, c % 81 55 79 96
Selling wheat, % 82 33 25 79

a/ Range: 1-7 (1 = no education 7 = college education. b/ Field days, lectures,
and demonstrations, c/ Membership in any of several agricultural societies
ranging from co-ops, 4-H clubs, etc.


Table 2. Characteristics of farms of sampled bread wheat
farmers.


Bread wheat Distance
Size Land in Plots Plots on from home
Region (ha) (%) (no.) hillside (%) to fields (km)

Mediterranean 14.6 67 1.9 67 2.6
Aegean 5.8 46 1.6 75 2.3
South Marmara 7.0 48 77 1.8
Thrace 15.2 48 96 2.3

a/Virtually all durum farmers eliminated and those durum plots
included are not considered as part of bread wheats.


Table 3. Percent of fields and percent of area devoted to various
varieties on sampled bread wheat farms, by region.

Mediterranean Aegean S. Marmara Thrace
Fields Area Fields Area Fields Area Fields Area
Bezostaya 5.7 2.4 7.1 20.7 76.0 78.7
Italian a 1.8 1.5 11.4 10.3
Mexican b 93.1 95.4 23.0 32.7 7.9 8.6
HYV's 94.9 96.9 28.7 36.1 26.4 39.6 76.0 78.7
Other c 5.1 3.1 71.3 64.9 73.6 60.4 24.0 27.3

a/Conte x Marzotta, Mara, and Libellula. b/ Penjamo, Lerma Rojo 64,
Super X, Sonora 64 with Penjamo by far the most prominent variety.
c/ Other includes land race varieties and pre-dwarfs improved varieties,
largely Turkish. Local improved account for about one-fourth of other in
Aegean and nearly one-third of other in South Marmara.







lel Ins of topography, weighted in le ms of wheat area, and
two sets of villages were selected at random. Roughly 30
villages from each legion wele selected in this manner.
Within selected villages, farmer households were stratified
into three groups according to farm size and then farmer
households were selected at random. Roughly 200 house-
holds were interviewed from each region. Some of those
farmers interviewed were subsequently eliminated because
they concentrate on durum wheats. If durum occupied
over half of the farmers' wheat area he was eliminated
from the sample. No HYV durums were available to
Turkish farmers in 1971. This left a sample size of 200
farmers in Mediterranean Region, 154 in Aegean, 177 in
South Marmara, and 178 in Thrace. In most cases, farmers
who reported some durums had all of their wheat land in
durums.
Dr. Resat Aktan figures prominently in all of the early
stages of this study, from planning its scope, to formulat-
ing hypotheses, to preparation of the questionnaire, and
to administering the survey. Survey administration was
completely in Dr. Aktan's hands Ennumerators were pro
vided by the Ministry of Agriculture, largely from its
extension staff. Turkey's State Institute of Statistics also
provided counsel on statistical techniques and accomplished
the task of transferring the data to punch cards. It should
also be mentioned that, while none of the data are reported
here, some 500 farmers of Anatolia Region and Southeastern
Region were interviewed. Those data are now under
analysis, with the aim of establishing the characteristics of
current practices.
Several statistical techniques were used in analyzing the
data. These included simple chi-square tests, analysis of
variance, multiple regression, and logit analysis. While a
number of variables were considered (see Foreword for a
description of hypotheses) major attention was focused on
agro-climatic region and farm size.



Regional Characteristics from Survey Data

Data from the survey can be used to augment the cursory
description in the first section of this chapter. Certain
characteristics of farms and farmers are given in Table 1
and Table 2 while data on the use of the HYV's by regions
are in Table 3.
Some comparisons among regions are interesting. Notice
that average age in Aegean and South Marmara is well
below that in Thrace. Family size in Mediterranean is far
larger than in other regions. Little off-farm work is under-
taken. Tractor use is far more frequent in Mediterranean
and Thrace than in Aegean or South Marmara. Notice that
Aegean is well below the other regions on extension-related
variables (radio programs, field days) and on membership in
agricultural societies. Thrace ranks quite high on agricultural
societies. Finally, both Aegean and South Marmara fall


,bIhe' 4. Distribution of sampled farmers by average
quantities of plant nutrients (nitrogen and phosphorus)
applied to bread wheat in 1971 72 production."

Medilterranean Aqgeain S. Miat nar I hracl
(kg/ha) (%) (%) ) ( )

0-30 6 54 38 6
31-60 7 26 20 14
61-90 12 14 23 29
91-120 17 4 16 21
121-150 20 2 3 14
151-180 19 11
over 180 19 1 1 5
Total 100 100 100 100

Mean (kg/ha) 131 34 50 99


a/ Recommended levels are: for HYV's,
120-140 kg/ha.


160-200 kg/ha; forothers,


well below Mediterranean and Thrace in percentage of farm-
ers selling wheat.
Average farm size for sampled farmers is a good bit
smaller in Aegean and South Marmara while percentage in
wheat is far larger in Mediterranean. South Marmara farm-
ers live closer to their plots on the average. Mediterranean
farmers have the lowest percentage of their land on hillsides
and rolling land.
Clearly Aegean and South Marmara lagged behind Medi-
terranean and Thrace in use of HYV's. Notice too that in
Mediterranean and Thrace there is little evidence that size
is related to adoption while in Aegean and South Marmara
there is some evidence of such a relationship as percentage of
area is larger than percentage of fields
Two additional points should be made. If one compares
the regions in terms of "adoption" of HYV's where adoption
is defined as having more than 50 percent of the spring or
winter wheat lands in HYV's, then the rates are 93 percent
for Mediterranean, 29 percent for Aegean, 31 percent for
South Marmara, and 72 percent in Thrace. The second
point relates to durum wheats. While virtually no durum
wheat was produced by the sampled farmers of Me-
diterranean, 30 percent of the plots of the Aegean farmers
interviewed were in durums, 15 percent of those in South
Marmara, and 8 percent of those in Thrace.
Similarly, the farmers of the four regions can be group-
ed according to fertilizer use. The result is seen in Table 4.
As with HYV's, fertilizer is most heavily used in Medi-
terranean and Thrace, falling sharply in Aegean and South
Marmara. It should be noted that fertilizer use on wheat
has increased appreciably since the introduction of the
HYV's. In 1966-67 it was estimated that wheat absorbed







2dr percent of the total plant nutrients used in agriicuiure.
By 1972, wheat was estimated to take up 60 percent of all
nutrients used in agriculture. Based on the survey results
the four regions under study used nearly 30 percent of the


nutrients applied to all wheat. Even so, there is aip.;' '
some way to go before applications reach recomniti:,
levels, which are 160-200 kg/ha for HYV's and 120 140(
kg/ha for other varieties.


III. FARMER INCENTIVES FOR ADOPTION OF RECOMMENDATIONS


This chapter treats results of analyzing the data from
the bread wheat farmers interviewe, ,n conjunction with
the 1971-72 production survey. Anticipating the con-
clusions it will be seen that differences among regions
exercise the most notable influence on the pattern of adop-
tion of HYV's and on fertilizer use
Before examining the data from the survey it is appro-
priate to review the elements of the new technology. These
were: 1) HYV's, 2) increased use of fertilizer, 3) better
seed bed preparation, presumably with tractor-drawn im-
plements, 4) seed treatment, 5) proper seeding dates and
rates, 6) seed and fertilizer applications with drills, 7) arid
weed control. These will be covered in turn below with
emphasis giv'n to use of HYV's and fertilizers.



Measures of Profitability of Recommended Procedures

There is a presumption that by following recommended
procedures f rmers will realize greater per hectare profits
than from following more traditional practices. This pie-
sumption is solidly backed by experimental data from ri;ch
of the regions. There is still the question of how farmers
fare when they incorporate these practices.
An earlier study undertaken by the Ministry of Agricul-
ture' in 1968 showed that a sample of 55 farmers had
average net return'; of 1810 Turkish Lira for HYV's while
a sample of 44 farmers had net returns of 791 TL per hec-
tare from local varieties. This study also showed a consider-
able difference in use of fertilizer and implements between
the two sets of farmers.
To get an impression oi relative yields under more nearly
similar conditions sample farmers were asked about their
yields for the 1971-72 crop year. It was not worthwhile to


group their responses in termn, of all of the elements of the
new technology as, with some 128 simplified combinations
per region, the substantial vai action among farmers permits
or,ly a few observations for each combinations of elements.
It ,, however, meaningful to group responses in terms of
variety planted and fertilizer use. The results are shown
in Table 5. Even so, ii must be remembered that data are
farmers' recollections of yields, hence are subject to a certain
margin of error. This presentation has the significant ad-
vantage of having maintained leitilizer use relatively constant
between the two classes ol seed. Even so, as many other
important elements are left uncontrolled in Table 5 e.g.
proceeding crop ind sec'liin date-the yields are little
more tl,;in imdicivive of the relative profttabihity of the
practices which ,, e being conu.idered. Notice for example,
that larger applications of feiitlizer seem to be warranted
in oilly three of the nine comparable cases-South Marmara
hillsides [or both HYV's and other varieties and Thrace for
other varieties. In the remaining six cases, the yield dif
ferences are snort of the roughly 200 kilos of extra wheat
needed to buy the extra fertilizer, e.g. for Aegean flat lands
under HYV's, the reported difference in yields associated
with higher fertilizer use is only 90 kilos. This is a measure
of the influence of the uncontrolled elements and, perhaps,
of errors in reporting yields.
In any case, before relative profitability can be assessed
some additional adjustments must be made. First .while
relatively few of the sampled farmers reported seflin5
straw (7 percent, 4 percent, 12 percent and 29 percent.
respectively in the foui regions), they do report that a
market exists and many use the straw for maintaining
their own animals. Assume that the relative proportion of
usable dry matter is partitioned 40 percent to grain and
60 percent to straw in rYV's and 32 percent and 68 per-







cent in local varieties and that tve price of straw is 30 per-
cent of the price of grain (sampled farmers repotled prices
varying from 25 to 35 percent) Then the added grain
needed to offset the implied straw foregone from the data
in Table 5 is seen in Table 6.
A second adjustment can be made because, according
to Turkish government estimates, even when adjusting
for fertilizer, HYV's cost roughly 100 kilos of wheat per
hectare more to produce than do local varieties. These are
said to be the additional costs of harvesting and tending
the HYV's. This estimate seems high for several of the
regions where reported yield differences between HYV's
and Other varieties are low e.g. Thrace hillsides.
Adding these costs together, deducting them from the
reported yields of HYV's and then deducting the reported
yields of ordinary varieties leaves the net extra yields
shown in Table 7.
These calculations-based on farmers' reports of yields,
an assumption about grain partition, the assumption that
average fertilizer use within the respective fertilizer groups
is the same for HYV and other gioups-indicate that HYV's,
produce relatively more profits than other varieties in all
but two situations. Those two are South Marmara hillsides
for low fertilizer users and Thrace hillsides for high fertil-
izer users. Again, we must emphasize the operative verb
"indicate" as these two results are certainly counter to
expectations.
One interesting aspect of Table 5 and Table 7 is that
in four of six hillside situations, the HYV's appear to be
more profitable than the other varieties. This runs counter
to the conventional wisdom of the countryside which has
held that HYV's are not useful on hillsides. The latter view
has received some support from the Extension Service which
has not given the same emphasis to HYV's for Aegean and
South Marmara hillsides that it has given to HYV's for
valley farmers.
It can be asked why the yields of Table 5 are so much
below the 4 to 6 tons yields cited by Experiment Stations
and regularly achieved on some farms. In significant
measure, this occurs because farmers have not taken up all
the practices recommended by agricultural scientists. Table
8 is a graphic demonstration of this assertion.
For the Aegean Region perhaps the most notable issues


Table 5. Average yields (kg/ha) of bread wheats reported
by sampled farmers for 1971-72 by variety, fertilizer use,
and agro-climatic zone.

HYV Other
Low High Low High
Region Zone f at. fert. a fert. a fert. a
1. Mediterranean Flat lands h 2518 b 1750C
Hillsides 1556 b 950 C b
2. Aegean Flat lands 2204 2294 1727 b
Hillsides 1448 1533 1057 1134
3. S. Marmara Flat lands 2430 2375 1477 1550
Hillsides 1063 1833 1008 1332
4. Thrace Hillsides 1947 1811 1278 1600

a/ Low fertilizer is 0-60 kg/ha of nutrients in Regions 1, 3, and 4,
0-50 kg/ha in region 2. High fertilizer is 110-150 kgs. in region 1,
60-100 kg/ha in region 2, 80-100 kg/ha in regions 3 and 4. b/ No
data reported. c/ Fewer than 10 observations.


Table 6. Grain equivalent of extra
HYV's as implied by yields of Table 5.


straw produced by


fertility
Region Zone Low High
---- -- ----
Mediterranean
Aegean Flat lands -91
Hillsides -18 -27
South Marmara Flat lands 127 67
Hillsides -137 20
Thrace Hillsides 51 170


Table 7. Net yield advantage (kg/ha) of HYV's over other
varieties after accounting for value of straw and certain
other costs.

fertility
Region Zone Low High
Mediterranean
Aegean Flat lands 286
Hillsides 273 272
South Marmara Flat lands 980 792
Hillsides -182 381
Thrace Hillsides 620 -59


Table 8. Farmer practices by type of wheat and region in 1971-72 production.

Mediterranean Aegean S. Marmara Thrace
HYV Other HYV Other HYV Other HYV Other
Fertilizer (kg/ha 133 62 22 70 42 109 73
Treated seecs (% of fields) 61 21 77 68 98 75 97 84
Drilled wheat (% of fields) 2 0 5 1 4 15 6 2
Recommended date (% of fields) 94 41 63
Weed control (% of fields) 20 10 25 12 87 52 33 32
Drilled fertilizer (% of farms) 2 0 0 0 0 0
Split applications fert. (% of farms) 78 48 92 70 81 59 67 5







are the low plates of fertilizer application and the low per-
centage of farmers planting HYV's at the recommended
time. Weather was also a problem according to reports
from the Aegean Regional Agricultural Research Institute
as some 200 millimeters less than the normal rainfall
was received.
South Marmara is also distinguished by low rates of
fertilizer applications as compared with recommendations
and by a significant portion of farmers who do not seed at
the recommended time.
Thrace shows up quite well except for split applications
of fertilizer. This may well be a less critical problem for the
winter wheats and the rainfall regime of Thrace than in
the spring wheat regions.
While irrigation is potentially available to many of the
farmers cultivating spring wheats, few take advantage of it.
According to the survey, Aegean Region farmers irrigate
less than 10 percent of their spring wheat area, and most
of this is applied to improved varieties. The remaining
regions report less than 1 percent of their spring wheat
under irrigation and Thrace reports roughly 1 percent of
its winter wheat under irrigation.
One final point must be made before proceeding to the
question of risk. In the summer of 1975, Mexican varieties
were selling at a discount as compared with local varieties in
the market towns of the Aegean Region. The discount
was on the order of 10 percent and is attributable to the
red grain of the Mexican varieties. This gives a reddish
tint to locally ground flour while the local amber-colored
varieties produce a whiter flour through local mills. Given
the local preference for white flour the local varieties are
sold in the market towns while the Mexican varieties are
shipped to major mills.
On the other hand, no price differential was indicated
by the survey. An explanation of this apparent contra-
diction is that the harvest of 1972 sold at the government
support price with sufficient grain available that TMO
could maintain the price. By 1975, however, domestic
and world shortages along with Turkey's inflation pushed
market prices above support prices as local preferences
were manifested. At the time of the survey in any case, as
price differentials were small or nonexistent, they had little
effect on adoption of improved varieties.
With respect to profits, then, experimental data, Ministry
survey data, and the 1971-72 production survey data com-
bine to support the argument that HYV's are more profit-
able than other varieties, even when both are managed in
roughly the same way. Notice, too, that the net yield
differences reported in Table 7 represent proportionately
larger increases in net profits per hectare than in net yields
per hectare. For example, using simple averages, the net
yield increase of Table 7 over the other varieties of Table
5 is over 25 percent. Since other costs must be netted
from the base yields, the relative increase in profits is of
a substantially larger magnitude.


Measures of Risks of Recommended Procedures


Looking now at risk, the situation is less clear than for
profits. It appears that the growing conditions in the
Mediterranean Region for HYV spring wheats and in Thrace
for HYV winter wheats are such that there is no more
risk associated with their introduction than with other
varieties. Mediterranean farmers face little danger from
frost and have not experienced major disease problems.
Thrace is blessed with good moisture hence the criticism
applied to Bezostaya-that yields are unstable because of
moisture stress-is not applicable. Farmers of these two
regions respond with high adoption rates and also are
using heavy fertilizer applications.
Circumstances are quite different, however, in the Aegean
and South Marmara areas where late frosts, especially in
the hills, can be devastating. One of the most critical factors
affecting yields of the Mexican varieties is time of seeding.
Bill Wright, working with Turkey's Wheat Research and
Training Center, states that "Perhaps the worst thing you
can do to a short duration type wheat (in Turkey) is to
sow too early". Seeded early and with high temperatures
during early growth, t.ii I nj is restricted and the plant
moves quickly into the reproductive phase. If frost occur-
ing while the plant is flowering, yields can be severely re-
duced. Sown later, this phase occurs after all but the
latest frosts.
On the other hand, late seeding can also be undesirable.
Demirlicamak of the Wheat Research and Training Center
points out that, if seeding is late, rains can delay planting
even more, leading to poor emergence and inadequate stands
if wet conditions persist.
These constraints, especially that arising from frost,
put the Mexican varieties at a relative disadvantage in
South Marmara and Aegean as compared with Italian or
other varieties. These varieties are of longer duration,
hence, even if seeded early, have a high probability of
escaping damage from frost. This is probably one of the
factors explaining the relatively low adoption of Mexican
varieties and the correspondingly higher adoption of Italian
varieties in South Marmara. As Table 3 shows, South
Marmara has more area in Italian than in Mexican varieties.
It, has more area in improved winter wheats than in the
total of Italian and Mexican spring wheats. The heavier
use of winter wheats is a measure of the relatively cooler
temperatures prevailing in South Marmara, especially in
the hills where nearly 90 percent of the Bezostaya is grown.
This last point, it should be noted, is an adverse com-
mentary on the homogeneity of the hillside areas of South
Marmara. If some of the hills of that region support winter
wheats (Bezostaya) while others support spring wheats
(Italian, Mexican, and other), then these two sets of hills
are clearly not part of a homogeneous agro-climatic zone.
Disease is also playing a role in farmers' assessments of
the desirability of adopting new varieties. Some early users






of these varieties, especially in Aegean and South Marmara,
suffered when their HYV's showed far more susceptibility
to Septoria than did local wheats in the attack of 1970.
Penjamo withstood the attack reasonably well. Still, in
the minds of some farmers, the new varieties are suspect
because some of them manifested great susceptibility to
this now-and-again scourge of wheat growers in the Medi-
terranean littoral.
Other diseases, the most important being yellow rust, are
special problems in Aegean and South Marmara. Here,
however, it appears that the Italian varieties and Penjamo
show better resistance than do the local varieties.
Looking now at precipitation, the monthly rainfall
patterns seems to be less favorable to HYV's in Aegean
and South Marmara Regions than in the Mediterranean
Region.
Discussions with scientists2 indicate two different stages
in wheat's growth cycle. At these stages minimum amounts
of moisture are critical. The first critical stage is at crown
root initiation, which starts 3 to 4 weeks after seeding. The
second critical stage is at flowering and early in the seed
filling period. This usually occurs some 130-150 days after
seeding, depending upon the environmental conditions.
The threshold moisture level below which yields are lower
than expectations needed for these two critical stages vary
depending upon several factors. Given the conditions exist-
ing in the coastal regions of Turkey, a rule of thumb that
seems reasonable is that 40-50 mm rainfall either in De-
cember or January or both, (roughly when crown roots de-
velop) and 60-70 mm of rainfall in either March or April
or both (flowering and early filling stage) is usually suf-
ficient for a good harvest. Assuming the lower figures as
the threshold levels, Table 9 snows the probabilities of
having amounts of precipitation more than or equal to
the thresholds. The probabilities are based on monthly
observations of the past 15 years averaged over several
localities.
As is observed from Table 9, rainfall is not likely to be
a limiting factor in the coastal regions at the crown root
development stage. The probabilities of having rainfall


equal or above 40 mm are very high. However, rainfall
seems to be an important factor at the flowering and
early filling stages. Probabilities of having rainfall equal to
or above 60 mm in March and April are highest in the
Mediterranean Region with 63 percent and considerably
lower in Aegean and South Maramra. This is consistent
with the higher adoption rate of HYV's in Mediterranean
and lower adoption rates in Aegean and South Marmara.
Perhaps even more significant is rainfall's influence on
rates of fertilization. With the higher probability of late
rains in the Mediterranean Region, farmers can be more
confident of adequate moisture, hence would tend to be
more disposed to accept the costs of high rates of fertil
izer. This is certainly consistent with the behavior of
Mediterranean farmers (see Table 4). The probabilities do
not properly order the Aegean and South Marmara regions
with respect to fertilizer use but this might be influenced
by the higher temperatures and evaporation rates of the
Aegean. These conditions can impose an even greater
need for moisture in that region than in South Marmara.
With this, the threshold level in South Marmara might
be overstated relative to Aegean. If so, the probability of
achieving threshold levels is understated.
Frost and disease are also playing a role in fertilizer
use. If farmers must be concerned about disease, frost, and
precipitation, whether on local or improved varieties, they
tend to use fewer complementary inputs in production.
These considerations might well explain the markedly lower
fertilizer use found in Aegean and South Marmara (Table 4),
where risks are relatively higher than in Mediterranean.

Summary

While not conclusive, the data presented in this chapter
suggest that HYV's are more profitable than local varieties
under virtually all of the circumstances covered by this
study. From the data in Table 5, fertilizer does not appear
to be profitable in Aegean and South Marmara, where
average rates of application are low, nor in Thrace where
average rates of application are high. Neither HYV's nor


Table 9. Estimated probability that rainfall will exceed specified
threshold levels by region.

Crown root development Flowering and earlyilling HYV
Pr. (precip 40 mm in Pr. (precip. >60 mm in Adoption
Region Dec. and Jan.) March and April) rates
Mediterranean 0.94 0.63 94.9
Aegean 0.93 0.43 28.7
South Marmara 0.93 0.37 26.4

a/Percent of fields under HYVs.






local varieties manifest a profitable response to fertilizers
according to the data of Table 5, certainly an anomalous
result.
The discussion on risk outlined several ways in which
risk might be influencing the adoption of HYV's. it was
seen that both regions with low adoption rates-Aegean
and South Marmara-experience relatively more climatic
risk from disease, frost and rainfall than do the regions
with higher rates of adoption.


Notes



1. Denizli Ilinde Pamuk, Mekiska Bugdayi Cesitleri ve Akbasak
Bugdayinin Mukayes ell Ekonomik Analizi (1967-68), Tarim Bak.
Planlama ve Ekon. Arast. Dairesi Bask. Yayin No. 36, Ankara
1969.

2. Private discussion with Dr. Glenn Anderson and Dr. Sanjaya
Rajaram at CIMMYT, Mexico.


IV. ADOPTION AND CHARACTERISTICS OF THE FARM, FARMER,

AND GOVERNMENT PROGRAMS


A number of variables expected to be related to farmers'
decisions to adopt new technological practices are presented
in the Foreword. Included are characteristics of the farm,
of the farmer, and of government programs related to the
technologies being considered. In this chapter we present
the survey data in a number of ways, showing the
relationships between adoption of new practices and each
of these factors in two-way tables. In the following chapter
we present the results of multivariate analysis of the effect
of the factors on adoption when considered simultaneously.


HYV's and Individual Variables

The new wheat varieties were introduced to Turkish farm-
ers on a large scale in 1967 after extensive trials in 1966.
Most of these HYV's were imported from Mexico along
with Bezostaya from Russia. Accompanying the importa-
tion of the seed, the Ministry of Agriculture launched a
dramatic program to promote use of the varieties and of
complementary agronomic practices. By 1972, Penjamo, a
variety from Mexico, occupied 70 percent of the coastal
area seeded to HYV of bread wheats while Bezostaya ac-
counted for all of the HYV's grown in Thrace (See Table 3).
Farm size and topography. With respect to topography,
the term "flat lands" refers to land actually in valleys or on
flat plains. In forming the two farm-size categories, each
sub-region's farms (e.g. the farms of Aegean hillsides) were
arrayed by size and divided evenly into two groups. The
range of farm sizes and the proportion of farmers falling
into each group is given in Table 10.
Within regions and holding farm size constant, the effects
of topography are consistent in that flat lands always have
12


higher HYV adoption rates than hillsides (Table 11). This
result accords quite well with the discussion. Differences
are small in Mediterranean where overall adoption rates
approach 100 percent and are substantial in Aegean where
disease and frost conspire against HYV spring wheats while
climate is not right for winter wheats. In contrasting small-
er farms with larger farms within region by topography
classes, the effect of farm size is usually positive. There
are two exceptions to this, that in Mediterranean flat lands
is small while that in South Marmara flat lands is notable.
Only one other difference exceeds 20 percentage points,
that between smaller and larger farms in Thrace.
It was hypothesized that flat lands would show larger
adoption rates than hillsides and larger farmers would lead
smaller farmers. In general this is true but differences in
adoption rates related to farm size are small with the single
exception of Thrace. It might well be that larger farmers
adopted first, followed by smaller farmers, in which case
earlier differences would have been larger. By 1972 how-
ever, 5 years after introduction of new varieties, large dif-
ferences awe not evident. It is clear that the impact of
topography exceeds that of farm size.
Age, education, and membership in agricultural societies.
Table 12 relates adoption of HYV's to three characteristics
of farmers. No simple relationship between adoption and
age is evident. Only in the case of Mediterranean Region
is it true that younger farmers lead older farmers in the use
of improved varieties.
Education and adoption of HYV's do show a consistent
relationship across each of the four regions. In each region
the average education of adopters is higher than the average
education of those who use other varieties.
Turkish farmers have access to a wide range of agricultural







Table 10. Range of farm sizes for sampled farmers by
sub-regions and proportion of sampled farmers in each
sub-region.

Size (ha.) proportion (%
Region Zone Smallest Largest Smallest Largest
Mediterranean Hillside 0.1-6 6.1-125 33.5 33.5
Flat lands 0.1-8 8.1-150 16.5 16.5
Aegean Hillside 0.1-3.8 3.9-37 37.5 37.5
Flat lands 0.1-4.5 4.6-22 12.5 12.5
S. Marmara Hillside 0.14.3 4.4-108 38.6 38.6
Flat lands 0.1-3.3 3.4-15 11.4 11.4
Thrace Hillside 0.1-7.9 8-128 50 50


societies. Among these are Chambers of Agriculture, Agri-
cultural Credit Cooperatives, Agricultural Sales Cooperatives,
Village Development Cooperatives, and 4H Clubs, (clubs
emphasizing the teaching of agricultural skills of young
people). For the sampled farmers, coops were the dominant
kind of society with over two thirds of those reporting
some kind of association being members of cooperatives.
As with education, a consistent pattern emerges across all
four regions with adopters of HYV's reporting greater
membership in agricultural societies than do users of other
varieties. Members of such societies, especially members of
cooperatives, are said to have easier access to credit and to
such inputs as improved seeds and fertilizers than has the
farmer who is not affiliated. It is notable that such a large
portion of the sampled farmers in Mediterranean and
Thrace regions report membership in such agricultural
societies.
It should be noted that age and education are negatively
correlated for Turkish farmers. The simple coefficients of
correlation are -0.52, 0.45, -0.43 and 0.61 for Medter-
ranean, Aegean, South Marmara, and Thrace regions res-
pectively.
Risk aversion: Earlier discussion has argued that HYV's
are more sensitive to the vagaries of weather and to some
diseases than are local varieties. For example, late frosts
were described as having more serious consequences for
Mexican HYV's than for local varieties. Mexican HYV's
are also thought to be more susceptible to Septoria than
the Italian HYV's or local varieties. On the other hand, local
varieties are more susceptible to rusts than are the HYV's.
The occurrence of late frost, of Septoria, and or rust is
unpredictable except in probabilistic terms. Thus, in those
regions where late frost or Septoria are thought to be major
problems, farmers planting Mexican HYV's tend to be
exposing themselves to more risks than those planting local
varieties. The converse is true in areas where rusts are
prevalent.
Late frosts will tend to be a greater problem in the
higher hills and in South Marmara followed by Aegean than


Table 11. Adoption of HYV's a among sampled farmers by
size of farm, b region, and topography b (percent).

All Smaller Larger
Region Zone farmers farmers farmers
Mediterranean Flat lands 96 95 97
Hillsides 91 92 90
Aegean Flat lands 69 60 77
Hillsides 14 4 23
South Marmara Flat lands 57 70 43
Hillsides 22 13 32
Thrace Hillsides 70 62 85

a/ Mexican, Italian, and Russian varieties. b/ See table 10 and
discussion.


on the plains or in the Mediterranean region. Even though
the last severe outbreak of Septoria, in 1969, occurred in
Mediterranean region, the agro-climatic characteristics of
South Marmara tend to be most favorable to Septoria's
development. Stripe rust tends to be most notable in
South Marmara and stem rust in Aegean. The Mediterranean
Region suffers little from either.
On the basis of these natural risks and the foregoing
description of the varieties, we would expect Mexican
varieties to have a clear advantage in Mediterranean Region
and a lesser advantage in South Marmara and the high
hills of Aegean because of Septoria and late frost. Italian
varieties would tend to have an advantage in South Marmara
because of Septoria resistance. On the basis of risk
considerations, Mexican HYV's would tend to be preferred
over Italian HYV's in Aegean flat-lands, where frost is not a
factor but stem rust is, with roughly comparable risks on
the higher hills where frost threatens the early maturing
Mexican HYV's more than the later Italian HYV's.







Table 12. Age, education, membership if farm organizations,
and adoption of HYV's by region for sampled farmers.

Average Average Membersb
Region Seed age (years) education a (%)
Mediterranean HYV 51 2.5 97
Other 58 1.9 85
Aegean HYV 50 3.0 40
Other 47 2.2 10
South Marmara HYV 48 2.6 30
Other 47 2.0 15
Thrace HYV 54 2.5 72
Other 53 2.4 69

a/ The averages are based on coded values; e.g. "never went to
school" is coded 1 while "graduate of college" is coded 7. b/ The
partitioned sets are adopters and non-adopters so that, e.g. foi
Thrace, of those using HYV's 72% are members of an agricultural
society.







With respect to Thrace, the Mexican and Italian HYV's
are simply not relevant as winter wheats are required. This
leaves the field there open to Bezostaya. Bezostaya is also
sometimes found in the higher hills of Aegean and South
Marmara where climate calls for a winter wheat. This,
of course, is an adverse comment on the homogeneity of
the agro-climatic regions as they were drawn.
In an effort to relate risk, farmer's perceptions of risk,
and the adoption of HYV's, an index of optimism was
constructed for each farmer. The index is the ratio of
bad and normal yields divided by the relative probability
which the farmer assigns to bad and normal weather.'
The range of the index is from zero to infinity with
zero occurring when yields in a bad year are held to equal
to zero or when the probability assigned to normal weather
is zero. As the relative yields in bad years approach 1
or as the subjective probability of the occurrence of a bad
year declines the value of the index rises.
Only normal and bad years are considered in the index
because of the assumption that income-seeking risk-avert-
Ing farmers worry most about the consequences of bad
weather. This is roughly in line with the treatments of
Roy2, Telser3, and Kataoka4 where it is argued that the
probability of getting income below some critical level
plays an important role in decision making.
To quantify farmers' perception about nature, sample
farmers were asked to guess yield levels obtainable in their
best fields under bad, normal and good weather conditions.
They were also asked to tell how many bad, normal, and
good years they expected in the next 10 years. Nearly
all of the farmers responded to the question on yield
levels. Only some 10 percent responded to the question
on the fequency distribution of bad, normal and good
years. For those farmers not responding to this question,
it was arbitrarily assumed that each outcome has equal
probability, i.e. the Principal of Insufficient Reason was
invoked. For farmers not responding to the question on
weather, then, the optimism index is simply the ratio of
yields in years of bad weather to yields in years of good
weather.
Farmers responding to the question on the relative fre-
quency of various kinds of weather were grouped according
to the class of wheat seed they used. The results are
shown in Table 13. For each region save the Mediterranean,






Table 13. Relative frequency of bad weather as related
to use of HYV's and local varieties by region.

Mediterranean Aegean South Marmara Thrace
HYV's 0.24 0.20 0.12 0.23
Others a 0.31 0.32 0.37

a/ No observations.


adopters of HYV's expected fewer years of bad weather
than did users of local varieties. No users of local varieties
responded to the question in the Mediterranean. Only in
the case of South Marmara is the number of respondents
sufficiently large to permit placing much faith in the frequen-
cies reported. It should be pointed out that interesting
comparisons are those between HYV and Others within
regions. Comparison between regions has little meaning.
A more interesting comparison, of course, would have been
between adopters and others within topography classes but
the limited number of observations makes these comparisons
dubious.
Indices for all farmers were arrayed by region and a
series of descriptive statistics were calculated. No pattern
emerged from comparisons of the several measures calculat-
ed. In terms of means, only that for Thrace differs from
the other three. Even here, while adoption rates in South
Marmara are clearly lower than Mediterranean and Thrace,
the average index of optimism is much higher in South
Marmara than in the other two regions.
Extension Variables. When the new wheat campaign was
launched in 1967, few farmers or extension agents in Turkey
knew about HYV's or about the agronomic practices es-
sential for attaining high yields. At the outset, Government
assigned 250 extension workers to the coastal regions for
full time work in the wheat program. International agencies
and experts from foreign universities also helped to diffuse
the new technology.
All in all a substantial effort was made to get information
to the farmers. One measure of the success of this effort
is that virtually all of the farmers sampled for this
study reported knowing about HYV's. Only in the Aegean
Region did those reporting no knowledge of improved
wheat varieties exceed 3 percent. There, some 30 per-
cent of the sample farme'< said they were not aware of
such varieties.
Most of the farmers reported that their first knowledge
of HYV's came from the extension service. In all regions
the proportion reporting fiist knowledge from agricultural
organizations exceeded 80 percent. This result might be
partially attributable to the fact that the enumerators for
this survey were extension agents.
A surprisingly large number of sampled farmers report-
ed extension visits, from 75 percent in Aegean to 92 per-
cent in Thrace. This might be because farmers were report-
ing all kinds of visits by representatives of agencies associat-
ed with agriculture. Many farmers reported participating in
extension activities-from 13 percent in Aegean to 43 per-
cent in Mediterranean. Finally, the reported knowledge of
the technology and of the equipment best suited to the
implementation of the technology varied from 37 percent
in the Aegean to 76 percent in the Mediterranean and up
to approximately 95 percent in the other two regions.
In each of the above cases the Aegean Region lagged
well behind the others, and it also lags in the adoption of
HYV's. But, in each case, South Marmara compared quite







well with Mediterranean and Thrace and it too lags behind
in the adoption of HYV's.
Table 14 relates several dimensions of extension to the
class of seed reported by sampled farmers. While clear cut
patterns do not emerge from the table, adoption of HYV's
usually report closer association with extension activities
than do users of other varieties.
Seed availability and credit. Government influenceson
the production of seed dates to 1963 when a seed produc-
tion and certification law was passed. Responsibility for
insuring quality seed falls to State Farms. These farms
produce seed and also gather the seed of contracted farm-
ers. They inspect, clean, treat, and bag the assembled seed
which is then sent to seed laboratories for testing. Seed
distribution is carried on by several cooperating govern-
mental agencies.
Seed production and sales reached 200,000 tons in 1969.
From 1969 on, production has held up well but sales de-
clined substantially to 39,000 tons in 1971. It is said that
this reduction is largely due to the reduction in the availa-
bility of agricultural credit for financing seed purchases.
There is some evidence that availability of seeds is in-
fluencing adoption of HYV's. For example, in Aegean
Region 58 percent of the farmers sampled expressed this
difficulty and 89 percent of these did not plant HYV's.
For South Marmara the comparable percentages are 56
and 92 respectively. By way of contrast while 25 percent
of Mediterranean farmers reported difficulty in getting
HYV's, only 18 percent of these did not plant HYV's.
A second factor which might limit the use of HYV's
is the distance which farmers must travel in order to ac-
quire the seed. For both Aegean and South Marmara, the
distance to a source of HYV's is lower for adopters than
for others. In the case of Aegean the comparison is 14.6
km. vs. 15.2 km while in South Marmara it is 16.9 km. vs.
28.3 km. Again, however, the pattern is made less mean-
ingful by the average distance reported by adopters in
Thrace, 19.9 km. This is considerably greater than the
distance reported by non-adopters in Aegean.







Table 14. Extension services by region and by class of
seed (percent of farmers).

Extension Field days, Aware of new
Region Seed visits lectures, etc. technology
Mediterranean HYV's 78 44 75
Other 80 27 67
Aegean HYV's 82 26 54
Other 68 7 30
South Marmara HYV's 98 33 100
Other 84 44 91
Thrace HYV's 93 45 96
Other 89 15 98


Among farmers reporting the purchase of HYV's an
overwhelming proportion of them paid with their own cash:
over 65 percent in Mediterranean, over 80 percent in
Aegean and Thrace, and over 70 percent in South Marmara.
These responses may, of course, signal that tight credit is
limiting the expansion in use of HYV's.
Markets. Even given the faithful and sincere efforts of
the government to protect farmers from the vagaries of
adverse price fluctuations, it is sometimes said that not all
farmers have access to the protection offered by the go-
vernment's wheat trading agency, TMO. In particular it is
said that Aegean and South Marmara farmers do not find
these services available and that, because of apprehension
over the marketability of the red grained Mexican wheats,
these farmers were reluctant to adopt improved varieties.
To examine the validity of these statements two market-
ing variables were included in the analysis.
It is instructive to consider the relation between adoption
of HYV's and sales of wheat. This is seen in Table 15. The
percentages clearly present a consistent pattern with high
rates of adoption and sales going together. Even so, of
the farmers selling wheat, 45 percent grew local varieties
in Aegean, 56 percent in South Marmara, and 20 percent
in Thrace.
Data from several Aegean and South Marmara villages
with low adoption rates support some interesting specula-
tion on the influence of markets on the diffusion of HYV's.
In the twelve selected villages with 122 farmers only seven
farmers used HYV's. Forty-six farmers, 38 percent, sold
wheat. All but one of these sold wheat to private dealers.
Of the seven adopters, four sold wheat and one of these
sold to TMO. This all might mean that the emphasis on
private sales led the farmers to prefer local wheats over
HYV's.
While the preceding conclusion is not completely con-
sistent with the absence of price differentials reported by
farmers surveyed, it does fit other data which suggest that
while prices of the two wheats tend to move together, for
some markets TMO does not operate for long periods.
Rather, in some local markets TMO is on hand for only a
few months after harvest.
Data from the survey are not, unfortunately, completely
clear on the effects of markets on adoption of HYV's. More
detailed field work is being carried out to clarify the role of
markets in the diffusion of HYV's.






Table 15. Percentages of sample farmers adopting HYV's
and percentage reporting sales of wheat.

Mediterranean Aegean S. Marmara Thrace
Fields with HYV's 95 28 27 76
Farms selling wheat 82 33 25 78







Table 16. Average fertilizer use on wheat by sampled
farmers cross classified by farm size, region, and topography
(kg N + P205 per hal.

Region Zone Smaller farms Larger farms
Mediterranean Hillsides 125 143
Flat lands 113 122
Aegean Hillsides 26 59
Flat lands 33 57
South Marmara Hillsides 44 60
Flat lands 57 53
Thrace Hillsides 91 106



Table 17. Fertilizer use on wheat by sampled farmers
classified by region, topography, farm size a, and type of
seed (kg N + P205 per ha).

Smaller farms Larger farms
Region Zone HYV Other HYV Other
Mediterranean Hillsides 133 153
Flat lands 114 124
Aegean Hillsides 26 64 30
Flat lands 64 16 60 27
South Marmara Hillsides 69 41 64 42
Flat land 80 37 65 48
Thrace Hillsides 107 67 110 89

a/See Table 10.


Table 18. Fertilizer use and membership in agricultural
organizations by region (kg N + P20s per ha).

Mediterranean Aegean S. Marmara Thrace
Members 151 47 53 102
Non-members 93 22 28 53



Table 19. Distance from farm to source of fertilizer by
region and quantity used (kilometers).

Fertilizer use Mediterranean Aegean S. Marmara Thrace
High 20.3 14.7 12.1 13.9
Low 18.5 18.3 14.9 14.5



Table 20. Percentage of sampled farmers using fertilizers
who expressed difficulty in obtaining fertilizer, by region,
class of seed, and level of fertilizer use.

Region Seed Low Use High Use
Mediterranean HYV 33 55
Other -
Aegean HYV 30 57
Other 41 34
South Marmara HYV 42 50
Other 60 74
Thrace HYV 66 75
Other 83 67


Fertilizer and Individual Varieties

Fertilizer use in wheat has increased dramatically in Turkey.
From application of 3 percent of all wheat in 1963 it
rose to 32 percent of all wheat in 1972. Subsides on
fertilizers were instituted in 1963 and, until 1974, it is
estimated that Turkish farmers paid some 40 percent of the
production cost. Roughly, three kilos of wheat were required
to buy a kilo of nitrogen. Half of the nitrogeneous fertil-
izers were produced domestically, half were imported.
Benefit/cost ratios estimated by State Planning Organiza-
tion were favorable to wheat, on the order of 3, but were
even better for such crops as beets (4), beans (5), rice
(10), and cotton (11).
Farm size, topography, and HYV's. Table 16 shows
average use of nitrogen and phosphorus on wheat by sam-
pled farmers. All farms are cross classified by farm size,
region, and topography. The effect of region and farm
size is quite clear as the farmers of Mediterranean and
Thrace use far more fertilizer than these of the Aegean
or South Marmara regions. With the single exception of
South Marmara flat lands, larger farmers apply more fertil-
izer to wheat than do smaller farmers. Topography has no
consistent affect on fertilizer applications as flat-land farm-
ers sometimes use more fertilizer than do hillside farmers
and sometimes the reverse.
Table 17 shows fertilizer use as influenced by type of
seeds. It is clear from the table that farmers who plant
HYV's, use more fertilizer on the average than do farmers
who plant local varieties. The difference is consistent for
all regions, topographies, and farm size groups. It is un-
doubtedly connected with research results which show that
HYV's are more fertilizer responsive than local varieties.
An unpublished State Planning Organization report shows
benefit/cost ratios of 2.73 for HYV's and 1.93 for local
varieties. It is also linked to credit programs which insist
that recipients of credit for seed use fertilizer at recom-
mended levels. While this stricture is probably not adhered
too, it has undoubtedly had some influence.
The most arresting result from Table 17 is that for
HYV's the relationship between fertilizer use and farm
size, seen in Table 16, is sharply reduced when variety
grown is considered. Much of the difference in fertilizer
use on small and large farms appears to be related to the
more widespread use of HYV's on the larger farms. Only
in Mediterranean Region is the size/fertilizer relationship
maintained. For local varieties, however, the size/fertil-
izer relationship is maintained in each of the five sub-
regions in which appreciable quantities of local wheats are
found.
One general point can be made with respect to fertil-
izer use and this is that only in the Mediterranean area are
farmers approaching recommended levels. Recommenda-
tions are on the order of 180 kilos of nutrients-roughly
120 kilos of nitrogen and 60 kilos of phosphate-ranging
from 160 kilos to 200 kilos. Contrast such recommenda-







tions with the data of the tables which show even the
highest area, larger Mediterranean hillside farmers, at 143
kilos with other regions falling away sharply from that
rate. Or consider the data of Table 4 which suggest that
some 19 percent of Mediterranean farmers, less than 3 pei-
percent of Aegean farmers, less than 4 percent of South
Marmara farmers, and less than 16 percent of Thrace farm-
ers are following recommendations.
What is restraining fertilizers, whether ignorance, faulty
recommendations, availability of fertilizers, or risk aversion,
cannot be said. It's likely that each of these factors is
contributing in greater or lesser degree to the differences
between use and recommendations.
Membership. Age and education as they related to fertil-
izer use were not examined individually. Membership in
agricultural societies was considered and the results are
shown in Table 18. It is clear that members use appreciably
more fertilizers than non-members in each of the four re-
gions. This is consistent with the idea that societies, especial-
ly cooperatives, have preferred access to fertilizers and
that society members have easier access to credit and to
the fertilizer itself than do non-members.
Distance to source. For the most part farmers transport
fertilizers from markets to farms. Only in Aegean Region
is an appreciable portion transported by someone other
than the farmer. There some 24 percent of the sampled
farmers report other means of transportation.
Variation in distances transported vary from region to
region and between those who use heavy and light applica-
tions. This variation is seen in Table 19.
In general differences are small and relate shorter dis-
tances to greater fertilizer use.
Availability and credit. While the use of fertilizer has
increased dramatically since 1963, it is said that many
farmers cannot obtain the fertilizer they would like to
have. Table 20 relates class of seed and relative use of
fertilizers to difficulties in getting fertilizers as reported
by sampled farmers.
While no consistent pattern emerges from the data of
Table 20 it can be said that those with greater use of
fertilizer and those seeding HYV's tend to have greater
proportions reporting difficulties in obtaining fertilizers.
Another classification on this same variable, i.e. farm
size against difficulty in obtaining fertilizer shows that
only in Aegean and South Marmara did small farmers report
appreciably more difficulty in obtaining fertilizers than
large farmers. For Aegean 79 percent of small farmers as
compared with 76 percent of larger farmers reported dif-
ficulty while for South Marmara 63 percent of smaller farm-
ers as compared with 44 percent of larger farmers report-
ed difficulties.
Credit is widely used for purchasing fertilizer according
to sampled farmers. This is seen in Table 21 which
reports source of funds and Table 22 which shows source
of credit for those using credit. The percentages related
only to those who responded to the question (some 28


Table 21. Source of funds for fertilizer purchases among
sampled farmers (percent)

Region Own funllis C( i,(ll (C)monl)IIIaIIloi
Mediterranean 47 3G 1/
Aegean 44 55 1
South Marmara 38 38 25
Thrace 44 45 12


percent of Mediterranean farmers and 33 percent of Aegean
farmers did not respond to the question while virtually
all South Marmara and Thrace farmers responded).
Credit clearly played an important role in financing the
purchase of fertilizers. Judging from the column "Own funds"
there is not much variation among regions except that
South Marmara appears to be somewhat lower than others.
Of more interest is the difference in merchant financing
reflected in the second part of the table. Merchants play
an important role in Mediterranean Region while coopera-
tives dominate financing in the remaining regions.
A cross classification of farmers by size with respect to
fertilizer credit showed that, among farmers using fertilizers,
larger farmers tend to be more likely to use credit for
fertilizers than smaller farmers. The differences are small
in Mediterranean and Thrace, less than 2 percentage points,
and larger in South Marmara, 21 percentage points. In
Aegean Region, the relationship is reversed with 71 percent
of the smaller farmers using credit for fertilizer versus 63
percent for larger farmers.
Extension service. Several classes of extension services
were included in the study. These and their relationship to
fertilizer use are included in Table 23.
A greater proportion of those using more fertilizers
reported visits from extension personnel than did those ap-
plying less fertilizer. Roughly the same pattern holds for
participation in field days, lectures, and etc. as well as among
those who were aware of the technology. In general, then, a
greater exposure to extension services is related to great-
er'fertilizer use.
Two aspects of the table are of interest. One is the
relative awareness of new technology by South Marmara
farmers. The data suggest they are more aware than
farmers of other regions but their adoption rates, for vir-




Table 22. Source of credit for fertilizer purchases among
sampled farmers using credit (percent).

Agricultural
Region bank Cooperatives Merchants Other
Mediterranean 17 39 38 6
Aegean 11 89 0
South Marmara 11 72 1 16
Thrace 17 80 1 2







Table 23. Participation in various classes of extension
service by region and fertilizer use (percent).

Partic- Aware of
Fertilizer Extension ipated field new tech-
Region use visits days, etc. nology
Mediterranean High 92 53 79
Low 69 36 72
Aegean High 67 11 52
Low 82 8 26
South Marmara High 95 56 99
Low 80 26 88
Thrace High 93 35 96
Low 92 40 96


tually all elements, are notably lower than those in Medi-
terranean and Thrace.
The second is the relatively low rates of exposure
evidenced in Aegean Region. This conforms with their
low rates of adoption of HYV's and fertilizers.
An alternative grouping vis a vis extension services is
shown in Table 24. Here, exposure to extension services
is cross classified by region and farm size.
As expected larger farmers fare better than smaller
farmers. Again Aegean farmers, with low rates of fertilizer
use and adoption of HYV's have lower percentages of
participation than do farmers from other regions. South
Marmara farmers, on the other hand, have high rates of
participation but low rates of adoption. Mediterranean
farmers, with the highest rates of adoption, tend to lag
behind both South Marmara and Thrace farmers.


Table 25. Farmers reporting certain practices classified
by agro-climatic region (percent of fields).


Table 24. Participation in various classes of extension
services by region and farm size (percentages of farmers).

Region Farm size Extension visits Field days
Mediterranean Smaller 71 33
Larger 86 54
Aegean Smaller 65 11
Larger 81 15
South Marmara Smaller 85 40
Larger 91 43
Thrace Smaller 92 33
Larger 93 41


Other Elements of the Technology.

Looking now at other elements of the technology, data
from sampled farmers have been cross classified with use
of HYV's by region. The results are seen in Table 25.
Seeds distributed by state farms are virtually always
treated. Seeds can also be treated by farmers with the
assistance of cooperatives or other agricultural agencies.
In general, and as might be expected because government
agencies distribute no local varieties of spring wheat, more
HYV users report seed treatment than do growers of local
wheats. The proportions are quite high for South Marmara
and Thrace HYV growers, notably low for Mediterranean
local variety growers.
Surprisingly few farmers drill seeds. Only in Aegean
Region is there any appreciable use of seed drills. In


by class of seed and


Treated Drilled Seeded Weed Fertilizer Fertilizer
Region Seed seeds seeds Nov-Deca control hand-cast split dose
Mediterranean HYV's 61 2 94 20 98 78
Other 21 0 10 10 100 48
Aegean HYV's 77 5 37 25 93 67
Other 68 1 30 12 86 5
South Marmara HYV's 98 4 59 87 100 81
Other 75 15 22 52 100 59
Thrace HYV's 97 6 b 33 100 92
Other 84 2 b 32 100 70

a/ Recommended date for varieties from Mexico. b/ Winter wheats seeded earlier in
the fall.







general more HYV users avail themselves of drills but the
differences HYV veisus users of local vai ictes-while con
sistent, are negligible.
Looking now at seeding date, Mediteilanean HYV users
are complying with the recommended seeding dates. Aegean
HYV users do not seem to be doing so but there are ex-
tenuating circumstances. First, some 20 percent of the
HYV's are Bezostaya and this has a different seeding date.
Second, some 25 percent of those using varieties from
Mexico are sowing in the spring. Thus, if all remaining
HYV users were seeding at the recommended time, the
percentage should be 60 percent rather than 37 percent.
Why nearly half of the winter sown Mexican HYV's are
not seeded at the recommended time is not evident from
the data. It is likely that yields are reduced but it might
also be true that risks of frost or excessive moisture at
planting time is also reduced.
For South Marmara 70 percent of the HYV's are from
Italy and Russia. Thus it can be inferred that all of the
Mexican plus a good part of the Italian varieties are being
sown in November and December. It should be pointed
out that most Italian and local varieties are photoperiod
sensitive hence the date of seeding is not so cr tical as with
Mexican varieties.
HYV users consistently report more weed control than
do users of other varieties, but, except for South Marmara,
the differences are not large. The quite substantial differ-
ence in weed control seen in South Marmara is inot readily
explainable.


As with seed, most fctlilien is hand caisl. Whail is lallhr
sui pi sing is that hand cast Ig( shows tp I l;a lively Iess ,11irr,110
Aeg(ean farimris. (Given the' tis of sl srd lill, wpoile'(I'd i
South Maimaia, one would haIUv expecli'd S 1outh Min1r,11,1
to have shown less hand casting of fei tilhre tlhan A(PI;I'in.
HYV users are moie consistent in reporting split ap
plications of fertilizers than are users of other varieties.
Moreover, differences tend to be large. This is undoubtedly
related to the larger applications of fertilizers reported by
HYV users.
All in all, HYV users tend to be more likely to apply
complementary recommended practices than are those who
seed other varieties. Except for split applications of fertil-
izer and weed control, however, the differences are not nota-
ble.





Notes

I. The weather risk index for a (liven fairnei is defined as
/ (Yb/YfIPn)I,/Pn) where Yb and Y,, are the yields (in klq/decare)
the farmer expects in had and normal conditions, and Pb and Pn
are the probabilities the farmer assigns to bad and normal conditions.
2. A.D. Roy. 1952. "Safety first and the holding of assets," Econo-
metrica, 29:431-438.
3. L.G. Telser. 1955-56. "Safety first and hedging," Review of
Economic studies, 23:1-16.
4. S. Kataoka. 1963. "A stockastic programming model," Econo-
metrica, 31:181-196.


V. MULTIVARIATE ANALYSIS OF FACTORS AFFECTING ADOPTION


In the previous chapter, we examined the relationship
between adoption of HYV's, fertilizer use, and individual
variables through the use of two-way tables. For several
reasons, it is useful to examine the relationships using
multivariate analysis. First, we can consider variables
not easily considered in two-way tables. Second, multivariate
analysis allows us to estimate the effect of any one variable
while holding others constant. Finally, we are able to
compare more directly the effects of each of the variables
being considered.
Two types of analytical techniques, ordinary least
squares regression analysis and logit analysis, are employed


in relating several independent variables to the adoption of
HYV's and to the application of fertilizers. The statistical
properties of ordinary least squares are reasonably satisfac-
tory for analysis of the amount of fertilizer applied and
this technique was applied to that relationship. Ordinary
least squares can, however, be quite unsatisfactory for the
analysis of a dichotomous variable such as the adoption of
HYV's.
Nerlove and Press' discuss the problems involved in
applying ordinary least squares to relationships featuring
a dichotomous dependent variable. They go on to offer
logit analysis as an alternate technique. In this study both







Table 26. Description of variables used in regression and
logit analysis of adoption decisions.


Independent variables
Age: Farmer's age; in years.
Education: Farmer's education; codes from 1 (no formal
education) to 7 (college).
Family size: Number of family members dependent on farm
Membership: Membership in agricultural society; 1 if member,
0 if not.
Radio: Number of times per month the farmer listens
to agricultural radio programs.
Off-farm work: Off-farm work by farmer; number of days per
year.
Other income: Off-farm income by members other than farm-
er; 1 = yes, 0 = no.
Weather risk: An index of farmer's assessment of weather risk
(see text), smaller values representing more
risky assessment. (Index values range from
0.07 to 6.0).
Farm size: Total farm size; in hectares.
pct wheat: Percent of farm allocated to wheat
Field distance: Distance between home and fields, in kilometers.
Owner: 1 = Land was owned, 0 = not owned.
Tractor: 1 = Tractor is used, 0 = not used.
Valley: 1 = Valley farm, 0 = foothill or hillside farm.
Sells wheat: 1 = Farmer sells wheat, 0 = does not sell any
wheat.
Government: 1 = Farmer sells wheat to government, 0 = does
not sell to government.
Fert. Avail.: Availability of fertilizer; 1 = is not difficult; 0 =
is difficult.
Seed avail. Availability of seed; 1 = was easy to obtain, 0 =
was not.
Extension: Farmer participation in field days, lectures,
demonstrations; 1 = yes, 0 = no.
Dependent variables
HYV adoption: Adoption of high-yielding variety; 1 = yes,
0 = no.
Fertilizer: Nutrients applied (N + P20s); kg/ha.


techniques were applied to the HYV adoption data.
In presenting results, logit coefficients have been convert-
ed to direct probability estimates so as to be comparable
to the ordinary least squares estimates. In each case, the
coefficients presented are to be interpreted as the change in
the probability that a farmer will adopt HYV's as a result
of a one unit change in the independent variable.
We assume that the logit coefficients are the better of
the two sets of estimates. Only in the case of South
Marmara are notable differences evident. There it appears
that ordinary least squares underestimates the impact of
several of the independent variables.
Several variables in addition to those discussed in the
previous chapter were included in the following analyses.
A description of all variables is given in Table 26.


Factors Affecting Adoption of HYV's.


We first present the general nature of the results of the
analysis of adoption decisions, then procede to discuss
them in more detail. We shall speak of a variable as having
an important effect if it is estimated with reasonable
accuracy and if its effect on the probability of adoption is
notable. Reasonable accuracy in estimation is defined here
as a logit t-ratio of at least 1.0 in absolute value. A notable
effect is defined as an increase of 10 points or more in the
probability of adoption when (a) a continuous independent
variable changes from its value at the 15th percentile to its
value at the 85th percentile in its sample distribution, or
(b) a dichotomous variable is increased from a value of zero
to a value of 1.0. Since about 95 percent of the farmers in
the Mediterranean Region were using HYV's, that region
was not included in the analysis. The regression and. logit
coefficients for the three remaining regions are shown in
Table 27. Discussion is based on logit results.
Of the farmer characteristics considered, for important
variables, education and the index of weather risk have the
greatest effects on the adoption decision, changing the
probability of adoption by 10 to 20 percent and 18 to 28
percent, respectively. The size of the family is consistently
positive but in only one case, that of S. Marmara, is it
important with an increase of five persons increasing the
probability of adoption by about 10 percent. Of the farm
characteristics, topography was the most influential, with
valley farms more likely to adopt by 35 percent in Aegean
and Thrace. Farm ers who sell wheat were from 9 to 18
percent more likely to adopt than were those who did not.
Of the three government policy variables considered, seed
availability is important in all regions with an effect of 32
to 51 points. Sales to government had a notable effect,
ranging from 17 to 21 points, and with t-ratios just at or
above 1.0 in all cases.
Farmer characteristics. As was suggested by the analysis
of the previous chapter, education is an important factor in
determining whether or not a farmer adopts HYV's. Each
unit increase in education increases the probability of
adoption by from 3 to 5 percent. This is presumably a
result of an increasing knowledgeability of the value of
HYV's, and an increasing ability to make use of this know-
ledge. Family size also had a positive effect on the proba-
bility of adoption, with each additional person increasing
the probability of adoption by 1 to 2 percent. This was
not anticipated, and we can present no rationale which
oould explain this positive (though small) relationship.
Farmer perception of risk, as reflected in the weather risk
index, had the anticipated effect on adoption in S. Marmara
and Thrace, but not in Aegean. Given the range of indexes
found among farmers, differences in risk perception (two
standard deviations) affect the probability of adoption by
28 percent in S. Marmara and by 18 percent in Thrace.
An inexplicable negative relationship was found in Aegean,
however, indicating that the more risky the farmer perceived








Table 27. Regression and logit analyses of the adoption of high-yielding
varieties. 3


Variable

Constant
Age
coefficient
t-ratio
Education
coefficient
t-ratio
Family size
coefficient
t-ratio
Membership
coefficient
t-ratio
Radio
coefficient
t-ratio
Off-farm work
coefficient
t-ratio
Other income
coefficient
t-ratio
Weather risk
coefficient
t-ratio
Farm size
coefficient
t-ratio
Percent wheat
coefficient
t-ratio
Owner
coefficient
t-ratio
Tractor
coefficient
t-ratio
Valley
coefficient
t-ratio
Sells wheat
coefficient
t-ratio
Government
coefficient
t-ratio
Seed availability
coefficient
t-radio
Extension
coefficient
t-ratio
Number of farms
R2
Chi-square


Aegean
regression logit

-0.17


S. Marmara
regression logit


Thrace
regression logit

-0.54


0.0002 -0.0035 0.0009 0.0025 0.0011 0.0018
-0.1 -0.9 0.4 0.6 0.4 0.6

0.046 0.050 0.049 0.072 0.035 0.029
2.0 1.7 2.4 2.1 1.2 0.9

0.0098 0.015 0.013 0.021 0.010 0.015
1.0 1.1 1.1 1.1 0.8 0.9

0.027 0.045 0.076 0.042 0.053 0.044
0.5 0.6 1.1 0.4 0.4 0.4

-0.0083 -0.0057 0.0089 0.013 0.0034 0.0017
-1.4 -0.7 1.6 1.3 0.6 .3

0.0002 0.0002 -0.0006 -0.0008 0.0000 0.0001
0.5 0.3 -1.2 -0.8 0.2 0.3

-0.053 -0.12 -0.048 -0.059 0.075 0.161
-0.8 -1.2 -0.7 -0.5 0.6 0.9


-0.056 -0.18 0.091 0.14
-0.9 -0.9 3.7 2.4


0.37 0.36
2.0 1.5


0.0090 0.0092 0.0005 0.0019 0.0012 0.0018
1.5 1.4 0.1 -0.3 0.5 0.6

-0.0010 -0.0012 -0.0002 0.0004 0.0028 0.0035
-0.8 -0.7 -0.1 0.2 1.8 2.0

0.12 0.15 -0.05 0.072 0.085 0.10
1.2 0.7 -0.5 0.4 0.7 0.8


0.025 -0.022 0.16 0.28
0.4 0.3 3.0 2.7


0.098 0.090
1.3 1.1


0.34 0.34 0.059 0.045 0.35
5.3 3.7 0.9 0.4 2.3


0.095 0.089 0.069 0.11
1.4 1.0 1.1 1.1


0.10 0.17
0.8 0.9

0.37 0.33
5.4 3.6


0.18 0.18
2.2 2.0


b 0.17 0.19
2.4 2.0


.41 0.51
7.6 4.9


0.11 -0.07
1.2 -1.2
153 176
0.60


0.33 0.32
5.1 4.3


0.15
-1.4
176

86.1


0.15
1.8
178

142.0


a/ Since the regression results are from ordinary least squares analysis of a dichotomous
dependent variable, the t-ratio cannot be used for tests of hypotheses. The logit
coefficients presented are the probability transformation of the logit model coefficients,
i.e. the change in probability of adoption given a one-unit change in the independent
variable. They are therefore directly comparable to the regression coefficients. The







t-ratios for the logit coefficients are those corresponding to the estimated coefficient
of the logit model itself.
b/ In South Marmara, all four farmers who sold grain to the government were also
adopters, and the logit procedure does not permit estimation of a model including
this variable.
c/ In Thrace, all seven valley farmers were also adopters, and the logit procedure does
not permit estimation of a model including this variable.


wheat to be, the more more likely he was to adopt HYV's.
Given that the risk from frost and drought in S. Marmara is
quite similar to that in Aegean, this difference in farmer
behavior is not easily explained.
Farmer membership in agricultural societies was estimat-
ed to increase the probability of adoption by about 4
percent in all three regions, but this is a small effect, not
significantly different from zero. The effects of the other
farmer characteristics, age, radio listening, off-farm work
and other income, were so small or so inconsistently
estimated as to be deemed negligible.
Farm characteristics. Each of the farm characteristics
considered was important in influencing farmer adoption
decisions in at least one of the areas. The most important of
these factors was topography in Aegean and Thrace, with
flat-land farmers about 35 percent more likely to adopt,
other factors constant. The effect of topography appears to
be much less in S. Marmara, and since this was not suggested
by the data of Table 11, this result bears some scrutiny.
The simple correlation between adoption and topography
was about 0.30, which is considerably lower than in the
other two regions, but still substantial. The low estimate
of topographic effect in S. Marmara may be due to the
very high correlation between flat land farmers and high
high weather risk indexes (correlation = 0.4). If most farm-
ers with high (optimistic) weather risk perception live in
the flat lands and if most adopters also live in the flat lands;
it can be very difficult to determine which factor, weather
risk or topography, is most closely related to the decision
to adopt. Since the estimated effect of weather risk in S.
Marmara is the largest of the three regions, and the estimat-
ed effect of flat lands is the smallest, it is quite plausible
that the estimation procedure is overestimating the former
and underestimating the latter in S. Marmara, due to the
correlation between the two.
The second most influential characteristic of the farm
is whether or not wheat is marketed. Those farmers who sell
wheat are estimated to be more likely to adopt HYV's by
9 percent in Aegean, 11 percent in S. Marmara, and 18
percent in Thrace. This suggests that any market discounts
which might exist for HYV's do not adversely affect the
decision to adopt, contrary to the inferences tentatively
drawn in the previous chapter (though we have more to
say in this below).
Farms with tractors appeared to be significantly more
likely to adopt HYV's in S. Marmara and Thrace, though


not in Aegean. The size of the effect in S. Marmara (28
percent) is surprisingly large, particularly when compared
to the estimates for the other two regions. As with the
case of the effect of topography mentioned above, if the
tractor variable is correlated with any other variables, the
estimation procedure may attribute too much effect to the
tractor variable and too little to one of the others. The
other variables which were correlated with tractor use were
wheat sales (0.25), weather risk (0.25), education (0.21),
topography (0.17) and farm size (0.15). Of these, only
the effects of topography and farm size appear to be under-
estimated compared with the other regions, so it is possible
that some of the effect attributed to tractor use in S.
Marmara is actually attributable to valley topography and
larger farm size among the farms using tractors.
The impact of farm ownership was not estimated with
great precision but the effect is to increase the probability
of adoption (by about 10 percent) as expected. The
percentage of cropland devoted to-wheat had no effect on
the adoption decision except in Thrace, where each ad-
ditional percent of wheat increased the probability of
adoption by one-third of a percent. Farm size appeared to
be an important factor only in Aegean, where each addi-
tional hectare of farm size increased the probability of
adoption by nearly 1 percent. While this would have a
considerable impact for extremely large farms of 100
hectares or so, the distribution of farm sizes in Aegean was
such that a farm at the 98th size percentile (about 17 ha)
would be only about 10 percent more likely to adopt than
an average sized farm (about 6 ha).
Government policy. Three factors related to government
policy were considered: whether or not the farmer sold to
the government purchasing agency, whether or not HYV
seed was easy to obtain, and whether or not the farmer had
participated in field days, lectures, or demonstrations. Of
these,seed availability had the most important effect. Those
farmers who said that seed was easy to obtain were 32 to
51 percent more likely to have adopted HYV's than those
who said it was not easy to obtain. This factor appears to
have the greatest impact on farmer decisions of all the
variables considered in this analysis. This result should be
interpreted with some caution, however, since it is in
general true that seeds will be easier to obtain where the
HYV's are well adapted, and more difficult to obtain
where they are unadapted. Thus this variable could be
serving as a proxy for the adaptability of the HYV's. This







could be happening in the S. Marmara estimates, for ex-
ample, where the effect of seed availability is high relative
to other estimates, and the estimated effect of flat land is
relatively low, as previously mentioned.
The effect of sales to the government marketing agency
TMO, was to increase the probability of adoption by about
20 percent in all three regions. This bears out previous
observations that the HYV's are more saleable on the
government market than on the private market. The effect
of extension activities was positive as expected in Aegean
(11 percent) and Thrace (15 percent), but was negative in
S. Marmara. One would not normally expect that participa-
tion in a field day would reduce the probability that a
person would adopt a new variety, yet this conclusion is
supported by the data of Table 14. Again, however, it is
possible that extension activity has been more intensive in
areas where the HYV's were less-well adapted, while other
variables in the analysis (such as topography) have failed
to reflect differences between these areas.
Summary. The results of the multivariate analysis of
HYV adoption decisions have shown that topography,
presumably representing the adaptability of the HYV's,
seed availability and government purchasing activities have
been the most important variables affecting the adoption
decision among the farmers studied. In addition, education,
family size, wheat sales, and land ownership had smaller
but consistently estimated effects in the expected direction.
In two of the areas, the effect of perceived weather risk had
an important effect on the adoption decision, and two
others, participation in extension activities had a signifi-
cant impact in increasing the probability that a farmer
would adopt HYV's. The estimates of the effects of several
of the variables in S. Marmara were inconsistent with esti-
mates for the other areas, perhaps due to correlation among
the variables or to poor specification of the variables in this
area.



Factors affecting application of fertilizers.


Again, the general nature of the results are presented first,
followed by more detailed discussion. A variable is said
to be having an important effect if it is estimated with
reasonable accuracy and if its effect on fertilizer use is
notable. Reasonable accuracy in estimation is defined here
as a t-ratio of at least 1.0 in absolute value. A notable
effect is defined as an increase of 15 kilos of plant nutrients
per hectare when (a) a continuous independent variable
changes from its value at the 15th percentile to its value at
the 85th percentile in its sample distribution, or (b) a
dichotomous variable is increased from a value of zero to
a value of 1.0. Regression results for all four regions
are given in Table 28.
Of the farmer characteristics considered, membership in
agricultural societies is the only variable which consistently
has a t-ratio greater than one. Moreover, in the cases of


Table 28. Regression analyses of the use of fertilizer by


region. a

Variable
Constant
Age
Coefficient
t-ratio
Education
Coefficient
t-ratio
Family size
Coefficient
t-ratio
Membership
Coefficient
t-ratio
Radio
Coefficient
ratio
Off-farm work
Coefficient
t-ratio
Other income
Coefficient
t-ratio
Weather risk
Coefficient
t-ratio
Farm size
Coefficient
t-ratio
Percent wheat
Coefficient
t-ratio
Field dist.
Coefficient
t ratio
Owner
Coefficient
t-ratio
Tractor
Coefficient
t-ratio
Valley
Coefficient
t-ratio
Sells wheat
Coefficient
t-ratio
Fert. Avail.
Coefficient
t-ratio
Extension
Coefficient
t-ratio
HYV adoption
Coefficient
t-ratio


Mediterranean Aegean S. Marmara Thrace

33.0 0.13 30.0 48.0

0.05 0.13 0.30 0.53
0.1 0.6 1.1 1.7

0.09 0.99 3.3 4.8
0.1 0.5 1.4 1.5


1.6 2.3 0.21
1.3 2.5 0.2


7.5 8.6 57.5
1.4 1.1 40


0.83 0.46 0.95
1.1 0.8 1 5


0.06 0.041 0.011 0.07
1.2 1 0 0.2 1.7


1.9 -15.0
0.3 1.0


0.29 0.43 64.0
0.1 0.1 3.2


0.19 0.045
0.9 0.1

0.38 0.06
2.7 0.5

0.06 0.65
0.2 2.6


0.53 3.2
0.1 0.3

39.0 0.50
2.4 0.1

13.0 4.5
- 1.6 0.7

0.50 6.2
0.1 0.9

33.0 11.0
- 4.0 2.0

12.0 4.7
1.5 0.6

15.0 27.0
1.2 3.7
0.43


-0.57
1.9

3.6
0.3

0.99
3.2

0.67
0.1

4.1
0.7

3.3
0.4

24.0
3.3


0.12
0.4

-27.0
- 1.5

0.33
0.8

8.4
0.6

4.1
0.5

32 0
18

23.0
2.7


5.5 12.0
0.8 1.5

26.0 4.5
4.0 0.6

21.0 28.0
2.7 3.3
0.36 0.33


a/ Fertilizer use is expressed in kg of N and P2 0 per ha.







Mediterranean and Thrace regions, where fertilizer use is
greatest, membership is important in the sense described
above. Listening to the radio is important in three regions
but, awkwardly, not in a consistent way. Age has appro-







private t-ratios in three cases but associated quantity changes
are small.
Characteristics of the farm had no consistently important
effects. Distance to the field is important in two cases
"Sells wheat" appears to be important in South Marmara
and in Thrace.
Of the three policy related variables, only fertilizer
availability has appropriate t-ratios in three regions; even
so, the signs are not consistent and in only one case does
the increase in fertilizer use associated with changing the
value of the independent variable from zero to 1.0 exceed
15 kilograms of nutrients per hectare.
Adoption of HYV's is consistently important, even in
the Mediterranean region where some 95 percent of the
farmers had adopted HYV's. It should be acknowledged
that results from models with HYV's are less easy to inter-
pret than were this variable not included because HYV use
itself is held to be a function of the remaining included
independent variables.2
Farmer characteristics. Of the eight variables describing
farmer characteristics, only three ever manifest importance
in the sense described above, viz. membership-in Medi-
terranean and Thrace, radio-in South Marmara and Thrace,
and weather risk-in Mediterranean and Thrace. Each
of the estimated coefficients for membership has the ex-
pected sign, which is to say that membership in a society is
associated with greater use of plant nutrients. The size of
the effect in this analysis is smaller than is implied by the
data of Table 18. In the case of radio, the signs of the
relevant coefficients are negative in two;regions, the op-
posite of what was expected.
For two regions the estimated coefficients for age
had t-ratios in excess of 1.0 and each had the expected
sign. In no case, however, is the impact on quantity of
nutrients a large one.
Off-farm work is the only remaining farmer variable
for which signs tend to be consistent among regions and
with expectations. For three regions, the estimated coef-
ficients have a positive sign.
One variable manifests consistency among regions but
its sign is contrary to what was expected. For all but the
Mediterranean region the sign lor the estimated coefficient
of other income is negative and it was hypothesized to be
positive. A rationable for this result is not evident.
For three variables-education, family size, and weather
risk-the signs of the estimated coefficients are half positive
and half negative. Only weather risk is ever important in
the sense defined, in both cases positively. It might be
argued that the negative sign estimated for education in
Thrace is consistent with expectations, given the data of
Table 5 which suggests that high rates of fertilizer applica-
tion don't pay for farmers there. Table 5, which is based
on farmer's reported yields, is not consistent with the rates
of fertilizer application reported for Thrace in Table 16.
As it seems likely that reported rates of application are
more nearly correct than are reported yields, the data of


Table 16 seem more reliable than those of Table 5. This
conclusion puts in doubt the negative sign for education in
Thrace, seen in Table 28.
Farm characteristics. Six variables describe character-
istics of the farm. Two of these, tractor and topography,
are important in one region and with the expected positive
sign. Distance to fields is important in two regions but
with different signs.
For each of the four regions use of tractor for plowing
and rates of plant nutrients applied are positively related.
Only in one case is the t-ratio greater than 1.0. Topography
also shows consistency among regions but there, for three
regions, the sign of the estimated coefficient is negative,
the reverse of what was expected. This result is consistent
with the data of Table 16. Distance to field is also consist-
ent but is positive in all cases, the reverse of what was
expected.
Tenure status tends to have a consistent sign with
nutrient use positively related to ownership in three of the
four regions. Never, however, is the t-ratio in excess of 0.6
in absolute value.
For each of the remaining variables, farm size and percent
of farm area in wheat, two regions show positive coefficients,
two show negative coefficients. This suggests that the
positive effects of farm size on fertilizer use in Table 16
are due to the other factors which are not controlled there
but are considered explicitly in the regression analysis.
For two of the regions, selling wheat is an important
variable and in both cases the estimated coefficient has
the expected positive sign. For all four regions the sign
of the estimated coefficient is positive.
Government policy. Fertilizer availability, which emerges
from asking the farmer whether or not he found fertilizer
easy to obtain, is positively related to fertilizer application
in three regions. In only one region, Mediterranean, is the
variable important but the sign is negative, indicating
that difficulty in acquiring fertilizer is positively related
to its use.
For the extension variable, three regions show the ex-
pected positive sign. The variable is important in only
one region, South Marmara.
Summary. In a general way the regression analyses of
Chapter V conform quite well with the tables of Chapter IV.
In only two cases, farm size and membership, do coef-
ficients of the regression models depart notably from ex-
pectations based on the simpler tables in Chapter IV.
Only five of the variables are entirely consistent from
region to region and with expectations as regards signs;
membership, field distance, sells wheat, tractor and use of
HYV's. On the other hand, six variables-farm size, radio
education, weather risk, family size, percent in wheat-show
great inconsistency with half being positive and half negative.
Of the remaining seven variables, other income and topogra-
phy show tendencies contrary to those expected.
No region stands out as peculiar, which was the case for
South Marmara in the analysis of HYV adoption. Even in







terms of the number of t-ratios with absolute values great-
er than one, there is startling consistency with no region
having notably more or notably less such estimates.
What is disconcerting in all of this is the size of the
coefficients of determination. Within regions the models
explained only a small part of the variation, from 33 to 43
percent. While such results are not uncommon in work of
this kind, it nonetheless suggests that the prudent will be
cautious in drawing conclusions. In better specified models,
it might well occur that variables with little importance
here would be found important.
Having recognized the constraint imposed on inter-
pretation by the low coefficients of determination, it can
still be said that regional differences in fertilizer use are
notable, tending to following the pattern of the adoption
of HYV's, i.e. ranking regions by percent of land in HYV's
is entirely consistent with ranking them in terms of


average application of plant nutrients. Even within regions,
this relationship holds up. Other variables with a notable
and consistent association with fertilizer use are availability
of fertilizer and membership in agricultural societies. Each
says something about access to fertilizers or about access to
credit or both.





Notes

1. Marc Nerlove, and S. James Press. 1973 Univariate and
Multivareate Log-Linear Logistic Models, The Rand Corporation,
Santa Monica, California.

2. A more desirable econometric approach would be to
estimate the fertilizer decision equation and the hybrid decision
equation together using simultaneous equation techniques.


VI. SUMMARY OF FINDINGS AND POLICY IMPLICATIONS OF THE STUDY


Wheat is the most important single element in the diet of
the Turkish family. Annual per capital consumption varies
between 160 and 225 kilograms of grain. If the present per
capital consumption is to be maintained, wheat production
must increase by 2.6 percent per annum just to equal
population growth rate. Moreover, this will have to be
accomplished through yield increases rather than area
expansion since no unused land is available in the country.
With the object of obtaining higher yields per unit
area, the government of Turkey introduced in 1967 to the
coastal regions new packages of wheat production techniques
consisting of high yielding seed, chemical fertilizer, weed
control, and several other agronomic practices. The new
seeds came from Mexico, Italy, and Russia. They respond
well to fertilizer and other production practices and produce
high yields under favorable climatic conditions.
The purpose of this study was to look closely at the
adoption of the new wheat technology, emphasizing high
yielding seeds and chemical fertilizers. The specific
objectives of the study were to see to what extent farmers
have adopted to the new wheat seeds, applied chemical
fertilizers to wheat and followed other agronomic practices
recommended to them, to identify and quantify association
between adoption of HYV's and fertilizer and selected
factors related to the farmer, the farm and government


policy, and to examine the extension system, credit and
input supply situation, and market conditions as they
relate to the adoption of new wheat technology.
To carry out these objectives, the study draws heavily
on data obtained through regional samples consisting of
800 farms with different characteristics. The regions under
study are: Mediterranean, Aegean, and South Marmara, all
spring wheat regions, and Thrace, a winter wheat region.
In this section we review the findings of this study, point
out policy implications and suggest recommendations for
the attainment of further diffusion of the new wheat
technology and of higher levels of wheat production.



Findings

What follows are general comments. More specific state-
ments are found in Chapter 5.
Between 1967 and 1972, the HYV's covered an area
of about 900,000 hectares in the three coastal spring regions,
Mediterranean, Aegean and South Marmara. This is around
65 percent of the land planted in wheat (including durum
wheat) in these three regions and corresponds to 46 per-
cent of all wheat fields in the regions. In Thrace, the
winter wheat region, HYV's are estimated to have covered







occur. Modifications may involve replacement of the
weaker varieties, and alterations of some of the recom-
mendations to fit farmers' conditions. It is reasonable to
believe that the State's leadership and constant motiva-
tion were the overall pushing mechanism for the successful
diffusion observed in the relative short period of 6 years.
This should continue to be so with a renewed spirit for
the years ahead for further progress.
There are still much potential gains from the new
wheat technology from which the country can benefit.
First, the presently available wheat area in the coastal
regions is more than 1.7 million hectares. Contrasted
to this the HYV's had covered 0.9 million hectares of land.
Therefore, further area expansion of HYV's is feasible.
Second, yields expected from HYV's may run as high as 5 to
6 tons per hectare. Turkish farmers have realized 40 to
60 percent of this potential on the average. With appro-
priate farming practices farmers can also reach 4 to 5 tons
of grain per hectare.
Research must be given priority in the overall efforts
by the state. Research in breeding for new varieties can
be speeded up through provision of more resources and
motivation. Along with this, agronomic research, relatively
neglected in the recent past, should be given more emphasis
than before. While the objectives with breeding should
continue to aim higher at yields with disease resistance,
frost escape and yield stability, the quality factor must
also receive equal recognition. The new seeds must appeal
to farmers' preferences in terms of color, size and bread
making quality. This is especially important for small
farmers who produce for the family's consumption. Further-
more, varieties to fit condition; prevailing at specific loca-
tions are needed. Wide differences in adoption rates of
HYV seeds among regions and subregions clearly indicate
that agro-climatic conditions existing in different locations
are decisive factors along with others on the extent of
adoption and diffusion. Thus it seems an appropriate
strategy in biological research to develop several varieties
for the several different kinds of regions.
To make use of the maximum yield potential from the
new seeds, timely and proper seeding is necessary. In
addition, adequate amounts of fertilizer should be used
along with weed and disease control. Seedbed must be
prepared properly and drill use should be encouraged. All
these practices have, however, their costs and benefits
which must be determined accurately. Agronomic research,
therefore, should aim for recommendations on rates, depth,
and time of seeding and fertilization. Moreover, research
should supply information about the effects on yield levels
of different topographies, number and timing of irrigations,
weed control, drill use for seeding and fertilization.
As the findings suggest, farmer's membership in an agri-
cultural organization is associated with higher rates of adop-


tion. Therefore more and wider membership in agricul-
tural organization, such as agricultural supply cooperatives,
might have a positive impact on adoption of seeds and use
of other inputs.
Extension services can be strengthened. More and
persuasive extension visits, demonstration plots, field days,
lectures can have an important bearing on the rate of
adoption. However, the whole system of extension seems
to be under-utilized. It can be mobilized through the
provision of resources and motivation. It should be empha-
sized also that extension workers should be supplied by
the research workers and scientists with the precise in-
formation about the new wheat technology before they
are sent to the farmers.
The situation in the market can affect adoption. Farm-
ers should be given the feeling that they are guaranteed
easily accessible markets for their wheat. This could be
done by increasing the number of state purchase offices
and by locating them properly in their regions.
The farmer should be supplied constantly and periodical-
ly with certified seeds of reliable quality. Constraints on
the timely distribution of good seeds and seed credit should
be removed. Perhaps one of the alternatives would be to
leave the production of seed to state farms and contracted
farm ers as before, but leaving the seed control and distribu-
tion entirely in the hands of farmers' organizations. In
addition, state farms located in the coastal regions should
be equipped and directed to provide constant flows for
farmers' use.
Any increase in fertilizer use on wheat appears to be
made difficult by economic reasons. With a shortage of
fertilizer supply, farmers as well as the country as a whole
would seek ways to use fertilizer economically. Wheat
then has less chances of receiving fertilizer than cash crops
in coastal regions given 1975 relative prices and productivity
gains. However, as farmers are induced to use the complete
package of the new wheat technology and as output prices
are balanced by the State, then wheat may well become an
effective competitor for fertilizer against other crops. In
any event, the overall supply of fertilizer will have to be in-
creased to reduce the competitive use among several crops.
Again, as in the case with seeds, membership in agricul-
tural organizations would help increase fertilizer use in
wheat. Perhaps efficient and timely distribution of fertilizer
could be maintained by assigning the job of distribution to
farmers' organizations with some control by the state.
Simple calculations indicate that if all traditional varie-
ties are replaced by HYV's and if modest yield level of 2 to
2.5 tons can be reached, then the three coastal regions
alone would supply almost one third of the nation's needs
from the existing amount of land, i.e. without pressure on
lands planted to other crops in the regions. To achieve
those aims will require new motivation and leadership.











COPYRIGHT NOTICE

CIMMYT International Maize and Wheat Improvement Center

The CIMMYT International Maize and Wheat Improvement
Center holds all rights to the source text and shall be
considered the copyright holder for the text and images of
these publications.

The CIMMYT International Maize and Wheat Improvement
Center has made this publication available to the University of
Florida, for purposes of digitization and Internet distribution.

The CIMMYT International Maize and Wheat Improvement
Center reserves all rights to this publication. All uses,
excluding those made under "fair use" provisions of U.S.
Code, Title 17, Section 107 are restricted.

Contact the CIMMYT International Maize and Wheat
Improvement Center for additional information and
permissions.




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