The publications in this collection do
not reflect current scientific knowledge
or recommendations. These texts
represent the historic publishing
record of the Institute for Food and
Agricultural Sciences and should be
used only to trace the historic work of
the Institute and its staff. Current IFAS
research may be found on the
Electronic Data Information Source
site maintained by the Florida
Cooperative Extension Service.
Copyright 2005, Board of Trustees, University
INTERNATIONAL AGRICULTURAL TRADE
AND DEVELOPMENT CENTER
PRODUCTIVITY AND DIVERSIFICATION CHANGES
IN THE AGRICULTURAL SECTOR OF TRINIDAD
AND TOBAGO: SOME EMPIRICAL FINDINGS
AND POLICY ISSUES
Max R. Langham and Carlton G. Davis
IW98-4 June 1998
INTERNATIONAL WORKING PAPER SERIES
' UNIVERSITY OF
Institute of Food and Agricultural Sciences
Food and Resource Economics Department
Gainesville, FL 32611
MISSION AND OBJECTIVE
INTERNATIONAL AGRICULTURAL TRADE
AND DEVELOPMENT CENTER
To enhance understanding of the vital role that international agricultural trade plays
in the economic development of Florida, and to provide an institutional base for
interaction on agricultural trade issues and problems.
The Center's objective is to initiate and enhance teaching, research, and extension
programs focused on international agricultural trade and development issues. It does
1. Serving as a focal point and resource base for research on international
agricultural trade, related development, and policy issues.
2. Coordinating and facilitating formal and informal educational opportunities
for students, faculty, and Floridians in general, on agricultural trade issues
and their implications.
3. Facilitating the dissemination of agricultural trade-related research results and
4. Encouraging interaction between the University community and business and
industry groups, state and federal agencies and policy makers, and other trade
centers in the examination and discussion of agricultural trade policy
Productivity and Diversification Changes in the Agricultural Sector of
Trinidad and Tobago: Some Empirical Findings and Policy Issues
Max R. Langham and Carlton G. Davis
Abstract: Increases in agricultural productivity in the agricultural sector are fundamental
to the improved well-being of both Trinidad and Tobago's agricultural producers and its
consumers to the extent that they depend on domestic food supplies. Estimates of
productivity growth in the agricultural sector indicate that productivity in the sector has
been declining at a rate of-0.41 percent over the period 1963 through 1994. If accurate
this estimate means that Trinidad is getting only 88 percent as much out of the same level
of inputs in 1994 as it obtained in 1963. The sector is clearly going in the wrong direction
if agricultural production is to be a part of Trinidad's future. This estimate also suggest
that the sector is not using its environmental resources in a sustainable way.
Our preliminary results suggest that there has been a fundamental problem that has
two faces. The problem is a lack of sufficient public investment in human capital to serve
the sector. This lack shows up in the general and continuing education of farm decision
makers, in the support of agricultural research activities, and in the collection and
dissemination of data needed to inform both agricultural producers and policy makers. A
small-island economy cannot afford to do the more basic research but it must invest in
good scientists who strive to keep up with global agricultural research and to adapt
promising new knowledge for use by local producers. Rather than invest in human capital,
public revenues have been used in ways that seem to subsidize sectoral inefficiencies such
as ill-advised policies to diversify and manage land use for agricultural production without
the necessary complementary macro policies.
Key Words: Trinidad's Agriculture, productivity, diversification, scale, human capital,
openness, sustainability, competitiveness, Dutch disease syndrome, agricultural policy,
Productivity and Diversification Changes in the Agricultural Sector of
Trinidad and Tobago: Some Empirical Findings and Policy Issues
Max R. Langham and Carlton G. Davis*
Food and Resource Economics Department, Institute of Food and Agricultural Sciences, University of Florida,
Gainesville FL 32611-0240 USA
A Small Island Developing State (SIDS) such as Trinidad and Tobago faces a unique
combination of sustainable development challenges. The topic of this paper is a central component
of these challenges. The numbers of people are increasing and most want to increase their
economic and social well-being in ways that will not handicap future generations in their struggles
for sustainable development. Conway (1991) recognized these challenges in his discussion of
stability, resiliency and equitability as important properties for sustainable agrosystems. His
stability refers to constancy of productivity gains, resiliency to the ability of natural systems to
withstand and recover from the shocks of natural disasters which will continue to occur from time
to time, and his use of equitability was a recognition that gains in economic well-being must
improve the lives of a large majority of the people if such gains are to strengthen the social order.
One of the first challenges to attaining greater stability in economic development is to
understand productivity and the policy instruments which affect it. This challenge in turn begins
with measuring and tracking productivity to use in analyses to better inform policy. At the same
time, the global imperatives of trade liberalization and international competitiveness have added
*Paper presented at the UWI 50th Anniversary Conference on Agriculture in the Caribbean: Issues and Challenges, St.
Augustine, Trinidad, West Indies, August 16-23, 1998. The work on which this paper is based contributes to a
research project being conducted under a Memorandum of Understanding (MOU) among the Food and Resource
Economics Department, University of Florida (FRED/UF), the Department of Agricultural Economics and
Extension of the University of the West (DAEE/UWI) and the Caribbean Agricultural Research and Development
Institute (CARDI). This MOU is a part of a larger Cooperative Agreement (CA) among UF, UWI, and CARDI.
This work would not have been possible without the support of the Central Statistical Office of Trinidad and
Tobago. In particular we acknowledge the very helpful assistance of the Director, Margaret Rampersad, Peter
Pariag, Adher Beepath, and Satee Boodoo.
greater urgency to the need for clarification of critical relationships between agricultural
diversification as a development model and productivity changes. Such clarification is a necessary
condition for sustainable development strategies for Caribbean agriculture in the years ahead. This
paper represents a step in this direction. The objective of the paper is to present some policy
implications based on estimates of agricultural productivity and diversification and on the
preliminary analysis of factors affecting these important variables in the agricultural sector of
Trinidad and Tobago.
The paper is organized into four sections. First, we discuss the concepts of productivity
and diversification and develop working definitions of both concepts as used in the paper. Section
II is devoted to discussing the data and methods used. Section III is used to present measures of
productivity and diversification for the period 1963-94. Also, preliminary results are presented
from analyses of causative and associative factors believed important in affecting changes in
productivity. Section IV is devoted to policy implications of the findings with regard to the role of
the agricultural sector in the longer-term development of Trinidad and Tobago.
Productivity and Diversification1
Productivity: Concepts and Issues
The concept of productivity is often confused with production. Production is an output
measure only, and in a total sense, it is a measure of the output of goods and services coming
from the economic processes of a particular system. For the agricultural sector, it is a measure of
the total output of economic goods and services from sectoral processes. In contrast, productivity
is a measure of output(s) per unit of a measure of the input(s) used to produce the output(s).
Productivity is always associated with some input(s). In economic jargon an input is also referred
1 This section borrows heavily from Langham et al, (1998).
to as a factor of production. Multifactor productivity (MFP, also called total factor productivity,
(TFP) is a measure of the production of goods and services per unit of a measure of the inputs
used in the production processes.
In order to aggregate across different kinds of outputs and inputs it is necessary to use
index measures. IfYt is an index measure of the total output of goods and services (production)
coming from the agricultural sector in year t, and Xt is an index measure of the total inputs
(factors) used in year t, then MFPt = Yt / Xt.
Productivity increases are central to economic advancement because they conserve
resources by giving more outputs per unit of the inputs used, and hence for a given output, they
save productive resources. These savings can be in human, capital, and/or natural resources.
Getting more output per unit of input is just another way of saying that costs per unit of output
are lower. In a competitive system, lower marginal costs lead to increased supplies and lower
prices. A desirable consequence of lower prices is that consumers can purchase more goods and
services with a given level of money income, and hence their money goes further, or as we
economist say, they experience real income increases. Such increases constitute an important link
between productivity and a country's standard of living. In general, in countries where workers
can produce a larger quantity of goods and services per unit of time, most people enjoy a higher
standard of living. It is in this regard that productivity is a key determinant in the living standards
attained by a country.
Since agriculture is chiefly comprised of land-based activities, MFP advances
conserves our natural resource base by permitting us to produce more food per hectare of land.
This conservation of natural resources is central to the issue of sustainability that has become an
important topic in international trade and development. There are two important and positive side
effects of this conservation of resources brought about by productivity increases. First, inflation
rates of prices are moderated. Secondly, the economy or sector of the economy which
experiences the productivity gain is better able to compete.
Getting more for less is also central to the issue of 'competitiveness'. This term can mean
many things. We use the term as "The degree to which a nation can, under free and fair market
conditions, produce goods and services that meet the test of international markets, while
simultaneously maintaining and expanding the real incomes of its citizens" (US Congress, 1985,
p.70). An operational way to define increased competitiveness of a sector or an economy is by
saying its MFP is growing faster than that of its competitors (Ezeala-Harrison, 1995). Different
rates of productivity growth among sectors of an economy lead to changes in the relative
importance of sectoral contributions to a country's output. Also, international differences in
productivity growth rates lead to differences in the competitive position of countries, to country
differences in real per capital incomes, and to changes in the relative purchasing power of
Two other important effects of increased agricultural productivity are that they permit us
to produce the food we need on a smaller land base and they help to keep food prices low. The
first of these permit the more marginal farming lands to be used in more environmentally friendly
ways and the second keeps food prices lower for consumers. Since latest projections indicate that
there will be more people living in poverty in the cities than in rural areas shortly into the new
millennium, affordable food for the poor in an urban setting is becoming increasingly important.
These two effects serve to help offset the problem resource-poor farmers have of surviving in
agriculture when new technical knowledge is used by farmers with whom they compete.
Productivity increases and associated benefits have significant costs in the form of
intangible investments to improve the economic performance of the human and non-human factors
of production. The human resources are of particular importance and require investments in
education and training, research, health, safety, and mobility. Technical knowledge must be
discovered, disseminated, and ultimately embedded in the stock of knowledge of those putting it
in practice, before it can be helpful in increasing productivity.
A significant cost of increases in productivity is the cost of adjustment of resources,
especially those human resources, made relatively less competitive by the processes of change. In
agriculture, increases in productivity have made it difficult for those farmers to survive who
cannot, for some reason, adopt the more productive processes. What often happens in such cases
is that the current generation of farmers do continue farming even with a relative decline in their
level of living unless opportunities exist to supplement their declining farm income. However,
their children tend to break the agricultural bond by seeking full-time employment off the farm. As
a consequence, increases in productivity in agriculture will have much the same effect on farm
population as a decline in MFP in the farm relative to the non-farm sectors--a decline in the
proportion of the population on farms and an increase in the average age of farmers. In the first
instance, people are pushed off the farm by an inability of their families to compete in agriculture.
In the second, they are pulled off by more attractive off-farm employment opportunities
(Langham, 1992, p. 70).
A major problem with the movement of people from the farm to the non-farm sector
associated with increased agricultural productivity is the cost distribution effects of adjustment.
These costs fall disproportionately on the farmers who have a relatively weak resource base from
which to compete. Productivity increases work counter to a policy objective of keeping resource-
poor farmers as primary actors in agricultural production simply because such farmers are at a
competitive disadvantage in capitalizing on new information and technologies. In the US setting,
we have taken great pride in our Land-Grant University System designed to make 'two blades of
grass grow where one grew before' and have claimed to be of great help to farmers. The system
has been of great help to the surviving farmers, but the number surviving continues to decline and
is now considerably less than one percent of the population.
The cost of adjustment in the farm population falls most heavily on those farmers who find
that they can no longer compete in agriculture and are forced to adjust out of farming-- especially
if they can find no comparable wage-earning opportunity for employment. In developing
countries weak intersectoral linkages in the labor markets could limit employment opportunities
outside the rural sector and make the adjustment even more difficult.
Diversification: Concepts and Issues
The concept of diversification has several meanings and interpretations. In this paper the
term is used with reference to the portfolio of commodities grown at the farm level of the
agricultural sector and is measured with an entropy index. The term may be used at any of several
different levels of the economy including the economy-wide level. It can also be used for other
types of portfolios. Davis (1990) for example discusses this concept from the point of view of
form and function as it has evolved in a Caribbean context. Delgado and Siamwalla (1997) make a
distinction between farm-level diversification and village diversification. The former they define as
pertaining to any economic activity undertaken by farm people. And in attempting to characterize
economic development under commercialization, they define village-level diversification as where
households become more specialized and the villages more diversified with a wider array of goods
and services for sale. Timmer (1988) in another aspect close to diversification, speaks of
structural transformation of the rural and urban sectors at the economy wide level.
Another distinction is that between domestic and export diversification. Here domestic
diversification of the agricultural sector connotes a more inward oriented approach toward
attaining food security or an import substitution objective. Export agricultural diversification on
the other hand refers to the diversification of the agricultural export portfolio as a means of
attempting to stabilize foreign exchange receipts. McCalla and Valdes (1997) state that the one
place the government should take a proactive role is with agricultural export commodities. Their
reason for suggesting government intervention in promoting export diversification is that, in the
case of agricultural commodities, unlike industrial goods, the private agricultural sector will
under-invest in the search for new markets for their commodities. The cost of searching for new
international markets is high and the organizational and free-rider problems among farmers
provide justification for government subsidies and/or direct assistance in the search and promotion
Bautista (1992) makes the point that while there appears to be much similarity amongst
issues associated with diversification however defined, there are noticeable differences. Whereas
diversification in general is seen as expanding opportunities for income generation and
employment creation as well as reducing the risk of having all of one's eggs in the same basket,
export diversification out of a few agricultural commodities to non-agricultural exports may be
It is not the purpose of this paper to critique the merits (or demerits) of alternative
conceptual and operational frameworks for viewing the diversification process. Rather, it is to
elaborate on the findings and policy implications flowing from a specified conceptual and
operational framework or point of view.
Diversification is often heralded as a strategy to get the agricultural sector moving and/or
to lower risks. However, too often such a strategy does not have benefit of a complete cost
assessment in part because a research base for policy formulation has not been a strong suit in
developing diversification strategies--perhaps because of the felt need for quick action which does
not permit time for adequate answers to questions. Nevertheless, cost considerations need to be
complete and factored into the strategy.
The work by Habasch (1989) for 42 states in the US and more recent research by the
authors in Florida and Jamaica, by a research team of Florida and Caribbean agricultural
economists (Langham, et al, 1997; Roberts and Langham, 1997; Langham et al, 1998) are
consistent with the principles of specialization and trade based on comparative advantage. These
principles remain a pillar of trade theories and are consistent with the New International Trade
Theories (NITS) which demonstrate that increasing returns to scale, a consequence of
specialization, is a fundamental source of advantage for trade (Helpman and Krugman).
Evidence on the cost side of the diversification issue is of particular importance to the
agricultural sectors of countries that depend heavily on production agriculture and have limited
alternatives. Also, it has implications for research policy and the nature of agricultural systems to
be encouraged through policies. Without adequate human and capital resources to support
research to solve the technical problems up front, it is probably best not to attempt diversification,
or if attempted, the expected benefits should be consistent with the knowledge of the reality of the
economic situation. The reality is that even in the more developed countries there are too few
resources to do this adequately and farmers in general find it easier to make improvements with
what they do best rather than seek further new products to grow.
In putting food on the table, Man has found it increasingly easier in history to do the job
with greater specialization. In fact, agriculture was invented by Man to get away from an
extremely diversified natural system as a source of food. The following quote (the underlined
emphasis is added here) from the US National Academy of Sciences (NAS, p. 1) serves to
emphasize this fact:
Throughout history man has used some 3000 plant species for food; at least 150 of them
have been commercially cultivated to some extent. But over the centuries the tendency has
been to concentrate on fewer and fewer. Today, most of the people in the world are fed by
about 20 crops--cereals such as wheat, rice, maize, millet, and sorghum; root crops such
as potato, sweet potato, and cassava; legumes such as peas, beans, peanuts groundnutss),
and soybeans; and sugar cane, sugar beet, coconuts, and bananas. These plants are the
main bulwark between mankind and starvation. It is a very small bastion.
There are even fewer commodities with large commercial markets. In choosing
commodities for development to provide opportunities for a sizable number of producers, one has
to be very sensitive to this fact. There is a natural tendency in promoting production activities to
over-estimate the size of the market--especially in diversification initiatives where those
responsible are trying to generate some excitement in the effort. Farmers are quite astute in what
they grow vis-a-vis the markets they face. For many minor commodities, only a few producers
are needed to satisfy the demand. Such commodities are not good candidates for programs for
public promotion to aid an ailing agricultural sector. However, as indicated earlier, McCalla and
Valdes see an important role for governments in export crop promotion. This aspect of Caribbean
agricultural diversification strategy needs to be studied further and situated within a sustainable
Important insights from T. W. Schultz, who shared the Nobel Prize in Economics with Sir
Arthur Lewis, in the 1960s are that farmers are rational in their economic decision making, that
the transformation of traditional agricultural systems requires investment in new methods, and that
for savings and investments to occur there must be opportunities for farmers to make profitable
investments and incentives for them to save to do so. "Where are the incentives?"
Anne Krueger raised another very good question when she asked, "Why do we as
economic analysts continue to look to government to correct problems of market failure rather
than to encourage governments to focus on problems which involve the provision of badly needed
social-overhead investments?" A very important social overhead investment is in a strong
agricultural research and education program to provide the information needed for farmers to
make an improved living by increasing there productivity and competitiveness. In doing so they
will produce food efficiently for the benefit of consumers--all of us. However, Stiglitz (1989)
suggested that the answer to Krueger's question might relate to the fact that market failure may
be so pervasive that local overhead investments have little impact.
If one applies the Schultzian result to the diversification issue, one would expect
that traditional systems would have over time exploited opportunities for diversification. Diverse
natural environments lead by necessity to the selection of crops and animal enterprises
accommodated by the environment. However, it is less clear whether traditional systems have
evolved to exploit opportunities for diversification created by price movements in world markets.
Langham's (1992, pp. 87-89) exploration into this issue in a Cameroonian context
supports the hypothesis that indeed they do--often by inter-cropping. On the price side nearly
every situation of negative covariance was exploited in the farmers' portfolio of crops. The major
exception (sisal) was not suitable for the high rainfall area studied. These results are what one
would expect with poor farmers who cannot afford to take risk. They know how to diversify their
portfolio to manage their risk of going hungry. Quiroz (1994) came to essentially the same
conclusion when he stated that his research "suggests that we should not be overly optimistic
regarding the returns to diversification in agriculture." A market phenomenon that limits
diversification opportunities is that commodity prices rise and fall together.
This new evidence adds to the robustness of the Schultzian hypothesis. Poor farmers are
not only technically and allocatively efficient in using the information they have but also they
manage their portfolio of enterprises efficiently. The evidence also suggest that the potential for
gains from diversification have largely been exploited by farmers. What they need is access to new
reliable information and the means (credit, roads, markets, etc.) to exploit it.
Today, it is increasingly recognized that productivity and diversification are simultaneously
determined (McCalla and Valdez ,1997; Langham et al ,1997; Langham et al, 1998). Neither
causes the other. Both are the consequences of other forces within the system. Empirical work at
Florida, cited immediately above, modeled these two variables as endogenously determined in the
Florida and Jamaican contexts, respectively. This line of approach adds further insights into the
productivity/diversification nexus. Diversification introduces greater complexity and more
problems to be solved. Without adequate efforts to anticipate and solve these problems up front
with research and other human capital and infrastructural investments, efforts to diversify will lead
to reduced productivity. In history this seems to have been the reason efforts to diversify have
largely failed and the failure has shown up as a negative association between diversification and
Data and Procedures
The data from which the indexes were computed were recorded from worksheets and
reports made available by the Central Statistical Office of the Government of Trinidad and
Tobago. The data cover a 32 year period from 1963 through 1994. Many other sources of macro
and sectoral data were also used from national and international sources--including some data on
agricultural research assembled earlier by Dr. Ranjit Singh here at the University of the West
Indies and information provided by the Caribbean Agricultural Research and Development
Multifactor Productivity (MFP) Indexes
Multifactor productivity measures estimated for this paper were from Tmrnqvist-Theil (T-
T) indexes2. A T-T index for the rate of change in MFP for a particular country and year is based
on a T-T index of the sum of weighted rates of changes of the quantity of agricultural outputs
minus the sum of the weighted rates of changes in the inputs used to produce the outputs. The
2 The Tornqvist-Theil index has been shown by Diewert to be exact for the homogeneous translog aggregator
function. This aggregator function has been shown by Christensen, Jorgenson and Lau to be a second order
approximation to an arbitrary twice-continuously-differentiable-linear-homogeneous function. Since we can never
know the true nature of the form of the agricultural aggregator function, the approximation provided by translog
aggregator function makes it a good choice.
weights used were revenue shares and cost shares for the rates of changes in the outputs and
One can use the indexes of the rates of changes in outputs and a base year in which the
index is set equal to 1 and compound forward from the base year (or discount backward from it)
to get a quantitative index of the outputs (Y,) for the t th year relative to the base year. In the
same way, one can use the T-T index of rates of changes in the quantity of inputs and the same
base year to estimate a quantitative index of the inputs (X ) used to produce the outputs in the t th
year --again relative to the base year. MFP, in the t th year may be estimated as Y/X, or it may
be estimated using its rates of change and compounding or discounting from the chosen base
year3. The base year 1992 was chosen; however, any choice is arbitrary since any year may be
Entropy indexes were used to measure diversity. This measure was developed in
information theory by Shannon (1948) and has since been used in many areas of science as a
measure of diversity. In this paper, ifp, is the proportion of total revenue from agricultural
commodity i in year t then the entropy index for n commodities for year t would be :
I, = -p,,(On p,i) If the agricultural sector produced only one commodity, the entropy index
would be zero and represent extreme specialization. If the sector were perfectly diversified each
commodity would represent 1/n of the sectors revenue, and the entropy index would take on the
SOne can also measure productivity by years across countries. In this situation, e.g., the comparison could be
made against the country that got the most output per unit of inputs (the frontier country) in that year. For an
example, see the Malmquist indexes estimated by Roberts and Langham for Barbados, Belize, Cuba, Dominican
Republic, Guyana, Jamaica, Suriname, and Trinidad and Tobago for the 1961-86 period.
value of the In (n). With 63 commodities for Trinidad and Tobago the value of the index
representing perfect diversification would be 4.14 = ln(63).
Productivity Trends and the Dutch Disease Syndrome
Output, input and MFP estimates from historical data of the agricultural sector are
presented in Table 1 and graphed in Figure 1. The base year was 1992 for this graph and the
indexes were set at 1 in that year. The most output per unit of input was attained in 1973 and
began falling precipitously in 1977. Growth rates for outputs, inputs, and MFP over two sub-
periods 1963-75 and 1976-94 are given in Table 2 along with an average over all years, 1963-94.
Productivity grew at an annual rate of 1.72 percent over the 1963-75 period. It fell at a rate of-
0.38 percent from 1976-94. For the entire period MFP fell at an average rate of-0.41 percent. A
decline of this amount means that the agricultural sector is losing ground to the extent that a unit
of inputs in 1994 yielded only about 88 percent of what they did in 1963. Had the sector
continued to grow at its 1963-75 rate, a unit of inputs would have yielded over 1.7 times as much
in 1994 as in 1963.
These estimates of negative growth in productivity since 1975 are cause for concern and
research is needed to isolate the causes. One admissible hypothesis is that offered by the "Dutch
Disease" phenomenon where the decline may be attributable to the cost of labor in agriculture
being bid up by developments in the petroleum industry and subsidized export agriculture.
Trinidad and Tobago has a well established track record dating from the 1960s and 1970s as an
"open petroleum based economy". The peculiar economic dynamisms associated with this type of
economic structure were addressed in the writings of Caribbean economic development scholars
Estimates ofMultifactor Productivity (MFP) Indexes
for the Agricultural Sector of Trinidad and Tobago with the Use
of Tnmqvist-Theil Approximations to Divisia Indexes
Year Output. Y Input X MFP = Y/X
1963 0.960 0.878 1.036
1964 0.964 0.910 1.004
1965 1.114 0.969 1.093
1966 1.043 1.024 0.967
1967 1.076 1.090 0.939
1968 1.185 1.082 1.046
1969 1.177 1.177 0.956
1970 1.276 1.168 1.069
1971 1.226 1.113 1.073
1972 1.330 1.075 1.211
1973 1.229 0.976 1.230
1974 1.228 1.036 1.159
1975 1.266 1.019 1.214
1976 1.298 1.052 1.206
1977 1.345 1.017 1.162
1978 1.302 1.077 1.061
1979 1.215 1.134 0.938
1980 1.112 1.041 0.933
1981 1.081 1.220 0.772
1982 1.156 1.003 1.008
1983 1.073 0.956 0.981
1984 1.053 1.011 0.999
1985 0.970 1.030 0.920
1986 0.954 1.041 0.898
1987 0.969 1.045 0.911
1988 0.967 1.068 0.890
1989 0.992 1.115 0.881
1990 1.048 1.200 0.871
1991 1.010 0.993 1.013
1992 1.000 1.000 1.000
1993 1.000 1.006 1.006
1994 1.028 0.911 1.128
Figure 1.-Outputs, Inputs, and Multifactor Productivity in the
AgriculturalSector of Trinidad and Tobago, 1963-94
0.9 -4- Outputs
0.8 -- Inputs
0.7 -- MFP
(V n 1- 0) M 0- cr M U) 1- 0) O
0) 0) 0 ) WMO CMO Ol 2 u)
Growth Rates in Indexes Measuring Outputs, Inputs, Multifactor Productivities
and Diversification in the Agricultural Sector of Trinidad and Tobago by Periods
1963-75. 1976-94 and 1963-94
Variable and Period Estimated Growth Rate Standard Error of Estimate
1963-75 0.02755 0.00415
1976-94 -0.00723 0.00314
1963-94 -0.00371 0.00197
1963-75 0.01014 0.00614
1976-94 -0.00342 0.00296
1963-94 0.00038 0.00151
3. MFP with Actual History:
1963-75 0.01723 0.00501
1976-94 -0.00383 0.00459
1963-94 -0.00409 0.00205
4. MFP with "What If' Estate Sugarcane Was Not Subsidized
1963-75 0.03323 0.00444
1976-94 -0.00089 0.00465
1963-94 0.00217 0.00229
1963-75 0.01081 0.00128
1976-94 -0.00086 0.00128
1963-94 0.00191 0.00068
such as Seers in the 1960s, and Bruce in the 1970s. However, full articulation of the economic
characteristics of the phenomenon within a contemporary global context is generally credited to
Corden and Neary (1982).
Following the economic mechanisms suggested by Corden and Neary, the 'Dutch Disease
Syndrome' of Trinidad and Tobago petrochemical dominated sector would adversely affect its
agricultural sector. During the 1970s Trinidad and Tobago experienced a five-fold increase in its
average crude oil price per barrel. This increase translated into a twenty-fold increase in
government oil revenues. The country continued to receive exceptionally favorable oil prices and
associated government oil revenues up through the mid-1980s (Pantin, 1989). The 'Dutch
Disease Syndrome' manifested itself in the case of Trinidad and Tobago in terms of the profit
squeeze that the petroleum sector exerted on the country's agricultural sector, thus leading to a
general sectoral decline which has not been reversed. Specifically, sectoral decline occurred via
two related economic mechanisms. First, buoyant oil prices and high rates of return on labor and
capital caused a direct bidding away of these and other factors from agriculture. Second,
economic spurts in the petrochemical sector exerted upward pressures on the country's exchange
rate. This in turn lead to further decline in agriculture output, employment, terms of trade and
real return to factors of production.
Our empirical findings on the relationships between Trinidad and Tobago's petrochemical
sector development and agriculture sector MFP/diversification supports the Dutch Disease
hypothesis. Specifically, our estimates, which will be discussed more fully later in this paper,
suggest that the country's oil sector development did have the expected negative total effect on
MFP in the agricultural sector.
The issues of openness and the Dutch Disease Syndrome hypothesis are intimately
connected to the issue of the impact of the country's government policy towards its agriculture
sector, particularly its agricultural export sub-sector. We also did a little "What if?" analysis. In
this analysis the "what if assumptions" were made of no subsidies to estate sugarcane production
and to its labor. These assumptions were reflected in the data by pricing estate sugarcane at the
mill the same as that received by the private growers and pricing estate sugarcane labor the same
as private farm labor. The result are shown in the bottom section of Table 2. This "what if'
analysis indicated that agriculture sector MFP and diversification would exhibit similar trends
"with" and "without" such subsidies but that MFP would have been higher in the absence of the
subsidies. The oil revenues associated with the Dutch Disease Syndrome were the basis of the
direct government revenue transfers to the estate sugar sub-sector. We suggest that these
revenue transfers were not used in an economically efficient manner to enhance the productive
and competitive structure of the country's agriculture sector. We go further to suggest that these
transfers were a major factor associated with stagnant productivity growth in the agricultural
Estimates of partial productivities (PP) for land and labor in the sugarcane sub-sector over
the 1964-1994 period are graphed in Figure 2. These partial productivities in contrast to MFP,
are measures of total sugarcane output divided by a measure of one input, or in the case of land
and labor, two inputs. Such measures can be useful in providing further insights into the
productivity of a system that is dominated by land and labor as they are in sugarcane production.
These estimates show a precipitous downward trend up to the mid-1970s, followed by a relatively
low-level constant trend.
Indexes of diversification are presented in Table 3 and graphed in Figure 3. The growth
rates for these indexes show the same signs in the two subperiods to those for MFP rather than an
inverse pattern within the subperiods. However, over the entire range of the data, the slight
increase in diversification is associated with a slight decrease in MFP. This result is consistent
with what Langham et al. (1997, 1998) found in Florida and in Jamaica.
Figure 2. Partial Productivities in the Sugarcane
Subsector of Trinidad and Tobago
3 -- Land & Labor
02 6 -- Labor
1 00 C '0 00
\D0 -t r- 00 00 00 C
.- -. .-
Entropy Indexes of Diversification
for Trinidad and Tobago. 1963 through 1994
Year Index Year Index
1963 2.48 1979 2.90
1964 2.64 1980 2.92
1965 2.60 1981 2.90
1966 2.64 1982 2.82
1967 2.67 1983 2.82
1968 2.71 1984 2.86
1969 2.71 1985 2.67
1970 2.74 1986 2.64
1971 2.79 1987 2.72
1972 2.85 1988 2.67
1973 2.92 1989 2.75
1974 2.87 1990 2.79
1975 2.82 1991 2.80
1976 2.79 1992 2.84
1977 2.85 1993 2.91
1978 2.83 1994 2.88
Note: The maximum value this index could have
is 4.14 = In (63). A value of zero
would occur with perfect specialization.
Figure 3.-Entropy Indexes of Diversification in the
Agricultural Sector of Trinidad and Tobago, 1963-94
3 -------------------- --------
2 .5 .. ': .. .' .. .
2.4 . .. ,
Productivity Analysis: Some Preliminary Findings
In research, data are nearly always a problem. This is doubly true when one is using time
series data and the information needed has been only partially collected and recorded. The results
presented in Table 4 were estimated with an incomplete data set where gaps were filled with linear
interpolations and extrapolations. It is for this reason, that we present the results as preliminary and
suggestive of the kinds of analysis needed to make adequate productivity policy4.
The Model and Variables.-The model used specifies MFP as a part of a system in which MFP,
scale, and diversification are all simultaneously determined in a system which includes the MFP
equation of interest. Specifically, equation (1) was used to estimate the MFP function in the
system. Here, the x's are used to represent the exogenous variables and they's the other two
endogenous variables appearing in the MFP equation:
(1) InMFP, = a, + a, lnx, +a4Inx,, +a, nx Q + 2r k lny +Int
where: x,, = research expenditures in constant dollars averaged over a one and
a five-year lag in period t,
x2, = openness of the economy measured as the ratio of imports plus
exports to gross domestic product in year t,
x,, = life expectancy at birth measured in year t, and
4 Human capital investments of the kinds societies make through their public school systems and universities and
through public research, extension, and development initiatives have played a dominant role in productivity growth. An
adaptation by Kendricks (1977) of earlier work by Denison (1974) suggests that well over 80 percent of the gains in
productivity in the US non-residential economy came from factors associated with advancements in knowledge during
the period 1929-69. Earlier, Solow (1957) estimated that about one-half of the growth in productivity came from new
knowledge in the form of more productive technologies. Evenson et al (1979) have summarized some 60 estimates of
the internal rate of return from agricultural research. Over 50 of these estimated internal rates of return were above 30
percent Our point is that investments in education and research are important determinants of productivity changes.
X4t = a zero-one variable which takes on the value zero through 1975
and the value 1 for the years 1976-94.
y, = an entropy index measure of agricultural diversification in year t,
Y2t = a proxy for the scale of farms measured as the estimated average
size in acres in year t, and
/t, = an error term.
The estimated a's provide a measure of the direct effect of a one percent change of the
exogenous variables on the percentage change in MFP. In addition to the direct effect of the
exogenous variables appearing in equation 1 on MFP, there is an indirect effect of all exogenous
variables in the system through the endogenous variables appearing on the right hand side of
equation (1)--i.e., diversification and scale. These indirect effects depend on the estimated y's and
the reduced-form parameters which measure how the associated endogenous variable is changed
by movements of the exogenous variables in the system.
The following text table provides a classification of the variables:
Endogenous Variables Exogenous Variables
In MFP Equation In the System but not in Equation
1. MFP 1. Research Expenditures 1. Unemployment Rate
2. Diversification 2. Openness 2. Percent GDP from Agriculture
3. Scale of Farming 3. Life Expectancy at Birth
4. Oil Development
Equation (1) was estimated as being part of a larger system in an attempt to avoid the
problem of a simultaneous equation bias in the estimates. All variables were measured in
logarithms except the zero-one variable for oil development. This simultaneous equation
specification suggests that the system which determines MFP goes beyond a single explanatory
TSP (Hall, 1993) software was used to estimate the parameters in equation (1) by a
combined use of instrumental variables with AR1 using the Cochrane-Orcutt (CORC) method
(hereafter IV(AR1), see TSP, Section 5.6.1, p.39). The evidence of serially correlated errors was
the major factor leading to the selection of estimates from the IV(AR1) procedure. To estimate
the total effect of an exogenous variable on MFP, one must add to its direct effect its indirect
effect through the other endogenous variables. As a basis for getting a measure of this total effect,
reduced form equations which express each endogenous variable in the system as a function of all
the exogenous variables were estimated. The estimated reduced-forms equations for the
endogenous variables appearing as arguments in the MFP equation were estimated separately by
AR1 using TSP software. These AR1 results were used in the analysis because of evidence of
serially correlated errors.
In general, the direction of the net effects indicated by the point estimates of the
coefficients of the exogenous variables on MFP were consistent with prior expectations.
These point estimates of the elasticities of MFP with respect to the explanatory variables are
presented in Table 4. Because of the squared term, the estimates for research was computed at
the logarithm of its geometric mean. With the exception of life expectancy, the signs of the
estimated elasticities were as expected from what is known about the determinants of MFP. The
indirect effects were estimated by using the estimated coefficients in the reduced-form equations
for the effect of each exogenous variables on the two endogenous variables in the MFP equation,
and then multiplying these respective rates of change by the associated rate of change in MFP
Preliminary Estimates of Effects of Causal Variables on Scale,
Diversity, and MFP in the Agricultural Sector of Trinidad & Tobago
Causal Indirect Effect on Total
Variable Direct Effect MFP Effect
MFP Divers. Y1 Scale. Y2 thru Y1 thru Y2 on MFP
Research -0.06113 0.08618 0.21171 0.09911 -0.02901 0.00897
Openness 0.15515 0.02985 -0.09202 0.03432 0.01261 0.20208
% Unemployment 0.03192 0.01176 0.03670 -0.00161 0.03509
Life Expect. -2.98792 -0.79199 -2.69496 -0.91082 0.36933 -3.52941
% GDP fromAg. -0.07278 -0.01373 -0.08370 0.00188 -0.08182
Oil Dev. -0.02074 -0.00573 0.03950 -0.00659 -0.00541 -0.03274
* The total life expectancy effect marked with an asterisk has an unexpected sign.
when the respective endogenous variable changes. For example, when there is a one percent
increase in research expenditures there is a direct negative effect of (-0.061) percent decrease in
MFP which one would not expect. At the same time the one percent increase in research
expenditures increases both diversity and scale which is as one would expect. These increases in
diversity and scale have an estimated indirect effect on MFP so that the total estimated effect of
research is 0.009 percent--i.e., a one percent increase in agricultural research directly and
indirectly leads to a slight percentage increase in productivity of .009 percent.
It is important to recognize that variables endogenous to a system cannot be manipulated
as policy instruments. The forces which drive agriculture toward a more diversified system are
the same as those which drive it toward lower productivity. Indeed, the most diversified
agricultural systems in the world are among the poorest and among the least environmentally
friendly, and there is evidence that the opportunities to spread risks through diversification have
been recognized and effectively exploited by farmers (Langham, 1992, pp. 81 and 87-89; and
Quiroz, 1994) However, farmers also recognize that increased diversification introduces more
problems and needs for information and investment capital in their business. They also know that
these added needs will compete for their management time and production resources and hence
make it more difficult for them to stay at the cutting edge of technology for the commodities with
which they have valuable experience.
Diversity and scale as endogenous variables.--Since the information on quantities and revenues
of the individual agricultural commodities produced in each year is basic to the computaion of the
numerator of MFP and the entropy index of diversification, there is little doubt that these two
variables are endogenously and simultaneously determined. However, the classification of scale is
more problematic. In an earlier study in Florida (Langham et. al., 1997), we modeled scale as an
endogenous variable since it is seen as a consequence of internal farm firm decisions associated
with competitive market forces and efforts by farmers to make an adequate living. However, in
our recent study in Jamaica (Langham, et. al., 1998) we elected to model scale as an exogenous
variable based on the substantial body of literature suggesting that the complexity of regional land
tenure patterns introduces significant rigidities in farm land market5. As such, the implication is
that conscious public policy is the only effective way to address scale issues associated with these
After a substantial review of the orthodox literature and some more recent findings which
challenge the orthodoxy (Newman and Le Franc, 1994), we elected to treat scale in this study as
an endogenous variable operating in a similar fashion to MFP and diversification. The rationale to
proceed along this line stems from an impressive body of new evidence by Newman and Le Franc
relating to the "strategic flexibility" exhibited within the Jamaican small farm sector over the
5 Scale was also modeled endogenously using the Jamaican data for purposes of comparison. The numerical
results in this effort were consistent with those for Florida and Trinidad in that they were more consistent with
1980-1991 period, relative to land acquisition for scale purposes. This new evidence, while
supporting the orthodoxy with respect to the multiplicity and complexity of the land tenure system
(some nine different tenure patterns were identified), non-the-less found that these systems were
less binding on scale (size) than originally thought. The Newman and Le Franc study found that
although farm households expands and contracts in response to changing needs and demands, in
the aggregate, family-related variables are manipulated in such a way as "...to inhibit
fragmentation and instead maintain a fairly stable resource base i.e. a critical land mass" (page
180). The primary family-related variables identified were reciprocity, responsibility and
These findings suggest that debilitating tenure patterns notwithstanding, farm scale
constraints associated with these patterns do not constrain output and productivity enhancements
but are determined along with these enhancements. Rather, the most binding constraint appears
to be the lack of sufficient capital for economic exploitation of existing land resources. Newman
and Le Franc argues (page 180) "...outward migration together with difficulties in obtaining labor,
capital and adequate market systems, have resulted in a situation where there are high levels of
underutilized capacity in the small farm sector: land in particular, remains idle for lack of sufficient
resources for its exploitation."
Exogenous variablesfor human capital.--As mentioned in footnote 4 analysis at the US economy
level has suggested that human capital (including new discovery) accounts for about two-thirds of
the increases in productivity. We attempted to measure investments in human capital with two
variables--investments in agricultural research and the life expectancy at birth in year t. This latter
variable was used as a proxy for both education and medical care--two forms of human capital
investment in the general population. Ideally one would want this variable to be specific to the
rural population. Our expectation was that this variable would be, on average, lower in the farm
than in the general population but the pattern of change would be similar.
Research does not have an immediate effect on productivity, and there is general
agreement that the effects of research accumulate and then decay with time. This decay occurs
because current information becomes obsolete. Biological knowledge becomes obsolete simply
because crop pests change and develop immunities as their environment is changed by the
application of new control methods and/or the planting of new crops or new varieties of crops.
Also, a new plant disease or pest may flourish in a new environment created with the broad
adoption of a new crop or hybrid variety. The consequence of the increased obsolescence of
biological knowledge over time is that considerable research is required in the future just to
maintain where we are today. For example Heim and Blakeslee (1986) have estimated that 70
percent of research on wheat in the State of Washington has been required to maintain yields at
current levels. Other examples of maintenance research are given by Plucknett and Smith (1986).
Mechanical and electronic knowledge also becomes obsolete as new ideas are built into new
equipment models. This type of obsolescence will likely occur at even faster rates as agricultural
decision-makers move increasingly toward greater precision in the use of agricultural inputs.
In this paper we represented the research effort with a variable which represents a rather
short response to research effort and which is defined as the average of a one- and five-year lag of
real research expenditures. This specification follows that used by Griliches (1964). The present
value of distant future payoffs would be modest today. And, knowledge by its very nature is
cumulative, and the payoffs for new ideas today are in part due to the past accumulations of
knowledge (especially from more basic research) on which the ideas are built, so it is not clear
what response time to research should be specified. It is clear that it is a complex set of dynamic
The more basic the research, the longer the period required for its potential to be realized
in production agriculture. It is largely for this reason that there is a very large 'public good' aspect
in basic research because if governments do not fund long-term basic nor track research to
provide an advanced knowledge base, applied research has less payoff. It is largely for this reason
that continued success of privately funded research depends on research efforts in the public
The exogenous variable, openness. --Intuitively one would hypothesize a positive relationship
between the degree of openness and MFP. We measured openness as the ratio of imports plus
exports to GDP. The economic logic for this expected relationship is the assumption that
increased openness would lead to increased specialization at the farm firm level (although
possibly increased diversification at the sectorial level), where this increased specialization would
be associated with inter-country technological spin-offs and higher sectoral investment thus
leading to higher MFP. We find that this hypothesized relationship held for Trinidad and Tobago
for the total effect. Specifically, a one percent increase in openness resulted in a 0.2 percent
increase in MFP.
In order to gain additional insights into the openness/MFP relationship relative to that of
sectoral diversification, we disaggregated the effects of openness in terms of:
1) its direct effect on MFP and diversification, and 2) its indirect effect on MFP via
diversification and scale (farm size)as reported in Table 4.
Some very interesting findings are exhibited in the case of Trinidad and Tobago.
Openness showed a positive direct relationship (0.15) on MFP but a negative effect (-0.09) on
scale6. Openness had a slightly positive direct effect (0.03) on diversification. The total effect
shows that openness fostered a higher level of MFP in Trinidad and Tobago agricultural sector
but that this response is simultaneously associated with a slight reduction in scale and an increase
in diversification. This direct effect of openness on the MFP/Diversification relationship is further
reinforced by the results showing that openness tends to impact positively but indirectly, on MFP
via scale (0.013) and diversification (0.034).
To what extent are the characteristics of the openness variable a major
determinant of the country-specific results observed in Trinidad and Tobago? Furthermore, how
might the findings in Trinidad and Tobago be related to the composition of the computional
specificity and the dynamics of the country macroeconomic data sets used in the analysis?
The exogenous variable, percentage of GDPfrom the agricultural sector.--This variable provides
another measure of the importance of the agricultural sector relative to the non-agricultural
sectors. In part, it measures the economic and political environment in which the agricultural
sector functions and is believed to be important in explaining the incidence of both part-time
farming and scale. Over the period of the study, GDP from farms has decreased from 10.1 to 2.2
The exogenous variable unemployment rate.- This variable was believed important since
marginal farmers are pulled off as well as pushed off the farm. This variable was included as a
measure of the economic health of the larger economy. In times of lower unemployment, the
6 These coefficients are elasticities, i.e., the percentage change in the corresponding endogenous variable when
the exogenous is increased one percent.
environment is on average more conducive to shifting from full time to part-time farming or for
moving out of farming completely. Higher unemployment in the general economy may also
increase the supply of farm labor and hence help keep labor costs down. However, from a societal
perspective this is not an attractive way to increase agricultural productivity because of its
distributive effects. The estimated coefficient for this variable had a positive sign. However, we
did not postulate any a priori sign based on empirical evidence of wage spillovers and
unemployment distortions in wage-gap economies such as Trinidad and Tobago and Jamaica7.
Conclusions and Implications
Increases in agricultural productivity in the agricultural sector are fundamental to the
improved well-being of both Trinidad and Tobago's agricultural producers and its consumers to
the extent that they depend on domestic food supplies. Yet, to our knowledge there have been no
efforts to measure, track, and explain sectoral productivity changes in this island nation. The
absence of such efforts has been to delay the important processes of observing and collecting the
kinds of data needed to do this job well. Our experience suggests that the data to track
productivity are best on the product side in terms of both quantities and prices of commodities
grown, not as complete with regard to the quantities and prices of inputs used, and very sketchy
in terms of the variables believed to be the determinants of productivity. We hasten to point out
that our experience does indicate that the staff of the Central Statistical Office does a very good
job with the resources available to them.
Our preliminary results indicate that productivity in the sector has been declining at a rate
of-0.41 percent over the period 1963 through 1994. If accurate, this estimate means that Trinidad
and Tobago is getting only 88 percent as much out of the same level of inputs in 1994 as it
7 For a review of this economic anomaly see Tidrick (1975).
obtained in 1963. The sector is clearly going in the wrong direction if agricultural production is to
be a part of the Country's future. This estimate also suggest that the sector is not using its
environmental resources in a sustainable way.
Our estimates suggests that there has been a fundamental problem that has two faces. The
problem is a lack of sufficient public investment in human capital to serve the sector. This lack
shows up in the general and continuing education of farm decision makers, in the support of
agricultural research activities, and in the collection and dissemination of data needed to inform
both agricultural producers and policy makers. A small-island economy cannot afford to
undertake the more basic research but it must invest in good scientists who strive to keep up with
global agricultural research and to adapt promising new knowledge for use by local producers.
Rather than invest in human capital, public revenues have been used in ways that seem to
subsidize sectoral inefficiencies such as ill-advised policies to diversify and manage land use for
agricultural production rather than focus on complementary macro policies and important
infrastructures which farmers need to support their production decisions.
Bautista, Romeo. "Rural Diversification in the Philippines: Effect of Agricultural Growth and the
Macroeconomic Environment." Southeast Asia Journal ofAgricultural Economics. 1,
Bruce, Carlton. "The Open Petroleum Economy: A Comparison of Keynesian and Alternative
Formulation." Social and Economic Studies, Vol. 21, No. 2, June 1972.
Christensen, L.R., D.W. Jorgenson, and L.J. Lau. "Conjugate Duality and the Transcendental
Logarithmic Production Function." Econometrica, 39(September 1971):255-256.
Conway, Gordon R. "Sustainability in Agricultural Development: Tradeoffs with Productivity,
Stability, and Equitability." Paper presented at the 11th Annual AFSR/E Symposium,
Michigan, October 5-10, 1991.
Corden, W.M. and J.P. Neary. "Booming Sector and Deindustrialization In An Open Economy."
Economic Journal, Vol. 92, December 1982.
Davis, Carlton G. "Product-Product Dimensions of Agricultural Diversification Strategy in the
Caribbean Community." Carlisle A. Pemberton (Editor). Agricultural Diversification:
Policies and Strategies. Trinidad: Caribbean Agro-Economic Society (1990):30-6.
Delgado, Christopher, and Amar Siamwalla. "Rural Economy and Farm Income Diversification in
Developing Countries." Paper presented at the XXIIIrd International Conference of
Agricultural Economists, Sacramento California, August 10-16, 1997.
Denison, Edward F. Accountingfor United States Economic Growth 1929-1969. Washington,
DC: The Brookings Institute, 1974.
Diewert, W.E. "Exact and Superlative Index Numbers." Journal of Econometrics, 4 (1976): 115-
Evenson, Robert E., Paul E. Waggoner, and Vernon W. Ruttan. "Economic Benefits from
Research: An Example from Agriculture." Science, 205(September 14, 1979):1101-7.
Ezeala-Harrison, Fidel. "Canada Global Competitive Challenges: Trade Preferences vs. Total
Factor Productivity Measures." American Journal of Economics and Sociology. 54,
Griliches, Zvi. "Research Expenditures, Education and the Aggregate Agricultural Production
Function." American Economic Review. 54(1964):961-74.
Habasch, Mona. The Impact of Diversification on Productivity in US Agriculture, unpublished
MS thesis, University of Florida, Gainesville, 1989.
Helpman, E. and P. Krugman. Market Structure and Foreign Trade. Cambridge: MIT Press,
Hall, Bronwyn H. Time Series Processor Version 4.2 User's Guide. Palo Alto, CA: TSP
Heim, M.N., and L. Blakeslee. "Biological Adaptation and Research Impacts on Wheat Yields in
Washington," Department of Agricultural Economics (Mimeograph), Washington State
Kendrick, John W. Understanding Productivity: An Introduction to the Dynamics of
Productivity Change. Baltimore: The Johns Hopkins University Press, 1977.
Krueger, Anne O. "Government Failure in Development," The Journal of Economic Perspectives,
3 (Summer 1990):9-23.
Langham, Max R. "Determinants of Productivity in the Agricultural Sector with Implications for
Research Policy and Analysis." Agricultural Policy Analysis in Sub-Saharan Africa,
Proceedings of an International Symposium, Max R. Langham and Francois Kamajou
(Editors). Gainesville: University of Florida, Office of International Programs, 1992,
Carlton G. Davis, Carlisle Pemberton, Ballayram, and Edward Evans. "Understanding
Productivity, Its Importance, and Relationships to Diversification: Some Agricultural
Policy Issues." International Working Paper Series, IW 98-2, Food and Resource
Economics Department, University of Florida, March 1998.
SFlorence Tangka and Sharon Roberts. "The Importance of Public Sector Investments
in Human Capital for Productivity Growth in the Agricultural Sector of Florida." Staff
Paper SP 97-16, University of Florida, Food and Resource Economics Department,
McCalla, Alex, and Alberto Valdes. "Diversification and International Trade." Paper presented at
the XXIIIrd International Conference of Agricultural Economists, Sacramento, California,
August 10-16, 1997.
National Academy of Sciences. Underexploited Tropical Plants with Promising Economic Value.
Washington, DC, 1975.
Newman, Margaret and Elsie Le Franc. "The Small-Farm Subsector: Is There Life After
Structural Adjustment?" Elsie Le Franc (Editor). Consequences of Structural
Adjustment: A Review of the Jamaican Experience. Kingston: Canoe Press of the
University of the West Indies, 1994, pp: 118-203.
Pantin, Dennis. "Into the Valley of Debt." Trinidad: Ferguson Publishers, Ltd. 1989.
Plucknett, D.L. and N.J. Smith. "Sustaining Agricultural Yields." Bioscience 36(January
Quiroz, Jorge A. "Agricultural Diversification and Policy Reform." ILADES/Georgetown
University, Santiago. The World Bank, December 1994, Revised June 15, 1995
[Ag_diver.email@example.com (January 7, 1997, 13 pages)].
Roberts, Sharon, and Max R. Langham. "Productivity Growth, Technical Progress and Efficiency
Change in the Caribbean: Key Ingredients for 'International Competitiveness'".
International Working Paper Series, IW97-16, Food and Resource Economics
Department, University of Florida, November 1997.
Schultz, T.W. Transforming Traditional Agriculture. New Haven: Yale University Press, 1964.
Seers, Dudley. "The Mechanism Of An Open Petroleum Economy." Social and Economic
Studies, Vol. 13, No. 2, June 1964.
Shannon, C.E. "A Mathematical Theory of Communication." The Bell System Technical Journal,
Solow, Robert M. "Technical Change and the Aggregate Production Function." Review of
Economics and Statistics, 39(1957):312-20.
Stiglitz, Joseph. "Markets, Market Failure, and Development." Perspectives on Economic
Development. 79, 2(1989):.
Tidrick, Gene. "Wage Spillovers and Unemployment in a Wage-Gap Economy: The Jamaican
Case." Economic Development and Cultural Change. 23,2(1975):306-24.
Timmer, C. Peter. "Crop Diversification in Rice-Based Agricultural Economies: Conceptual and
Policy Issues." R.A. Goldberg (Editor), Research in Domestic and International
Agribusiness Management. Vol. 8. Greenwich CT: JAI Press, Inc., 1988.
US Congress. US Export Competitiveness. Washington DC: House Committee on Agriculture,
US Government Printing Office, 1985.