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 WFAS research may be found on the Electronic Data Information Source (EDIS)
site maintained by the Florida Cooperative Extension Service.
Copyright 2005, Board of Trustees, University of Florida
INTERNATIONAL AGRICULTURAL TRADE
AND DEVELOPMENT CENTER
PRODUCTIVITY AND DIVERSIFICATION CHANGES
IN THE AGRICULTURAL SECTOR OF TRINIDAD
AND TOBAGO: SOME EMPIRICAL FINDINGS
AND POLICY ISSUES By
Max R. Langham and Carlton G. Davis
IW98-4 June 1998
INTERNATIONAL WORKING PAPER SERIES
UNIVERSITY OF FLORIDA
Institute of Food and Agricultural Sciences
Food and Resource Economics Department Gainesville, FL 32611
Nff SSION AND OBJECTIVE OF THE
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 so by:
I 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 thecollection 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 diversifyr 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 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 (DABE/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 Diversification'
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 pectoral 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
This section borrows heavily from Langham ct 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. If Yt is an index measure of the total output of goods and services (production) coming from the agricultural sector in year 1, and Xt is an index measure of the total inputs (factors) used in year t, then MIFPt 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, MIFP 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 MEP 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 capita incomes, and to changes in the relative purchasing power of currencies.
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 surv iving 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 resourcepoor 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 Siarnwalla (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 activities required.
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 different.
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 (groundnuts),
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-i-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 development mode.
Important insights from T. W. Schultz, who shared the Nobel Prize in Economics with Sir Arthur Lewis, in the 1 960s 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 governent 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 productivity.
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 Institute (CARDI).
Multifactor Productivity (MFP) Indexes
Multifactor productivity measures estimated for this paper were from Tsrnqvist-Theil (TT) 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 inputs, respectively.
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 YWX, 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 chosen.
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, ifpi, is the proportion of total revenue from agricultural commodity i in year t then the entropy index for n commodities for year I would be:
I, (on ,) 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 l1n 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 MIFP 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 MIFP over two subperiods 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 M1FP 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 openn 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 of Multifactor Productivity (MFP) Indexes
for the Agricultural Sector of Trinidad and Tobago with the Use
of Tomqvist-Theil Approximations to Divisia Indexes 1992= 1.00
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
1.1.12131 :::::::::::::: '
0.9 --.... Outputs
..... ..... ::..... .. .. ... .... .. . . . .................... 'i:
09 75 0.0255 00041
1 6 40.70 3 ....... ...... ..51IF
and M Dierictio incthel AicutrltetrofTinddanroagyyPeid
09.-76 .969 an 1963-94.....
1963-75 0.02723 0.0045
1976-94 -0.00723 0.003414
1963 -94 -0.00371 0.00197
1963 -75 0.01014 0.0064
1976-94 -0.00342 0.002965
1963-94 0.00037 0.00151
3. DivPrithctuan ty
1963-75 0.01723 0.001
1976-94 -0.00089 0.00465
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-1I980s ( 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 MIP/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 MIFP 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 MIEP and diversification would exhibit similar trends "with" and "without" such subsidies but that MIFP 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 sector.
Estimates of partial pro ductivities (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.
Figure 2. Partial Productivities in the Sugarcane Subsector of Trinidad and Tobago
...... .. .. .. . ...Diversification Trends
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.
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
2.6 gii 1'I
2.3 -.. ..
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 = ao+I'alnx, +a41nx, +aSlnxQ + rkiny, + lnu,
where: x,, = research expenditures in constant dollars averaged over a one and
a five-year lag in period t,
x2t = openness of the economy measured as the ratio of imports plus
exports to gross domestic product in year t,
x3t = 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.
XQt= a zero-one variable which takes on the value zero through 1975
and the value I 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
ut= 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 MIFP, 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: Endouenous Variables Exoenous Variables
In MFP Equation In the System but not in Equation
1. MIFP 1. Research Expenditures 1. Unemployment Rate
2. Diversification 2. Openness 2. Percent GDP from Agriculture
3. Scale of Fanning 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 relationship.
TSP (Hall, 1993) software was used to estimate the parameters in equation (1) by a
combined use of instrumental variables with ARI 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 ARI using TSP software. These ARI 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 MEP 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 MEP in the Agricultural Sector of Trinidad & Tobago
Causal Indirect Effect on Total Expected
Variable Direct Effect MFP Effect Sign
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 from Ag. -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
MEP 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 tenure patterns.
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
5Scale 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 conceptual expectations.
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 residence.
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 variables for 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 forces.
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 domain.
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 MEP. 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 WFP.
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 MEP and diversification, and 2) its indirect effect on WIP 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 GDPfroni 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 percent.
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 Jamaic a.
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 resourses 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 7For 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 11 th 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 XXIIId International Conference of
Agricultural Economists, Sacramento California, August 10-16, 1997.
Denison, Edward F. Accounting for 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): 11545.
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 0. "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.
Florence 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 XXIIId 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
[Agdiver.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.