Institutions are considered a fundamental determinant of economic growth. And agriculture is the key sector for poverty reduction and sustainable economic growth in low and middle-income countries. The vital role of agriculture would not materialize without sustained agricultural productivity growth driven by public investments in agricultural Research and Development (R&D). The objective of the thesis is threefold: (i) to measure the effect of institutions on agricultural productivity and on public agricultural R&D investments in low and middle incomes countries, (ii) to estimate the impact of public R&D investment on agricultural productivity, and (iii) to test the hypothesis that agricultural R&D spending would differ across countries at different stages of economic development (i.e., examine the role of per capita GDP in determining agricultural R&D spending).Agricultural productivity is measured by yield (cereal yield); institutions are measured by four indicators: protection of property rights, impartial public administration, judicial independence, and legal enforcement of contracts; agricultural R&D is represented by public investments in R&D. Data were collected from the World Bank, the FRASER Institute, and the Agricultural Science and Technology Indicators (ASTI) over the period of 2000 to 2011 and in 49 low-and middle-income countries (25 from sub-Saharan Africa, 11 from LAC, and 13 from Asia). Panel data with fixed effects models were estimated to address the three objectives. The multiple linear regression analysis reveals the protection of property rights and legal enforcement of contracts have a substantial but opposite impact on agricultural productivity (cereal yield) across low-and middle-income countries. The same result suggests that impartial public administration affect positively public agricultural R&D investment while property rights, judicial independence, and legal enforcement of contracts have no significant implications on public agricultural R&D investment. The analysis also indicates that agricultural R&D investment positively impacts cereal yield across low-and middle-income countries. The same analysis carried out in the three regions has shown that the four institution indicators have different effects on agricultural productivity (cereal yield) and public agricultural R&D investment. In addition, the analysis suggests that public agricultural R&D investment significantly impacts agricultural productivity in Asia and the LAC regions but not in sub-Saharan Africa. Furthermore, the result confirms that agricultural R&D spending differs across countries at different stages of economic development.

Agricultural research and development (R&D) investment is becoming an increasingly important policy issue as food prices push upwards and food security problems emerge. An important source of agricultural R&D funding is from producer check-offs, which are increasingly being used to fund applied agricultural research such as disease management, genetic improvement, and weed control. Existing studies of producer-funded agricultural R&D indicate that there are high private and social rates of return to agricultural R&D investment by farmers, and thus that farmers are under investing in R&D. The focus of this thesis is at the producer level. This study examines one of the factors -- the horizon problem -- behind the apparent disincentive for farmers to invest in producer-funded R&D activities. It has been argued that given the long period of time over which the benefits of R&D investment occur, the increasing age of the farm population implies that the horizon problem could be indeed an important factor in producer underinvestment. Contrary to this widely acknowledged argument, this study shows the horizon problem is likely not a factor affecting farmers R&D investment decisions. Two models are developed to examine the horizon problem. The first model consists of a framework for determining the marginal internal rate of return of investing in R&D. Specifically, the model calculates the internal rate of return -- i.e., IRRh -- associated with the farmers' planning horizon and compares this to the internal rate of return -- i.e., IRR bar-- associated with the benefit horizon of the R&D. The impact of the horizon problem is determined by examining the difference between IRRh and IRR bar. The results of the horizon problem model show how that, contrary to what some authors have argued, the horizon problem is likely not a disincentive for R&D investment, unless the time horizon of farmers is very short. Given that the membership horizon for the average Canadian producer is 15 to 20 years, it is expected that the horizon problem is not an issue for Canadian producers. Furthermore, the analysis assumes farmers only are concerned with profit maximization. However, farmers may also consider other factors when making R&D investment decisions, such as future generations of agricultural producers and environment issues. The results of this study show that, even under the assumption of profit maximization, the horizon problem is not an issue for Canadian farmers, let alone in a more realistic model implemented by including factors other than profit. The results of the horizon problem model also show that the impact of the horizon problem is not affected by land tenure relationships. The second model consists of a multi-region, multi-product trade model that is used to examine the impact of Canadian pea R&D funding on consumers and producers in Canada and in various countries around the world that produce and consume pulses. To address the underinvestment issue, it is important to understand the question of who benefits from the research that is undertaken, and who bears the cost. Given that Canada is the largest pea exporter in the world an increase in R&D investment can be expected to have a significant impact on international trade and overseas producers and consumers. The simulation results from the second model illustrate that with increased pea R&D investment, Canadian producers, as well as consumers in all regions, are better off as a result of the R&D investment, while overseas producers are worse off. The results of the sensitivity analysis show that a pivotal supply shift associated with an increased levy, combined with a parallel supply curve shift due to increases in the knowledge stock, does affect the IRR in the large country versus the small country case. This result differs from the result that occurs when there is a parallel shift in supply at both the levy and R&D stages, indicating that it is important to understand the interaction between the manner in which R&D is funded, the way in which R&D affects supply and the trade status of a country. The results of the sensitivity analysis also indicate that the IRR to Canadian producers depends critically on how large an impact pea R&D has on the production of other crops (e.g., wheat and canola). The larger is this impact -- i.e., the more that wheat and canola production falls as a result of higher yields/lower costs of pea production -- the smaller is the IRR. The results also indicate that the elasticities of demand for peas and lentils in the importing countries do not have an impact on the IRR in the case where Canada is a large country exporter for peas only; however, they do have an impact on IRR in the case where Canada is a large exporter for both peas and lentils. In all cases, the more elastic is the demand, the higher is the IRR.