Causes of and constraints to agricultural and economic development: discussion.

This session revisits some of the most well-known propositions of the agricultural development literature, such as

* Agricultural productivity growth must precede or accompany economic development.

* OECD agricultural protection stunts Less Developed Country (LDC) agricultural progress.

* Foreign aid is necessary for economic development.

* Food aid can be delivered in ways that support LDC agriculture.

These papers reveal that our discipline still has plenty of work to do to understand these issues.

Tsakok and Gardner examine the proposition that agricultural productivity growth is necessary for economic development, a frequently used rationale for investments in agricultural research. They review recent econometric evidence that calls this proposition into question as a universal truth, and then explore case studies to find evidence of refutation. The four countries considered span the history of economic development, from the first industrialized country to the most recent entrant into "middle-income" status. They find only mixed evidence that agricultural productivity played a key role, and in some cases, simply not being "a drag" on other sectors appears to have been sufficient.

This descriptive analysis, one that considers the special circumstances of each case, is illuminating and fun to read. In their forthcoming book, I hope they can also examine some key elements of the structural transformation framework that are missing in the paper. First, a substantial proportion of the labor force and the population are "trapped" in agriculture as the process of structural transformation proceeds. Thus, while general economic growth may not rely on agricultural productivity, it is possible that poverty reduction is achieved more readily with such productivity gains. Second, the interface of the demographic transition with the process of structural transformation should also be included in the descriptive analysis. Reduced mortality and improvements in public health accompany agricultural and economic growth. Where they are missing, as in the stalled demographic transition in Sub-Saharan Africa (SSA), the process of economic transformation is stalled. I hope these authors can find a set of necessary conditions for agricultural productivity to be an engine of economic growth and poverty reduction.

The paper by Dewbre, Thompson, and Dewbre has an ambitious agenda, as it examines two of the above four propositions. Growth in agricultural sector GDP per worker is regressed on various measures of foreign aid to agriculture and OECD protection for 1986-2004. The paper's original contributions include a focus on agricultural aid, as well as accounting for similarities in crop production among LDCs and OECD countries. They find that foreign aid has a significant, negative relationship to agricultural growth, and that the impact of OECD protection is mixed. The foreign aid result holds even after accounting for endogeneity--implying that foreign aid leads to poor performance (although this result does not appear to hold for SSA). Clearly, in the current Sachs versus Easterly debate on this issue, their results support Easterly's findings.

Regarding protection, the maintained hypotheses appear at odds with stylized facts. First, the authors state that rice, cotton, and sugar are the principal crops grown in LDCs that are also protected in OECD countries. Second, they suggest positive spillovers from OECD technology and trade for countries that grow crops also grown in OECD countries. Let me point out just one (of many) counter-example to both hypotheses: maize is one of the most widely grown and consumed staples by the poor in the least developed countries, and the yield gap between rich and poor countries is astonishing. In 2005, maize yields varied from around 0.5 tons/ha in African countries such as Malawi, to 9.3 tons/ha in the United States. Surely, the protection and support for feed grains in OECD countries as well as the lack of adaptation and transfer of technology for maize call into question the Dewbre, Thompson, and Dewbre hypotheses. Their positive finding for the relationship between agricultural growth and production overlap is intriguing, and deserves a closer look.

The paper by Kirwan and McMillan takes a fresh look at an old topic that has seen renewed interest in recent years (e.g., the FAO's 2006 State of Food and Agriculture report focused on food aid). How to make food aid supportive of development was outlined in the agricultural development literature in the 1970s, but this knowledge has not been translated into good policy. Kirwan and McMillan expose the shocking disconnect between food aid and need on the one hand, and between food aid and agricultural development on the other. Perhaps, the question is why massive amounts of food aid fail to motivate late-developing SSA countries (and their donors) to invest in agricultural productivity, in contrast to experiences in South Asia in the 1960s. This brings me to my last point regarding the timing and context for agricultural development.

Overall, the session suggests that all four of the above propositions are false, or at least not universally held truths. This is troubling because it is unclear what high-income countries (and high-income agricultural economists) can do to support agricultural development in LDCs. These three papers use very different time periods and datasets. In revisiting these propositions, it may be beneficial to consider the historical and geographic context for early and late developers.

Late arrivals can "leap-frog" over stages of technology development or social progress. Transformation of all kinds has occurred more rapidly in countries developing later. But late arrivals also may face significant barriers to "catching up," in terms of achieving competitive levels of productivity in the face of lower real prices in world markets dominated by early adopters of new technology. The dynamics of world markets and global diffusion of innovation have changed in recent decades. Agriculture's stylized role in development may have certain core characteristics and may require certain necessary conditions, but surely the way its role plays out will differ as the global context for development changes. This session has made a good start on revisiting the four propositions above with a fresh perspective informed by the current global context.

Laurian J. Unnevehr is Professor in the Department of Agricultural and Consumer Economics, University of Illinois at Urbana-Champaign.

This article was presented in a principal paper session at the AAEA annual meeting (Portland, OR, July 2007). The articles in these sessions are not subjected to the journal's standard refereeing process.