The Issue

Economics is a data-rich discipline with quantitative evidence that can check errant theorizing. Other social sciences are less fortunate, an especially challenging problem in analyzing changes in data-weak developing countries. Development assistance agencies with programs in these countries can be too easily guided by theories of growth that run ahead of supporting evidence, a problem for all the social sciences.¹

In this context, it is perhaps not surprising that there are conflicting views and evidence on the causal linkages between democracy (or democratic governance, hereinafter D/G) and economic growth. The United States Agency for International Development (USAID) assumes such a linkage. Two of the Agency’s five major goals include building democracy and encouraging broad-based economic growth. Field missions are challenged to develop and implement strategic plans that include initiatives (intermediate results below the level of strategic objectives) that contribute to both D/G and economic growth. The usual assumption is that D/G is the independent causal factor that improves an economy’s performance. In the absence of widely accepted evidence supportive of these linkages, there are advocates of the opposite causal relationship, as partially reflected by those who would delink trade and human rights policies toward growing overseas markets for U.S. exports. These advocates argue that market-based economic growth will lead to greater democracy, although the reasons for this are not always clear.

The current focus on markets and exports helps to narrow the meaning of economic growth and the means to achieve it. Economic growth is thought to be generated by increased private sector trade and investment (hereinafter T&I). Increased T&I in turn is thought to be contingent on the creation of market-friendly policies. These policies are produced by governments, and, therefore, D/G plays a central role in increasing economic growth through the policies issued and implemented by governments.

How market-friendly policies are produced and then implemented involves political processes not adequately captured by economics, theoreticians of democracy, and those who would link D/G to economic growth. Analysts tend to limit themselves to assessing the relationships among various quantitatively and qualitatively-based indices of economic growth, policy environments, government performance, democracy, and governance. The black box that helps explain how independent policy variables act on dependent growth ones is left unopened.

These independent variables can be drawn from various readily available indices, such as:

The conservative Canadian Fraser Institute’s Economic Freedom of the World: 1975-1995 (James Gwartney, Robert Lawson and Walter Block) assigns scores to seventeen dimensions, weights for each, and an overall Freedom Index. The dimensions cover such items as foreign currency accounts, government consumption, equality under the law, tax rates, and constraints on capital mobility.

The conservative Heritage Foundation’s 1996 Index of Economic Freedom (Bryan Johnson and Thomas Sheehy) bases its overall scores on ratings assigned to ten dimensions, including trade, taxation, property rights and regulations.

The human rights group Freedom House bases its World Survey of Economic Freedom, 1995-1996 (Richard Messick and Kaku Kimura) on six dimensions ranging from freedom to own property to freedom to participate in a market economy.

Under its PEDS III USAID project, SRI International ranked countries on nine dimensions ranging from import and export regulations to start-up, foreign direct investment, and labor matters.
Private-sector risk-assessment agencies also provide multi-dimensional and overall subjective assessments.
The International Country Risk Guide (ICRG) rates political, financial and economic risks separately, and also has scores for corruption in government, the military in politics, political party development, quality of bureaucracy, etc.
The Business Environmental Risk Intelligence (BERI) index takes a similar multi-dimensional approach to its overall indices.

USAID/Washington offices have similar indices — for example, the Africa Bureau’s multi-dimensional Economic Performance Index.²
There are analogous measures of democracy and political regimes — for example, two transfers of power after free and open elections, which assumes that everyone is playing by the rules of the democratic game. Freedom House’s annual survey of political and civil rights is a leading example of these measures.

The plethora of variables, indicators of them, and analytical approaches help explain why it has been difficult to gain a consensus on linkages between D/G and economic growth (or T&I). However, several threads in the recent economic growth literature suggest some new approaches to conceptualizing D/G and the mechanisms through which if affects an economy’s performance. This paper reviews selections from this literature to suggest how independent and dependent variables appear to be linked at the (micro) project level, how national-level D/G environments affect project-level performance, and how certain features of D/G and economic growth are related at the national level.

Although the variables to be introduced can be labeled independent (causal) or dependent (result), the relationships between them have a static, black box quality. Successful strategic plans and the management of change require an understanding of mechanisms and dynamic forces that help explain the observed statistical relationships. New (or almost new) approaches to providing this understanding will be introduced, followed by observations on the implications these approaches have for program strategies and their management. The simple but basic distinction to be introduced is between institutionalized participation and how participants are structured.


Key Concepts and Distinctions

Aside from different measures of actual economic performance, the research literature distinguishes between the policy environment and (domestic) sociopolitical processes that influence this environment. Policies have a substantive content — for example, credit, investment, and deficit-reduction policies on paper and as implemented. The different meanings of process include participation, one of USAID’s current priority operational approaches. The Agency’s Strategies for Sustainable Development devotes two paragraphs to participation, giving examples of what field missions should include in their strategic plans and the management of their implementation.

Participation is generally thought to be a means to achieve different ends (in USAID’s current lexicon, results and higher-level strategic objectives). These results (for example, specific market-friendly policies) can, of course, be produced with a minimum of participation. A few key decision-makers, for example, can decide to remove price controls. Other policies may require citizen pressure (participation) to effect their reform, and participation will be required to implement and enforce the reforms.

There is much to be learned about how participation affects policies and their implementation. The literature review here is still at a fairly high level of generality. One study reports on how participation at the project level affects that particular project’s performance. A World Bank economist proposes to scale-up project-level relationships to the level of (national) institutions. A third study reports on how national-level participation seems to affect project-level performance.

This highly abbreviated review of this research is followed by some suggestions on the implications this research has for institutional analysis more generally and for D/G programs in particular. Here the emphasis is on how institutions can structure participation in different institutional spheres. The implications this research has for USAID’s traditional approach to institutional development awaits attention in a separate forthcoming paper.


Recent Research

1. Participation and Performance at the Project Level.
Researchers at the World Bank and the USAID-funded Center for Institutional Reform and the Informal Sector (IRIS) at the University of Maryland analyzed the extent to which project performance is influenced by beneficiary participation in 121 World Bank rural water projects. (Jonathan Isham, Deepa Narayan and Lant Pritchett, Does Participation Improve Performance? Establishing Causality with Subjective Data, IRIS Report No. 67.) Much of the analysis is devoted to establishing the quality of the subjective ratings on both independent and dependent variables. The authors conclude “that increasing participation directly causes better project outcomes, at least for the water sector.” Their research, however, does not identify the “policy instruments” that “help to achieve more effective participation.” The authors note than “[a]n analytical approach that incorporates participation might examine the various mechanisms whereby cooperative action by groups can overcome the inefficiency of individualist solutions — for example, from free riding or strategic (mis)revelation of private information — while avoiding the limitations of centralized government.” Elinor Ostrom, Larry Schroeder and Susan Wynne, among others, have addressed this challenge in their Institutional Incentives and Sustainable Development: Infrastructure Policies in Perspective, Westview, 1993.

2. Scaling-up Participation from Projects to Institutions.
In his introduction to Robert Picciotto’s Putting Institutional Economics to Work: From Participation to Governance (World Bank Discussion Paper 304, 1995), Ismail Serageldin notes “the changing role of development projects, which began as instruments of public or investment finance but have become a key vehicle for institutional and policy reform.” Picciotto

proposes a systematic approach to the design of institutional arrangements. . . . The implications for development practices are clear. Markets should be used where feasible and effective. But there is no efficient market, nor a strong civil society in the absence of good government. Projects can thus be viewed as inputs into balanced institutional development strategies. Picciotto builds on a project data base (probably used by Isham and his associates) to support the view that the traditional project is becoming an instrument for institutional and policy reforms, rather than just a "production function.” His report, however, does not present systematic evidence on the success of project-level participation being scaled-up to improve governance and broader institutions. Nor is there evidence on how the structure and operations of institutions relate to democracy, although one might assume that Picciotto’s “governance” incorporates the meaning of D/G.

3. National-Level Participation Effects on Project-Level Performance.
In another collaborative World Bank-IRIS paper, Jonathan Isham, Daniel Kaufmann and Lant Pritchett (Governance and the Returns to Investment: An Empirical Investigation, IRIS Working Paper No. 186) link measures of societal-level participation to project-level performance. The participation variable comes from Freedom House’s Civil Liberties Index, one that covers a checklist of thirteen participation-related items, including the right of peaceful assembly, freedom of opinion and expression, the right and opportunity to take part in the conduct of public affairs, the right to freedom of opinion and expression, and the right to form trade unions. (Freedom House’s Political Rights index relates to elections, electoral laws, opposition parties, freedom from the military, etc.) Project-level performance measures refer to economic rates of return from 1,155 World Bank projects in eight economic sub-sectors where the stream of benefits could be readily quantified and valued.

The researchers found that “[o]n average, improving a civil liberties indicator by one standard deviation increases the predicted ERR by over 3 percentage points (the mean rate of return is 16 percent).” Further, they conclude:

. . . this relationship suggests a causative effect from better civil liberties to better project performance. First, prior work suggests that beneficiary involvement and accountability of implementing officials, two aspects of governance related to civil liberties, are both key elements of project success. . . . We conclude with a growth accounting exercise that shows that the economic rate of return is related one-for-one to the economy-wide rate of return, suggesting that it is a good proxy for overall investment performance. Overall, we believe that this paper is an additional piece of evidence for the view that increasing public voice and accountability — through both participation and better governance — can lead to greater efficacy in government action, including development assistance. . . . This two-level relationship helps overcome a problem inherent in the earlier analysis of project-level participation’s effects on project performance. Those were local infrastructure projects (the kinds studied by Ostrom and associates). Such projects typically require local community responsibility for many operational and maintenance functions. Participation is central to the sustainability and success of the projects. Participation is not something external to the projects. The evaluations of project performance, therefore, address key features built into the project design and implementation. In contrast, the two-level relationship covers projects in eight different sub sectors, including non-infrastructure investments. And the measures of participation are external to individual projects, although one might hypothesize that open, participatory societies would more likely have the highest levels of project-level participation.

The above research points to promising ways of linking one dimension of D/G (i.e., participation) to economic growth. A number of questions, however, require further study. For example, the means and success of scaling-up from project-level participation to the larger institutional environment need to be identified, verified and better understood. The available evidence points to downward rather than upward-flowing linkages: society-wide participation (the civil rights index) affects project-level performance. Also, participation is more than a process, for it entails participants, the degree to which they are organized, the mechanisms they are able to use to influence public policies and their implementation, and different forms of participation. Traditional social science research (e.g., on constitutions and laws) provides one approach to investigating these questions. The New Institutional Economics (NIE), including the collective-action theories formulated by the late Mancur Olson and others, is another promising approach.

Examples from both traditional approaches and the NIE will be mentioned to suggest how participation and participants are structured in ways that provide mutually-supportive links between D/G and economic growth.


The Structure of Participation

1. African Ethnic and Political Structures
William Easterly and Ross Levine of the World Bank produced two recent Africa-specific policy research working papers: “Africa’s Growth Tragedy, a Retrospective, 1960-89,” dated August 1995, and “Africa’s Growth Tragedy: Policies and Ethnic Divisions,” dated April 1996. The authors argue that there is still much we don’t know about how African economies operate.³ We knew less thirty years ago. In the 1960s, the World Bank’s chief economist listed seven African countries that “clearly have the potential to reach or surpass” a seven percent growth rate. In fact, they had negative growth.

Economists are quick to lay the blame for this failure on bad policies. Easterly and Levine argue that the explanation must go well beyond this one set of claims:

The World Bank recently concluded that most African countries have made little progress in reversing the decline [in economic growth]. These studies identify a diverse set of potential causes of Sub-Saharan Africa’s ills ranging from bad policies, to poor education, to political instability, to inadequate infrastructure, to weak institutions, but are unable to explain why these factors all went so badly wrong in Africa. If economists are to claim much success in explaining why some countries are rich and others poor, why some countries choose growth-enhancing policies and others growth-retarding ones, an explanation of Africa’s tragedy must be part of the package. The authors do not argue against good policies and investments. Instead, the question is why these policies and investments have not been made or effectively implemented.

Easterly and Levine analyzed thirty years of data on standard measures of growth and other variables economists use to explain it. The explanatory variables include school attainment, political instability, poorly developed financial systems, distorted exchange rates, government deficits, inadequate infrastructure, etc. “These variables account for about half the growth differential between the countries of sub-Saharan Africa and East Asia.”

Social structures — that is, the way participants are structured — enter this analysis through an index of ethnic diversity. Easterly and Levine found that Africa’s ethnic diversity tends to slow economic growth and reduce the likelihood that good growth policies will be adopted. Although the authors lack hard data to explain their findings, they hypothesize that

It may be more difficult to achieve a consensus for good policies in a polarized environment.... We suspect that ethnically fragmented societies are prone to competitive rent-seeking by the different ethnic groups and have difficulty agreeing on public goods like infrastructure, education, and good policies. Furthermore, ethnic diversity may favor policies destructive to long-run growth like financial repression and overvalued exchange rates if such policies create rents for the group in power at the expense of other groups. Academics relying on more traditional qualitative research techniques come to similar conclusions. For example, Princeton political scientists Henry Bienen and Jeffry Herbst (The Relationship Between Political and Economic Reform in Africa, n.d.) argue that African politics and elections are non-ideological; they focus on individuals in power or seeking power, not on economic and policy issues; there is not a significant constituency that backs both economic and political reforms as a single package; civil society is weak, and there is a tendency for the disaffected to “exit” from the political and economic system rather than “voice” their protest. In fact, constituencies for political reform often oppose economic policy reforms supported by the World Bank, IMF, USAID and other donors.

Other relevant conclusions by Bienen and Herbst include:

1. Political reform has mainly allowed autocrats to remain in power rather than bringing in ‘new brooms’ who might be assumed to be more committed to economic reform.
   
   
2. Political change in most African countries has not been accompanied by the creation of constituencies that favor economic reform. Many of the groups in favor of democratization are opposed to structural adjustment.
   
   
3. It is extremely difficult for African leaders to impose economic reform from above because they are so uncertain about the stability of their democratic systems. Indeed, African democrats are probably more insecure than those elsewhere, with the result that the need for new leaders to engage in old-style patronage politics will be relatively acute.
   

Similarly, Easterly and Levine link their findings to corruption, an especially damaging situation when different ethnic groups are competing for government payoffs.

Both the World Bank and Princeton researchers have black box models that link the structures of participants (constituencies) to government policies. Neither research team has systematic information on how participation and participants are really structured, forms of participation, and mechanisms through which participants influence decision-makers and implementers. Constituency analysis and the related stake holders analysis are general guiding ideas with gaps yet to be filled.

Easterly and Levine point to one such gap that has significant implications for D/G. Some countries are able to transcend their high level of ethnic diversity.

[I]n countries with very highly developed institutions, ethnic diversity does not significantly hurt policy choices. Institutional arrangements can overcome the negative implications of ethnic diversity. Although this suggests a reform strategy that focuses on improving a country’s institutions, altering institutional arrangements is fundamentally more difficult than changing, for example, exchange rate policies. Their indicators of institutional development refer to clear property rights, effective rules of the game, and an efficient bureaucracy.

2. Legal and Institutional Systems that Structure Participation.
The democracy literature tends to focus on vertical relations between the state and its citizens (and non-citizens). Donor agencies reflect this perspective in their support for programs on human rights, elections, and a free media. Similarly, many policy-reform programs call for removing governmental constraints on the private sector and different markets.

Less attention is given to horizontal relations between private individuals and entities. This is the arena (that is, the market) in which most economic activities are conducted. Government policies often distort the market. Equally important, the legal and institutional infrastructure to facilitate transactions between private individuals is weak or absent in (most) developing countries. How these infrastructures are structured (how they structure private transactions through positive and negative sanctions) provides a missing link in the participation, D/G and economic growth chain.

This is not to say that the key linkages are adequately understood, as much of the literature is still in the theorizing stage, and the relevant research relies on correlations, regressions, and qualitative judgments. This new generation of economic development theory argues that institutions play a critical role in economic performance. The difference between poor countries and rich ones, according to the IRIS perspective on the New Institutional Economics (NIE), has less to do with the existence of factors of production (land, capital and labor) or even of technology, than with the institutions that affect their deployment. Rules and implementing agencies determine how these factors are obtained, how they are transferred, and how they are used and exchanged. The NIE (including collective-action theories relating to the organization and effects of interest groups) provides a bridge between the rule of law, democratic governance, economic and financial policies and economic growth.

The best recent overview of this approach appears in the USAID-supported book Institutions and Economic Development: Applications of the New Institutional Economics to Growth and Governance in Less-Developed and Post-Socialist Countries, edited by Christopher Claque of the previously-mentioned IRIS Center at the University of Maryland. IRIS Center and World Bank staff have a number of other relevant publications — for example, “Institutions and Economic performance: Cross-Country Tests Using Alternative Institutional Measures” by Stephan Knack and Philip Keefer in Economics and Politics, 7:3, November 1995. Some of this work is in the tradition of Nobel Prize-winner Douglas North’s contributions, including his conclusion that “the inability of societies to develop effective, low-cost enforcement of contracts is the most important source of both historical stagnation and contemporary under development in the Third World....” (Institutions, Institutional Change, and Economic Development, Cambridge University Press, 1990).

Using a measure of contract-intensive money (the ratio of non-currency money to the total money supply) to indicate the state of contract compliance and security of property, the IRIS researchers

[F]ind strong evidence that those underdeveloped nations respecting property rights and contract rights accumulate capital and increase the incomes of their populations at substantially higher rates than other developing countries — even when controlling for the large variation in starting points, in levels of school, and in the price of investment goods. The IRIS researchers further establish the significance of institutions and the rule of law:

In addition to financial development and inflation, we have elsewhere considered the possibility that our institutional indicators are spuriously correlated with investment and growth due to omitted policy variables. In all our regressions, adding measures of government size, trade openness, and macroeconomic stability fails to reduce substantially the estimated impact of our property rights and contract enforceability indicators. Property rights and contract enforceability cover more than written laws and regulations (as well as customary law). They also include structures and mechanisms to implement and enforce the legal system. This larger institutional structure refers more generally to the rule of law.

One might expect that the rule-of-law relating to economic transactions would have spill-over effects on political rights, that good governance is tied to democracy so that the two concepts become democratic governance (D/G). Swaziland (and others) may be exceptions to this combination, as private property and contract enforcement would seem to score ahead of scores on variables of democracy, such as voice and choice. This apparent anomaly is one reason why some groups argue that donor emphasis on D/G can be detrimental to expanding U.S. T&I in particular countries — whether in Swaziland or in China.

What may seem to benefit U.S. T&I in the short run may be detrimental in the intermediate or longer term. Introducing this time dimension raises questions about the forces behind changes in political systems, as well as the legal, institutional, and policy environments for market-based, private sector-led economic growth (via T&I). Stake holder and constituency analyses tend to be static descriptions of current political structures.

Donor agencies — either consciously or not — have time trends built into their normatively-oriented programs. They have a mental image of what a market-friendly institutional environment is. Using this end point as their surprise-free projection, the agencies work backward to the intermediate steps required to move from the current status to the desired one. USAID’s traditional logical framework is a widely-used approach to making these linkages. The Agency’s current approach to strategic planning substitutes some new terms (e.g., intermediate results) and (in this writer’s opinion) confuses some concepts (e.g., the procurement mode as an input), but the underlying logic remains the same.

Again, the challenge is to use this logic to make linkages between D/G and economic growth (T&I) below the strategic objective level. Several suggestions on how to understand what has to be done are briefly noted in the section that follows.

Continue reading
How to Link Democratic Governance
with Economic Growth (Conclusion)

---

END NOTES
(Part I)

1. For example, some experts question agricultural production and yield statistics reported by the Food and Agricultural Organization (FAO), the size and performance of the informal sector, and estimates of GDP.

2. These different indices rely on subjective assessments. Although agencies — such as the Fraser Institute — may base their scores on quantitative data, the cut-off points and score values are subjectively determined. Scores are not continuous variables with equal intervals between consecutive scores. For example, the real policy differences between a score of 2 and 3 may be much greater than between 7 and 8. At least one ranking (by the Heritage Foundation) is by an organization with strong ideological biases (against foreign assistance and USAID). Given the nature of the rankings, it is not surprising to find minimal inter-index consensus for the individual dimensions and, even, the overall composite scores, with exceptions.

3. For a discussion of some of the policy implications flowing from these knowledge gaps, see the author’s “What is In and Who is Out of Africa,” Foreign Service Journal, forthcoming.

---

Return to American Diplomacy home page