Change Font: A A A A Contact Us What's New FAQs Subscribe home
HomePublicationsBrowse ListingAt Different Speeds: Policy Complementarities and the Recovery from the Asian CrisisSecond-Best, Policy Complementarities, and Growth

Second-Best, Policy Complementarities, and Growth

2.1 Theoretical Background

As Bergstrom (2002) pointed out, one of the more disconcerting results in the theory of welfare economics was articulated by Lipsey and Lancaster (1956) in their paper "The General Theory of Second Best." They demonstrated that if there are distortions in more than one market, removing a distortion in a single market may not be beneficial if distortions remain in other markets. This theory generates a disheartening result: piecemeal reforms do not necessarily increase welfare and can even reduce it. The only way to unambiguously ensure an increase in welfare is to eliminate all distortions at once. In 1970, Foster and Sonnenschein proved that under reasonably general circumstances, at least one kind of piecemeal reform—a radial one (i.e., made of proportional reductions in all distortions)—would improve welfare.2

However, a direct application of this highly theoretical approach to policymaking involves extreme and possibly insurmountable difficulties. The information requirements would be immense and even creating a common definition of proportional reduction in very different policy areas would be extremely difficult.

A less demanding framework is thus required. According to Macedo and Martins (2008), engaging several reforms in parallel reflects the idea that reforms are mutually interdependent and, therefore, complementary. This goes back to 19th century economist Francis Edgeworth, who developed the notion of complementarity: activities are Edgeworth complements if doing (more of) any one of them increases the returns to doing (more of) the others. The concept has been generalized in such a way that it does not require any particular differentiability or convexity assumptions—the modern concept of supermodularity3 stipulates that a change in only one coordinate of a system is less than the change associated with a parallel move across several dimensions. This is to say that raising one variable increases the return to raising another.

The basic idea is easy to formalize. Assume an objective function F depending on two policy instruments (x, y). A given policy can have two possible states, either reform (x) or no-reform . The two policies are complementary if:

This means that, for y, the return of moving from minimum is less than of moving from to the maximum (x, y) (and symmetrically for x). Or, in other words, the return from making reform y (or x) is greater when reform x (or y) is already in place. For n policies, F is supermodular if the relations above hold for every pair of reform areas. In such a system, optimizing can be achieved by increasing all reforms in parallel (but not necessarily in the same proportion, as in radial reductions in distortions).

2.2 Complementarities and Policymaking

There are many practical examples that illustrate the importance of policy complementarities. In transition countries, for instance, if an economy becomes more liberalized (i.e., the proportion of prices determined by the market increases dramatically) but a policy of stabilization is not undertaken simultaneously, then inflation may accelerate. This happens because demand pressures become immediately real and measurable.

Or, if a country liberalizes the financial sector but does not have good exit mechanisms—or good bankruptcy laws—its financial system will accumulate bad debts. Moreover, if good exit mechanisms are not complemented with good entry mechanisms, the reallocation of resources will be blocked, with negative consequences in terms of growth and employment. Appendix A is not intended to be exhaustive, but nevertheless provides several examples of policy complementarities. Virtually all possible pair combinations of policies may be understood as being complementary.

By stepping out of a more Walrasian world—that is, a world of more efficient (static) allocations of resources—it is possible to integrate technical progress in this framework. For example, a low inflation environment can permit larger investment horizons and, if the financial sector is sound and competitive, contribute to augmenting the number of financed research and development projects. In addition, low inflation can help to keep the currency strong, making it easier to buy research and equipment goods (if tariffs on imports are not high) in more developed countries, as well as to finance higher-level education and training in Europe and the United States (US). Successful technical progress strategies will help to produce the same with fewer inputs, thus putting a downward pressure on inflation.

The notion of complementarity is based on powerful economic arguments and is doubtless an intuitive idea; however, as Macedo and Martins (2008) note, in the literature relating the design and scope of reforms to economic performance, little attention has been paid to this concept. Nevertheless, it is also true that the subject is attracting the attention of an increasing number of economists.

Azis and Wescott (1997), for instance, found that Washington Consensus-type, individual policies are of little help in promoting fast growth.4 Using both an outcomes-based probability analysis approach and a standard regression approach, the authors demonstrated that favorable combinations of policies can significantly increase a developing country's economic growth performance. They found that the probability that a developing country experienced fast per capita income growth if the country had only a single high-quality policy over the period 1985–1995 was in the range of 0.20 to 0.35. But, this probability jumped to the range of 0.55 to 0.90 when there was complementarity at a high quality level among three key policy areas: trade openness, macroeconomic stability, and a relatively low degree of government involvement in economic activity. The authors also demonstrated econometrically that, although none of these three policies individually is significant in explaining the pace of economic growth, collectively they are significant in explaining growth when they are summarized in a policy complementarity variable.

The authors concluded that while Washington Consensus-type policies are generally the right ones for developing countries to pursue, progress along a multifaceted set of policy dimensions is more critical than it was perhaps thought to be. In fact, as Azis and Wescott point out, it is possible to imagine cases in which adopting some Washington Consensus-type policies, but neglecting to implement other important policies, might actually lead to a growth outcome that could be inferior to a case of making fewer reforms.

Importantly, the authors note that, whereas they suggested a set of three core policies that appear to greatly improve a country's chances of exhibiting rapid economic growth, and that their findings supported the overall logic of the Washington Consensus, there may be other policy combinations that are even more effective in promoting growth.

More recently, Macedo and Martins (2008) carried out econometric tests focused on transition economies in Eastern Europe (European Union [EU] and non-EU members) and in the former Soviet Union. The tested equation was:

where initial conditions are simply the initial level of gross domestic product (GDP) per capita before the transition in 1989, RL (reform level) stands for the simple average of nine sectoral indicators taken from the European Bank for Reconstruction and Development's Transition Report,5 and RC constitutes an index of reform complementarity (captured through the inverse of a Hirschmann-Herfindahl indicator; the same index was utilized later in this paper). The results confirmed that countries with a higher reform level tend to have higher GDP growth, but the variation of RL displays a negative sign. Thus, an increment of reforms usually induces a negative impact on growth, which is typically the second-best result. Over the long run, when reforms become more broad based, higher levels of reforms are related to higher growth rates. The complementarity indicator displays a symmetric pattern, as its level displays a negative sign while its variation has the expected positive sign. Indeed, a high complementarity by itself does not necessarily lead to higher output growth, because, in the authors' sample, unreformed countries may have had, for some period of time, higher complementarity than did reforming ones.6 In brief, only the level of reforms and the changes in their complementarity have a positive impact on growth. Therefore, the former effect provides a long-run target for reforms, while the latter provides guidance on the conduct of the transition process.

For the new EU members, the reform process was characterized by a significant decrease of complementarity or coherence at the beginning of the transition. According to Macedo and Martins, not all reform areas could be changed at the same time, so complementarity decreased. Once again, this is typically a second-best situation, which can entail loss of welfare. Therefore, this transitional cost should be reflected in income losses at the beginning of the transition, a theoretical intuition that is indeed verified in the authors' sample for the new EU members,7 as the relationship between the average level reforms RL and GDP growth shows an initial decline followed by an increase until the end of the policy cycle. In fact, GDP growth and RC have the same evolution: they both decrease at the beginning of the transition and increase in the latter stages of the policy cycle.

Chang, Kaltani, and Loayza (2005) studied how the effect of trade openness on economic growth depends on complementary reforms that help a country take advantage of international competition. The authors presented significant panel evidence, using a non-linear growth regression specification that interacts a proxy of trade openness with proxies of educational investment, financial depth, inflation stabilization, public infrastructure, governance, labor-market flexibility, ease of firm entry, and ease of firm exit. An interesting pattern of reform complementarity emerged: the coefficients on the interaction between the trade volume ratio and, in turn, the secondary enrollment rate, the private domestic credit ratio, and the number of phone lines per capita are positive and significant. This indicates that the growth effect of an increase in openness depends positively on the progress made in each of these areas. That is, more openness results in a larger increase in economic growth when the investment in human capital is stronger, financial markets are deeper, and public infrastructure is more readily available. The shared explanation for these results is related to the competitiveness of domestic firms in international markets: when domestic firms find a better educated labor force and less costly credit and communications, they are able to compete with foreign firms and expand their markets effectively.

The estimated coefficients on the interaction between the trade volume ratio and, in turn, the proxies for governance, labor-market flexibility, and firm-entry flexibility are also positive and statistically significant. The beneficial impact of an increase in trade openness on economic growth is larger when society has a more efficient, accountable, and honest government, and where the rule of law is more respected. Likewise, the positive growth effect of trade opening is stronger when flexible labor markets make it easier for domestic firms to transform and adjust to changing environments, particularly those in highly competitive foreign markets. The results also point out the importance of unrestricted firm renewal in order for trade opening to have a positive impact on growth.8

2.3 What Approach for Growth Strategies?

We have seen that, by simultaneously eliminating all distortions, welfare will increase unambiguously. The best possible economic growth rate is achieved by eliminating all obstacles that stand in its way. However, this is not realistic. Hausmann, Rodrik, and Velasco (2005) recalled that such a strategy requires not only having complete knowledge of all prevailing distortions, but that it also demands having the capacity to remove all the distortions in their entirety.

A second strategy would be to simply ignore the second-best theory and undertake whatever reforms seem to be feasible, practical, politically doable, or enforceable through conditionality. This is, according to Hausmann, Rodrik, and Velasco, the "do as much as you can, as best as you can" approach, which implicitly relies on the notions that: (i) any reform is good, (ii) the more areas reformed, the better, and (iii) the deeper the reform in any area, the better. In the words of Rodrik (2004: 6), that "opportunistic strategy may end up being targeted on areas of reform that are not particularly significant for economic growth at that point in time and that produce low economic returns." He also said that this strategy has been commonly accepted at international financial organizations, namely the World Bank.

If one wants to take into account economic theory and thus guarantee that partial reform will have good results, it is necessary to select those areas for which the second-best interactions across markets magnify the direct positive effects rather than weaken or reverse them. But, as Rodrik also points out, in any real economy, figuring out these interactions (and quantifying them) ex ante is extremely complicated.

Hausmann, Rodrik, and Velasco proposed an alternative and practical "diagnostic" approach—that is, a focus on the most-binding constraints. The best option would then be to focus on the reforms for which the direct effects can be expected to be large. This way, there will be less worry that second-best interactions will greatly diminish or possibly reverse the welfare effects. The elementary principle to follow is: "Go for the reforms that alleviate the most binding constraints and hence produce the biggest bang for the reform buck. Rather than utilize a spray-gun approach, in the hope that we will somehow hit the target, focus on the bottlenecks directly" (Hausmann, Rodrik, and Velasco 2005: 7). In practice, the authors' approach starts by focusing not on specific distortions (the full list of which is unknowable), but on the proximate determinants of economic growth (e.g., saving, investment, education, productivity, infrastructure, etc.). Once it is clear where to focus, then it is possible to look for associated economic distortions whose removal would make the largest contribution to alleviating the constraints on growth.

These authors have advanced in rejecting the widely applied orthodox paradigm based on one-size-fits-all policies. It is not exaggerated to say that their proposal constitutes a step further—maybe a decisive one—toward a much more open and realistic policymaking paradigm, which is to be based on country-specific solutions. Nevertheless, this is not to say that market-friendly policies or conventional solutions are not needed to ignite growth; it may be the case that a country needs a more orthodox policy agenda, as the same authors suggested for Brazil, or a combination of orthodox and unorthodox elements (India, Republic of Korea [hereafter Korea], and Viet Nam seem to be good examples of this).

While the authors' focus is on starting a growth process, sustaining it is another matter entirely. On recalling the "growth and crash" experience of the Dominican Republic, Rodrik (2004: 12) justly adds that "igniting growth may not require the full laundry list of reforms promoted, but sustaining it and endowing the economy with resilience to adverse shocks require addressing over time the institutional and governance constraints that will inevitably become more binding in a growing economy"; therefore sustaining growth is more difficult than getting it started (Rodrik 2003).

Even if the targeting of the binding constraints is well done, some, or many, distortions will remain. The distortions could be regarded as not being very important at first, simply because they are not visible. However, the evaluation of estimated welfare losses arising from the remaining distortions should be made with a medium- or long-term horizon, taking particular note of the risks that those distortions imply. A high growth can be ignited, but one should ask how sustainable that growth process is. For instance, when an economy is growing, not much attention is typically paid to the necessity of putting in place good bankruptcy laws and well-staffed bankruptcy courts, because neither is needed. In a high growth period, failures are not that frequent. But, in the event of a crisis—say, a financial panic (a non-innocent example in the context of this paper)—the absence of such a key reform would amplify and deepen the recession. Capital flights would be larger and more abrupt because creditors would know that they would not be able to recover their loans; on the other hand, the reallocation of resources from closed firms to new or more efficient companies would be prevented. In the crisis, the previously invisible distortion would reveal itself.

In addition, the conjugation or interaction of high growth with a set of remaining distortions can create new distortions. Imagine that a poor country managed to dramatically increase its growth rate in a short span of time. Imagine that its government's strategy was to augment its trade openness to take advantage of strong comparative advantages and growth in its richer neighboring countries. In such a case, the country would specialize in a labor-intensive, low-tech, fragmented, and light industry sector (thus limiting, in a decisive manner, its long-run growth prospects). As the working-age population got more jobs and higher wages, these people (and their families) would start to consume more (i.e., buying better houses and hi-tech products, traveling, etc.). However, as the financial system does not function well because it is laxly regulated, protected from external competition, and closely connected to the government, domestic banks have neither the capacity nor the incentive to respond correctly to the new demand for credit. Loans to consumers (as well as loans to firms) would grow—also for political convenience—without the banks taking into account that debtors rely on a salary which, in turn, depends essentially on the evolution of international prices of export goods with low added value. This would constitute a risky situation. This merely hypothetical example aims to show that a distortion which was inexistent or incipient before the growth take-off—in the sense that demand for credit was virtually inexistent—when stimulated by rapid growth and bad institutions or incentives, could emerge and expand to worrying proportions.

In short, at first, it is acceptable to put aside the "spray-gun approach" and focus on a given bottleneck in order to trigger growth. Thereafter, policymakers should think about making complementary reforms. The underlying idea is that there is a link between high coherence (or high complementarity) and sustainable growth.

The views expressed in this paper are the views of the authors and do not necessarily reflect the views or policies of the Asian Development Bank Institute (ADBI), the Asian Development Bank (ADB), its Board of Directors, or the governments they represent. ADBI does not guarantee the accuracy of the data included in this paper and accepts no responsibility for any consequences of their use. Terminology used may not necessarily be consistent with ADB official terms..

[previous chapter] [next chapter]

Post a Comment

We welcome your feedback on this publication. Post a comment. ADBI is not obliged to acknowledge or publish comments and may abridge or edit them before web posting.


There are [0] comment(s) for this entry. Post a comment.

    Back to Top 
    © 2015 Asian Development Bank Institute.