Many scholars (including me) have proclaimed the failure of a package of market‐oriented reforms proposed in the 1980s and 1990s—variously known as the Washington Consensus, the International Monetary Fund and World Bank structural adjustment, globalization, or neoliberalism. I seek to update the stylized facts on policies and growth that influenced this verdict. While I do not claim that there is any causal interpretation on policy reforms and growth, I argue that theories of policies and growth should at least seek to keep updating and explaining such stylized facts.
The earlier stylized facts featured the zero or low per capita growth in the regions that were the focus of reform: Africa and Latin America. This produced some strong conclusions on the Washington Consensus (in both the academic literature and the applied policy debates). For example, economist Dani Rodrik argued in a 2006 paper that “proponents and critics alike agree that the policies spawned by the Washington Consensus have not produced the desired results.… It is fair to say that nobody really believes in the Washington Consensus anymore. The debate now is not over whether the Washington Consensus is dead or alive, but over what will replace it.”
I also expressed doubts in 2005: “Repeated [structural] adjustment lending … fails to show any positive effect on policies or growth.” And I noted in another paper the Lost Decades (originally a decade of slowdown in Japan’s economy during the 1990s) for Latin America and Africa, a discouraging outcome of stagnation in spite of policy reform from 1980 to 1998. Rodrik declared that even the “most ardent supporters of reform now concede that growth has been below expectations in Latin America” and that “success stories in sub‐Saharan Africa [have been] few and far in between.” The World Bank noted that “some countries managed to sustain rapid growth with just modest reforms, and others could not grow even after implementing a wide range of reforms.”
The doubts about the Washington Consensus had begun even earlier. In 1997, Rodrik asked whether globalization had gone too far. And in 1995, Paul Krugman noted that “the real economic performance of countries that had recently adopted Washington consensus policies … was distinctly disappointing.” Since economist John H. Williamson had defined the Washington Consensus in 1990, many economists doubted it almost as soon as the first postreform numbers appeared.
More recently, Rodrik and economists Suresh Naidu and Gabriel Zucman called for a new economic order in “Economics after Neoliberalism,” a Boston Review article, because “many of the dominant policy ideas of the last few decades are supported neither by sound economics nor by good evidence.”
The literature on policies and growth has some well‐known shortcomings. It failed to resolve causality from macropolicies to growth or even to measure macropolicy effort directly as opposed to indirect and endogenous measures of policy outcomes. If it were possible to resolve these problems, the literature would have probably done so by now.
Yet the examples above show that the stylized facts on policy outcomes and growth—that poor growth outcomes accompany improved policy outcomes—influenced beliefs on the policy‐growth relationship. If so, it seems strange that these stylized facts have not been updated in the literature, as much more data have become available. Increased emphasis on resolving causality is welcome, but it should not prevent the updating of influential noncausal stylized facts. My purpose is to fill this gap in the literature and report new stylized facts.
First, my principal finding is that there has been additional and quite remarkable progress on reform outcomes since the late 1990s. Earlier judgments on the reforms often happened before the reform process was complete and/or had enough postreform growth data to evaluate reforms. This first stylized fact could also be consistent with an exogenous international trend toward reform, although again causality cannot be proven.
The second stylized fact is that there is a strong correlation between improvements in policy outcomes and changes in growth outcomes.
The third stylized fact is that growth has recovered in Africa and Latin America in the new millennium and that the regression of growth on policy outcomes explains a substantial part of the growth recovery.
One of my main contributions is the documenting of the both extremely bad and moderately bad policy outcomes that were surprisingly common in the 1980s and early 1990s; such outcomes have mostly disappeared. This policy revolution has received little attention in the previous literature. Explaining how this happened would be a fascinating topic for future research. There are many possible explanations—did the Washington Consensus advice have a delayed positive effect on reform after all? Did the human capital of policymakers increase over time so as to change destructive policies? The common trends could be consistent with an exogenous international trend toward reform, although this cannot be resolved definitively.
The results are also an interesting case study in the use of evidence in the political economy of reform. When new reforms are announced with as much fanfare as the Washington Consensus, there is pressure to evaluate the reforms as soon as possible. This can lead to what I document: premature pessimism about reform before the reform process is even complete and before enough postreform growth is available. Later results may show this pessimism to be mistaken, but there is much less interest in evaluating the reforms by that point. This may help explain why it is so difficult to do reforms and why corrections to extremely bad policy outcomes are delayed. Exploring such political‐economy outcomes would also be a fruitful topic for further research.
I show that these changes in policy outcomes—especially changes from extreme policies—were accompanied by growth increases. Policy reforms can explain the growth increases in Africa and Latin America, and old data available through 1998 were indeed consistent with the reform pessimism partly because of weaker results on growth payoffs associated with reform outcomes and partly because less reform had happened.
None of these statements resolve causality. But if the earlier stylized facts induced doubts about the value of reform, the new stylized facts should cause some updating of beliefs toward a more positive view of these policy reforms.
The new stylized facts seem most consistent with a position between complete dismissal and vindication of the Washington Consensus. Even if the new stylized facts were interpreted as causal, they would still hardly constitute a blanket triumph of the Washington Consensus. And the most robust results only come from the most extreme policy ranges. Even critics of the Washington Consensus might agree that extreme ranges of inflation, black market premiums, overvaluation, negative real interest rates, and repression of trade were undesirable. The finding that moderately bad policies are not very robust predictors of growth could even possibly support a criticism of the Washington Consensus that it was too obsessive about getting policies exactly right.
Despite these caveats, the new stylized facts are consistent with a more positive view of reform, compared to the previous consensus on doubting reform. The reform critics (including me) failed to emphasize the dangers of extreme policies in the previous reform literature or to note how common extreme policies were. Even if the reform movement was far from a complete shift to “free market policies,” it at least seems to have accomplished the elimination of the most extreme policy distortions of markets, which is associated with the revival of growth in African, Latin American, and other countries that had extreme policies.
This research brief is based on William Easterly, “In Search of Reforms for Growth: New Stylized Facts on Policy and Growth Outcomes,” NBER Working Paper no. 26318, September 2019, https://www.nber.org/papers/w26318.