Tuesday, February 14, 2017

Mark Weisbrot and the debacle of heterodox economics: "Ecuador’s Left-Wing Success Story"

This week, I ran across an article by Mark Weisbrot, Jake Johnston, and Lara Merling titled "Decade of Reform: Ecuador's Macroeconomic Policies, Institutional Changes, and Results."  I also read Mark's article titled "Ecuador’s Left-Wing Success Story" on The Nation.  This post reflects on the main conclusion of these articles: That the economic policies of Rafael Correa have been successful to create a prosperous Ecuador.

The message of this post is very simple: Without properly discriminating the causes of what the authors call "success", or the effects of the economic policies implemented by the government of Rafael Correa, the descriptive analysis they present is just that, a simple description of the economy that in no way sheds light on the consequences of the policies carried out.  Therefore, it would be economically (and commonsensically) meaningless, and even irresponsible, to recommend these policies to a country, including, of course, Ecuador.

I start by reproducing the points of the executive summary and debunk them.  Then, I show the lack of scientific rigor in some of the statements made by the authors.

The literal executive summary is as follows (my comments appear in blue):
  • Annual per capita GDP growth during the past decade (2006–16) was 1.5 percent, as compared to 0.6 percent over the prior 26 years. This is a significant improvement, despite the fact that the economy was hit by major external economic shocks.
    • First, it is very arbitrary to take the prior 26 years, especially because the economy before dollarization (year 2000) was very different to the economy afterwards.  Also, there are 11 years in the period "2006-16", and one of them does not belong to the government of Rafael Correa: 2006. The correct comparison is between the years 2000-2006 and the years 2007-2016, although, for 2016 there is still incomplete information: there is no GDP estimate yet, for example.
    • Second, let us take the per capita GDP in constant dollars adjusted by purchasing power parity of the World Bank.  The average growth rate of per capita GDP between 1990 (there is no information prior to this year) and 2006 is about 1% per year, whereas it is about 2.3% between 2007 and 2015, which is higher than the period before, as the authors state.
    • However, if one considers the growth rates between 2000 and 2006 (the post-dollarization period) of the indicator of per capita GDP above, the average is about 3%, which is higher than the average growth rate under Rafael Correa.
    • Fourth, in the same period, 2007-2015, Chile's average growth rate is 2.5%, Colombia's average is 3.2%, while Peru's is 4.3%, all of them higher than Ecuador's.
  • The poverty rate declined by 38 percent, and extreme poverty by 47 percent. Much of the decline in poverty was a result of economic growth and employment, but some was also a result of government programs that helped poor people, such as the cash transfer program Bono de Desarollo Humano, which more than doubled in size as a percent of GDP.  
    • The poverty rate between 2007 and 2015 declined from 37.6% to 23.3%, about 14 percentage points or 38%. But the poverty rate declined from 64.4% to 37.6%, about 27 percentage points or 42%, between 2000 and 2006.
    • If we consider the headcount poverty ratio in constant dollars adjusted by purchasing power parity of the World Bank, Ecuador reduces its poverty rate by 62% between 2007 and 2015, whereas Peru reduces it by 69%.  Colombia is less successful (45%), whereas Chile basically halves it, though to a level that is about one third that of Ecuador.
  • The reduction in poverty was many times larger than that of the previous decade.
    • This is just false. First, reliable poverty rates going back to, say, 1996 or 1997 (the "previous decade") are very scarce.  The World Bank provides scattered information about poverty rates based on constant dollars adjusted by purchasing power parity.  Those show that the reduction in poverty between 1994 and 2006 is about 46%, which is not "many times" smaller.  Again, the same measure shows that between 2000 and 2006 (the post dollarization period), poverty was reduced by 67%, higher than the 62% achieved in 9 years of the government of Rafael Correa.
  • Inequality also fell substantially, as measured by the Gini coefficient (from 0.55 to 0.47), or by the ratio of the top 10 percent to the bottom 10 percent of the income distribution (from 36 to 25, as of 2012). 
    • This is just true and nothing can be refuted.
  • The government doubled social spending, as a percentage of GDP, from 4.3 percent in 2006 to 8.6 percent in 2016. This included large increases in spending on education, health, and urban development and housing.
  • There were significant gains in enrollment at various levels of education. Spending on higher education increased from 0.7 to 2.1 percent of GDP; this is the highest level of government spending on higher education in Latin America, and higher than the average of the OECD countries.
  • Government expenditure on health services doubled as a percentage of GDP from 2006 to 2016.
    • I would like to answer these three claims at once.  First, the only comparable information provided by the government exists from 2008.
    • Second, the information shows that about 3% of GDP was dedicated to public education and 1.4% to public health in 2008, for a total of about 4.5% of GDP.  Those numbers in 2016 are 4.5% for public education and 2.5% for public health, for a total of about 7% of GDP.  This is an increase of about 36% in education and health services.
    • Third, if you consider the percentage of GDP that the government dedicates to repay the public debt (capital, interests, and anticipated sales of crude oil), you will find that the values are 5.1% in 2008 and 8.5% in 2016, an increase of 67%, much larger than the increase in education and health.
    • Lastly, it can be very easy to increase spending, especially when the costs of the projects are not competitive and contaminated with corruption.
  • Public investment increased from 4 percent of GDP in 2006 to 14.8 percent in 2013, before falling to about 10 percent of GDP in 2016.
    • This is true, but says nothing.  It is only a report of expenses without an analysis on the return of this investment, as any investment project requires.  This poor way of analyzing the economy leads me to my second set of criticisms:  the lack of economics in this report. 
Before moving to the next point, let me provide an additional measure that the authors miss: the human development index, which is a much broader indicator of the development of a nation.  It has increased about 5% between 2005 and 2014, the same proportional increase than between 1995 and 2005.

In what follows, I reproduce some of the statements of the authors that bear no single cause-effect analysis, as any decent recommendation of economic policy should supply.  In economics, one needs to provide the effect or consequence of a given economic policy recommendation.  There are several ways of doing this.  One can use a model to simulate a policy, for example, or use empirical methods to isolate the effect of a policy on the outcome(s) of interest (regression analysis, synthetic control methods, randomized control trials, etc.)  The economic science has tried to make progress along these lines with researchers devoting their lives to come up with relevant methods to fulfill the many request serious policy makers make.  The authors of this report resort to none, just as the government of Rafael Correa has done since its inception.

These are some of their unfunded claims (some of them come from the article in The Nation):
  • Much of the decline in poverty was a result of economic growth and employment, but some was also a result of government programs that helped poor people, such as the cash transfer program Bono de Desarollo Humano, which more than doubled in size as a percent of GDP.
    • How did they measure the effect of the policies or programs on the reduction of poverty?
  • Over the last decade, the Ecuadorian government has instituted a series of financial and regulatory reforms that proved to be important when the economy was hit with severe external shocks.
    • How did they determine that the financial and regulatory reforms were important? How important? (A variance decomposition would be useful here.)
  • Tariffs on imports, imposed under the balance of payments safeguard rules of the WTO, provided a substantial stimulus to the economy over the last two years.
  • Ecuador is a good example of how a left government achieved success over the past decade through positive and creative changes in economic policy.
    • How is "success" measured here?  The macroeconomic results are not definite, as I have shown above.  Besides, I encourage the authors to incorporate in the analysis a measure of sustainability of what they call success.
  • Ecuador’s central bank created billions of dollars that it lent to the government for spending (and also to state-owned banks).
    • This is just false.  The central bank did not create a penny (well, there is domestic currency, but is minimal).  What the central bank did was to take the money in its vaults from private and public depositors and lent it to the government.  Currently, the debt of the government to the central bank reaches about 5% of GDP.
There are economists in Ecuador evaluating what has been these 10 years of economic policies.  This blog reproduces several of those analysis.  The topics cover:
  • Analysis of the impact of fiscal policy on the economy.  This is a fiscal VAR where identification is achieved with a variation of the Blanchard-Perotti methodology to include oil revenues.  Results show that public investment is not as beneficial to the economy as the government claims, once one discounts the effect of oil revenues.  Higher future taxes and uncertainty can be a harbinger to growth.
  • Analysis of the sustainability of the fiscal accounts.  This is an analysis of cointegration that incorporates a break in the cointegrating relationship.  Results show that the fiscal accounts are no longer sustainable.
  • Synthetic control method to analyze the effect of the economic policies implemented by Rafael Correa on the per capita GDP of Ecuador.  The results show that the policies were, if not detrimental, at the most neutral.
Again, an economist should provide an analysis of this nature to provide recommendations or to evaluate an economic proposal.  It is the debacle of a serious economist not to do it.  I thought Jayati Ghosh was the only one who could have done that. Update 2/15/17: I was forgetting Bill Black and his article, which was also debunked in this blog.