Tuesday, September 24, 2013

Ignorance and bias in economic models

Greg Mankiw's latest post discusses the potential effects of increasing the minimum wage. He refers to a recent paper by Lee and Saez that argues that increasing the minimum wage can be optimal (for everyone not just those receiving a higher income). But according to Mankiw, the argument of Lee and Saez is based on a model that uses an implausible assumption that drives all the results. Given how unrealistic the assumption is we should dismiss the results.

I do not want to go into the details of the particular assumption of the Lee and Saez paper and how plausible it is (I have no views on it), but let me comment on the reliance of economic models on implausible assumptions.

All models rely on assumptions and economic models are known (and made fun of) for relying on very strong assumptions about rationality, perfect information,... Many of these assumptions are unrealistic but they are justified as a way to set a benchmark model around which one is then allowed to model deviations from the assumptions. The strategy of setting up a benchmark model might sound like a reasonable one as we need to start with a tractable view of the world before one can get "closer" to the complexities of the real world.

But the problem is that the model becomes (or has become) the reference in a way that sets a high burden of proof for any deviations from it. If you think individuals are not rational, go ahead and model their behavior but you should do it in a way that is realistic and backed by data (good luck). If you want to allow for government spending to be productive, you can do it but you better have perfect econometric results that prove that returns to government investment is indeed high. Given that experiments are not possible in economics, it will always be very difficult to produce supporting evidence for some of these assumptions that is not controversial. So what do we do? We claim that we do not know enough about the real world, we claim ignorance, and that would be ok except that we do not stay quiet when we are asked for our opinion. We go back to the results of our benchmark model and if someone asks us about a relevant policy question we use them to justify our answer. Do we care about the fact that the assumptions of that model have never been proven and have no connection to reality? No, we don't.

This subtle (or not so subtle) bias in economic analysis is my biggest source of frustration with my profession. Not being able to predict crisis, the stock market or exchange rates does not bother me, it is just a reflection of the limits of our knowledge and I can live with it. But using the same naive predictions of models that refer to a fictitious world as the reference and only moving away from them when someone produces an unquestionable piece of empirical evidence is in my mind the true cost of our profession to society.

Antonio Fatás

Thursday, September 19, 2013

Does competition get rid of waste in the private sector?

It is very common to hear comments about the waste of resources when referring to governments and the public sector. Paul Krugman does his best to argue against this popular view by showing that most of what government do is related to services that we demand and value as a society (it is not about hiring civil servants that produce no useful service). As he puts it, the government is an "insurance company with an army". But critics will argue that even if this is the case, the functioning of that (public) insurance company is extremely inefficient. In fact, we all have our list of anecdotes on how governments waste resources, build bridges to nowhere and how politicians are driven by their own interest, their ambitions or even worse pure corruption. If only we could bring the private sector to manage these services!

In addition to the anecdotal evidence there is something else that matters: we tend to use framework that starts with the assumption that in the private sector competition will get rid of waste. An inefficient company will be driven out of business by an efficient one. An inefficient and corrupt manager will be replaced by one who can get the work done. And we believe that the same does not apply to governments (yes, there are elections but they do not happen often enough plus there is no real competition there).

But is competition good enough to get rid of all the waste and inefficiencies in the private sector? I am sure there are many instances where this is the case but I am afraid there are also plenty of cases where competition is not strong enough. And just to be clear, I am not simply talking about large companies that abuse monopoly power, I am thinking of all the instances where the competitive threat is not enough to eliminate inefficiencies. 

I think this applies to financial institutions: the financial crisis has undermined our perception that these institutions were acting in the benefit of their shareholders (even Alan Greenspan said so). Profits and rents going to that sector (and to a reduced group of individuals) did not seem justified by the value they added. Why don't we see entry of new banks? Why don't we see new entrants taking over this market, finding the necessary funding to build scale and attracting the depositors or investors of the current established institutions? Why don't we see new CEOs rising to the leadership of these institutions with a platform that promises to behave in a different way and possibly be much better at maximizing the creation of long-term value? It must be that competition is not strong enough.

Same in the healthcare sector: How good is the private sector at managing healthcare? Are medical doctors truly competing with each other (in both quality and price)? Luckily we can find some data here from Uwe Reinhardt on excess costs in US health spending that include among other things $190 billion of excess administrative costs.

But this table looks to me like an exception, but not an exception in the sense that we do not find similar waste in other sectors, we simply do not know about it, we do not even attempt to measure it (at least at the macro level). And the reason why we do not bother measuring it is because we assume that markets and competition must make this number close enough to zero. Maybe it is time to challenge this assumption.

Antonio Fatás

Sunday, September 15, 2013

The only uncertainty is why some cannot see facts.

The idea that policy uncertainty is the main reason why advanced economies and Europe in particular are not recovering fast will not go away. Marco Buti and Pier Carlo Padoan in Vox bring back this argument in their attempt to figure out why the recovery in Europe is so weak. 

They explore the differences between the Eurozone and US recoveries to understand what are the factors that explain the divergence in performance ("the growth shortfall of the Eurozone compared to the US is striking"). So what are the differences between the two economic regions? Three according to the authors: policy uncertainty, weak financial system and lack of investment opportunities.

The evidence? 

1. Measures of uncertainty are correlated with growth. But as many others have argued, measures of uncertainty are (in most cases) endogenous to growth. 
2. Reform in the years 1998-2003 is a good predictor of the potential growth gains in the five years that followed (2003-2008). I am not sure how relevant this is to explain the post-2008 growth performance but let me argue that the chart presented in the blog post is not too convincing either. Here it is:

First, notice that their measure of growth is potential (not actual). Second, while the estimated regression suggests a positive correlation, when one looks at the Euro countries, there is no obvious correlation between the two variables. And if we compare some of them to the US (this was the comparison that the authors wanted to understand), the evidence runs contrary to their argument. Spain or Italy or France or even Greece improved their (product markets) regulation much more than the USA or Canada but their (potential) growth performance was significantly weaker.

So while talking about uncertainty and reform always sounds right, the evidence is either weak or inexistent.

But what about alternative explanations? Could it be that policy, in particular fiscal policy, can explain differences in growth performance since 2008? It must be that the evidence is not that strong, otherwise how could it be that is nor mentioned in their article?

Here is my attempt to provide an alternative explanation to the cross-variation in GDP growth rates.

I compare the change in real GDP in the 2008-2012 period with the change in real government consumption during the same years. This variable is an indication of the fiscal policy stance during those years. I include all OECD countries in the sample (removing some, including only the Euro countries makes no difference to the slope or fit of the regression). The correlation is very strong with a a coefficient that is not far from 1.

But isn't government consumption a component of GDP? Isn't the correlation always there? Yes and No. Yes, there is plenty of endogeneity in my chart and if you want to get the numbers right you need to read all the literature on fiscal policy multipliers (which will tell you that the effects are in fact stronger than what the casual analysis above shows). But, No, we also know that many of these countries implemented explicit cuts in government spending in response to changes in their interest rate and the financial markets reaction to sovereign debt crisis. So there is a significant exogenous component to government spending in that picture, enough to be used as an explanatory variable. And remember that there are those who believe that this correlation should be zero or even negative -- those who believe in zero or negative fiscal policy multipliers. So when Greece (bottom left dot) cut government consumption by 17%, they expected private spending to increase by at least the same amount. But it did not happen, private spending also fell and GDP has fallen by almost 25% during that period. 

The chart above is just a reminder of how strong the evidence is in favor of the hypothesis that fiscal policy can explain most of the relative underperformance of some economies in the post 2008 period. The evidence is certainly much stronger than any of the factors that Buti and Padoan present. But somehow fiscal policy did not make it to their list of top three factors.

Antonio Fatás

Thursday, September 5, 2013

The Euro counterfactual

Since the financial crisis started we have heard many commentators telling the Euro countries: "I told you so, this was a very bad idea". The argument is that the Euro area is not an optimal currency area - a jargon used by economists to argue that the costs of having a single currency are larger than its benefits. While until 2008 things have looked fine, the crisis is the real test for the Euro area and it has failed. And it has failed because of what any standard macroeconomics textbook tells you: that once you give up your exchange rate you lose a stabilization tool and when a crisis that is asymmetric in nature comes along you suffer a prolonged crisis as the only way out is to let prices and wages fall (internal devaluation), a painful and inefficient process.

In a recent post, Paul Krugman reminds us once again of these arguments by comparing Ireland during the current crisis to Thailand or Indonesia during the Asian crisis. His argument is that the Asian economies recovered quite fast from their crisis while Ireland has not (and Greece has not even started any recovery). As Kevin O'Rourke puts it, Ireland looks like Thailand without the Baht.

The arguments seem solid and the evidence strong but I am somehow skeptical that we can that quickly conclude that the Euro was a failed experiment and that life without the Euro would have been better (and maybe I am reading too much into those posts and they are not really going that far in their statements). 

What one wants to do is build a counterfactual: where would Greece or Spain or Ireland be if they had never joined the Euro? What would their currency have done for them before and after the 2008 crisis? Unfortunately we cannot build such counterfactual so the best we can do is to look for similar examples (such as Thailand during the Asian crisis). But let me argue that if one extends the set of examples and anecdotes some of the data does not speak that clearly against the Euro.

Before I start let me make two points:

1. I have no disagreement with the argument that rigidities in prices and wages make an internal devaluation a painful way to get out of a recession. A depreciation is a much faster way to reset relative prices. This is what I teach to my students.

2. I agree that the Euro area comprises a set of countries that are performing way below their potential. A combination of failed reforms, lack of economic leadership, a not very proactive central bank is producing a growth rate that is below potential (both before and during the crisis). 

The point that I am trying to make is that among the list of problems that the Euro countries have, the Euro itself might not be the biggest one. While a devaluation could have helped established a faster recovery, its effects would have been uncertain and possibly small. I reach that conclusion by looking at data the same way Krugman and O'Rourke do. It just happens to be that I look at different data. Are my comparisons much better than theirs? Not sure (this is just a blog post, not a research paper). But I find them more relevant even if I have to admit that we have a great deal of uncertainty here, that our knowledge about the potential performance of countries with flexible versus fixed exchange rates is limited. And by knowledge I mean empirical not theoretical. We understand very well how exchange rates work but my reading of the literature is that we are not very good at quantifying these effects.

Below is a list of random and unconnected empirical facts that suggest that the Euro itself might not have caused as much damage as the comparison between Thailand and Indonesia to Ireland and Greece might suggest. 

1. Comparing across countries with different GDP per capita is tricky. It is hard to imagine Ireland to grow at rates similar to Thailand and Indonesia at any point in time after 2008 given that Irish GDP per capita is as high as that of Germany or the US (in 2012 and according to the IMF GDP per capita in Ireland is higher than in Germany; In contrast, in 1996 the GDP per capita of Indonesia was less than 10% of the US level). Finding a better comparison is difficult. Very few advanced economies adopt fixed exchange rates or go through large devaluations. But we can still find some anecdotal evidence.

2. The European countries were in a crisis also during the years 1991-1993. A crisis that brought their system of fixed exchange rates down. Some countries abandoned the fixed exchange rate and devalued while others stayed. This is the closest we have to an experiment of countries leaving the Euro area (yes, the experiment is not perfect, the crisis was much smaller, but it can help us understand the role of exchange rates). Below is a chart with the path of GDP that different countries followed.

There are 9 countries in the chart. Some of them let their currencies be devalued (and a subset of those completely left the system). These are represented by dotted lines. The others stayed within the system and their value of their currencies remained fixed to each other (they are represented by the solid line). The chart reveals that there is no clear pattern between the two groups. Growth differences do not seem to correlate well with the behavior of exchange rates. It might be that only those who devalue "had to devalue", possibly, but what is remarkable is how similar the growth patterns are for all countries.

3. Here is a second piece of evidence from the most recent crisis. Both Spain and the UK have suffered the consequences of the 2008 financial crisis. Both are large economies that before the crisis saw real estate prices booming. But once the crisis started one them let its currency depreciate (the UK) by about 30% relative to the Euro while the other (Spain) was stuck with a currency that it did not own. The theory says that the UK should have benefited from a strong boost in exports as a result of the depreciation of its currency. Below is a comparison of exports (real) after the crisis erupted at the end of 2007.

Contrary to our priors, Spanish exports have grown faster than UK exports. So it seems that the depreciation did not help that much the UK (or the lack of control of your currency did not hurt Spain that much either).

4. So maybe exports in Spain did not behave that badly but isn't unemployment extremely high, at levels that are as high as during the Great Depression in the US? Yes, they are, but unemployment rates in Spain were also extremely high when Spain had a currency (the peseta).

Employment Growth
Unemployment Rate

Above is a table that summarizes the Spanish labor market before and after the Euro. In the pre-Euro years unemployment was on average 18.9%. After the Euro, it has been "only" 14%. Employment growth before the Euro was 0.81%, after the Euro 1.45%.

None of the above facts provide a perfect test of what life without the Euro would have been for some of the Euro countries but they, at a minimum, question the conclusion that the recent financial crisis has clearly proven that the creation of a single currency in Europe was a really bad idea. For some of these countries life is hard and volatile (with or without the Euro).

Antonio Fatás