Is Pro-Business Reform Pro-Growth?
John Cochrane and Brad DeLong are clashing over whether pro-business reforms, like making it easier for entrepreneurs to start businesses or register property, raise living standards. Using data from the World Bank's Doing Business Index, Cochrane claims that reforms that bring the U.S. to the frontier — defined as a score of 100 out of 100 from the World Bank's assessors — could double per-capita income. DeLong challenges these conclusions, pointing out they rely on a rather extreme extrapolation of a log-linear model and assertions that the relationship between ease of doing business and income is causal.
You can see the essence of the controversy in these dueling graphs below. The first, Cochrane's, makes the large growth claims look somewhat reasonable; the second, DeLong's, makes them look ludicrous. The only difference is the scale of the y-axis: Cochrane uses a logarithmic scale, and DeLong does not.
This isn't exactly a productive debate. Extrapolation from the cross-section of countries will tell us almost nothing useful. There are roughly a million reasons why that will fail, not least of which is the complexity of much-deeper institutions, like the strength of rule of law, that play a central role in determining both living standards and public policy towards businesses.
Instead, we should be looking at the many countries that have implemented large pro-business reforms and see if their per-capita incomes rise relative to similar countries afterwards. If we want to anticipate the likely effects of a large pro-business reform, it is hard to do better than study countries that have actually tried them.
To the best of my knowledge, there's no event dataset that lists the names and years of countries that implement large pro-business reforms, as there is for fiscal consolidations or monetary tightenings. So I came up a simple trick to identify reforms directly from the World Bank data, adapted from a strategy in my undergrad thesis on Spain's housing boom and bust.
Consider how the ease of starting a business has evolved in Albania from 2004 to 2016. Notice the big jump the series (the solid blue line) from 2008 to 2009, a 15-point increase in that index. We'll call this a "reform," marked by the dashed gray vertical line below.
Was there an actual policy reform in Albania in 2008, or are we just picking up unimportant variation in the index? From the World Bank's 2009 report:
A new company law strengthened the protection of minority shareholder rights. The law tightened approval and disclosure requirements for related-party transactions and, for the first time, defined directors' duties. It also introduced greater remedies to pursue if a related-party transaction is harmful to the company. Albania made start-up easier by taking commercial registration out of the court and creating a one-stop shop. Companies can now start a business in 8 days-it used to take more than a month. The country's first credit registry opened for business. And tax reforms halved the corporate income tax rate to 10%.
So it does seem that "big" jumps in the World Bank's index correspond to actual reform programs. To measure the effect of this reform, we'll first need a counterfactual, an understanding of what would have happened to Albanian business law absent the big-bang 2008 reform. I'll make a simple assumption, which we will back up with evidence later, that the answer is "nothing" — that is, (1) the index would have not substantially changed in the reform year and (2) large reforms do not alter the future path of the index. You can see this counterfactual represented by the dashed blue line.
More formally, for country i in year t, I construct a measure of the cumulative reforms as:
which is to say that I calculate the rolling sum of all the increases in the index that are larger than some threshold kappa, which I define as the 95th percentile of all annual changes in the index. (I obtained similar results with alternative choices of the threshold.) Effectively, this decomposes every country's index into a step-function component and a drift component:
This is, I think, a reasonable way of doing things: Even if you are distrustful of the index, as am I, if the World Bank says that your country is in the top 5 percent of reformers in some year, there's probably something to that. In my sample, it took at least a 10-point increase in the ease of starting a business to qualify as a "reform" year. That is like going from India to China.
You can see the distribution of changes to the index below. Most countries see little to no change in their index value in most years, hence the concentration at zero. Some countries see their index values drop somewhat, but what you can really notice is a long right tail. All of the observations to the right of the dashed gray line at 10 points count as "reform" years for that country.
According to this measure, 36 percent of countries had at least one reform in my sample, the average reform entailed an 18-point increase in the index, and there are 109 reform years across countries. It turns out that these reforms are quantitatively important to explaining the overall time trend in the global-average ease of starting a business since 2004. In particular, these reforms explain half of all the increase.
Also, these reforms tended to be more important than drift for lower-income countries than for higher-income countries, respectively providing 55 percent and 38 percent of total gains in the ease of starting a business since 2004. You can also see, in the graph below, that lower-income countries reformed more during this period than higher-income countries. (I use the World Bank's designation of income groups.)
I claimed earlier that reforms don't seem to affect a country's index trend, other than through the jump. This is important to check because it means, on average, reforms are "permanent," and there is no positive feedback or backsliding. Evidence for this claim comes from a regression of the index on lagged values of the index and of the reform variable:
where k is the number of lags we include; I chose five. From this regression, we can estimate the effects of a permanent 10-point reform on the World Bank's index and compare it against what it would look like if we mechanically added 10 points to the index at the reform date:
Clearly, the estimated response looks like mechanical one. Since the evidence is supportive, for transparency and simplicity, I will go with my approach of pulling out large increases in the index and cumulating them over time.
Reforms and growth
The question that ultimately interests me, Cochrane, and DeLong is whether pro-business reforms are pro-growth. It is worth spending a moment, then, on how I measure economic growth.
I am using data from the Penn World Table on output-side real GDP per capita at chained PPPs; this is the correct series, according to the source, to "compare relative productive capacity across countries and over time," which is what we'd like to do. (The Penn World Table is the canonical source for comparable cross-country output data.) We can match 128 countries across both datasets, which is about half of the sample. The latest batch of data covers up to 2012, and the World Bank data begins in 2004, so my analysis is limited to that time period. (I'll discuss the limitations of the data more fully as we go on.)
How does Cochrane estimate the effect of pro-business reforms on output? His picture at the top of this post shows the regression line of the equation:
where Y is per-capita output for country i in his observation year T. As I said before, there's no reason to believe that the coefficient beta identifies the true effect of pro-business reforms. Burkina Faso is not many times poorer than the U.S. only because it is many times easier to do business in the U.S. than in Burkina Faso.
Instead, to identify the causal effect, we should consider whether a country's per-capita output rises after a reform relative to similar countries in the same year. More formally, we can express that as:
where Y is again per-capita output, and notice that the decomposition of the World Bank's index into a reform component and a drift component can be dropped right into the regression equation. Including the drift component helps make the comparison between countries that had like policy paths.
What I find is that neither term has a significant coefficient. In fact, I can bound the effect of pro-business reforms quite precisely around zero, with a 95-percent confidence interval for the effect of a 10-point reform on the level of per-capita output of -1.4 percent to 3.5 percent. That is far away from the claim that such a reform could double per-capita output.
These results are not changed when I set up the analysis differently. For example, it's reasonable to think that the effect of reforms on output lags the reform, but these results are unchanged when I include lags of the reform and drift components. You might also guess that there is an interesting "pre-trend" in per-capita output before reforms: Perhaps governments wait until booms or busts, or something like that, to institute reforms. This actually isn't true — per-capita output evolves similarly in countries in the five years preceding a reform than in countries that don't reform in that period — but I can control for several lags per-capita output, and this doesn't change the result. Another claim one might make is that the effects of reform are being obscured by differences in the path of output across regions or income groups (e.g., low-income countries). For example, maybe countries tend to initiate reforms when regional growth is particularly strong or weak. This also isn't true — but I can control for it by interacting region or income-group dummy variables with the year dummies, which constructs region- or group-specific time paths for income. What these tests show is that my results aren't unduly sensitive to the manner in which I measure the effect of pro-business reforms.
Another technical way to do this, which I won't go into here, is to use my reform measure as an instrument for the World Bank's index, and do two-stage least squares to estimate the effect of a large pro-business reform. The estimated effect is nearly the same as above.
What this means for economic development
The issue of pro-business reform is most relevant to developing countries with weak institutions, who face meaningful pressure from international organizations like the World Bank to implement them. Yet the data give little reason to believe that countries can raise per-capita output in any notable way by making it easier for entrepreneurs to do business. Pro-business reforms probably have very limited effects, if any, on economic growth within a reasonable time horizon.
There still may be compelling reasons to do them. Maybe the most compelling is the moral case: The freedom to start and operate a business without undue government intrusion should, I think, be seen as a basic and essential freedom for individuals to make for themselves lives that they judge meaningful, one that is as important as freedoms of speech or assembly. The fact that these reforms are economically neutral does not mean they are morally neutral. More prosaically, there are important limits to my empirical analysis. If the impacts of these reforms only begin to manifest themselves more than five years later, for instance, I won't be able to see that. Measuring longer-run effects will have to wait for more data.
It might at first be puzzling why, given such a robust cross-sectional relationship being the World Bank's index and per-capita output, that I find no meaningful effects of pro-business reforms. To me, the most reasonable explanation is that the raw cross-country differences in the index probably captures a more meaningful measure of institutions than the jumps in the index, which reflect specific, discrete policy reforms. Countries with illiberal policies towards business probably have other problems that restrain economic development, but countries that change those policies in a sharply pro-business direction don't necessarily solve all those deeper problems in the same sweep.
Maybe the lesson here is to beware the TED-talk version of development economics. Shortening the time it takes to incorporate a small business is not a substitute for deeper institutional reforms, such as those that support investment in human and physical capital, remove economic barriers that hold back women and ethnic or religious minorities, or improve transportation, power, and sanitation infrastructure. Easy pro-business reforms should not distract countries from pursuing changes that, while harder to make, we know to be richly rewarding in the long run.