The Great Reversal Page 11
“But then, the competition from abroad that Whirlpool had relied on to get the merger approved became a little bit too tough to handle,” as Soumaya Keynes explains. Whirlpool lost market share to LG and Samsung and petitioned for safeguard tariffs. When the government granted the request, the stock price of Whirlpool went up and LG announced it would raise the price of its washing machines. In April 2018, “the price of washing machines increased by 9 percent. The next month they increased by 6 percent. Both are the largest monthly price increases since the Bureau of Labor Statistics began collecting such data in 1977.”g This example shows the danger of relying on foreign competition to discipline domestic firms.
The Failure of Free Entry
Let us finally combine the insights from Chapter 4 on investment with those of this chapter on entry. We have described the fundamental law of investment. When q is above 1 in an industry, it means that there are rents left on the table. If the industry is competitive, these rents should be competed away: either incumbents expand (as in our example) or new firms enter.
The fundamental law of investment therefore begets a fundamental law of entry. As industries adapt to various economic shocks, some become more profitable and some less profitable. Economic efficiency requires exit from less profitable industries and entry into more profitable ones. This naturally leads to a q-theory of entry, similar to that for investment. Just as scaling up a high-q firm generates economic value, reallocating firms from low- to high-q industries also generates value. Gutiérrez and I have studied this idea and concluded that free entry has been failing in recent years.
Figure 5.7 shows the elasticity of the number of firms to the industry-median q over the past forty years. This elasticity used to be around 0.5: when the median value of q in a particular industry increased by 0.1 (say from 1.1 to 1.2), the standardized change in the number of firms would be 5 percent higher over the following three years relative to other industries. Firms used to enter more and exit less in industries with larger values of Tobin’s q, exactly as free entry would predict. In recent years, however, this elasticity has been close to zero. The decline is consistent across data sources and is stronger outside manufacturing.
What explains the failure of free entry? Several factors have probably contributed, but one is central to the theme of this book. Figure 5.8 shows the rise in federal regulations in the US along with the decline in the firm entry rate. As emphasized by University of Chicago economist Steven Davis (2017), the Code of Federal Regulations has grown eightfold over the past fifty-six years and now consumes nearly 180,000 pages.
FIGURE 5.7 Declining allocation of entry to high-value industries. The figure plots the coefficient of year-by-year regressions of changes in the log-number of firms / establishments on the industry-median Tobin’s q. Data sources: Compustat and SUSB series based on the number of firms by NAICS level 4 industry. QCEW series based on the number of establishments by SIC level 3 industry up to 1997 and NAICS level 4 industries afterward. Changes in the number of firms are standardized to have mean zero and variance of one to ensure comparability across data sources. Industry-median q is based on Compustat. See Gutiérrez and Philippon (2019b) for details.
FIGURE 5.8 Regulation index and establishment birth rate. Data sources: Establishment entry rates from Census’ Business Dynamics Statistics. Regulatory restrictions from RegData. See Gutiérrez and Philippon (2019b) for details.
We would like to study if and how federal regulations affect industry dynamics. For that we need to build an index of regulations. How does one go about building an index of federal regulations? By using computers to read and classify the data! RegData is a relatively new database—introduced in Al-Ubaydli and McLaughlin (2017)—that aims to measure regulatory stringency at the industry level. It relies on machine learning and natural language processing techniques to count the number of restrictive words or phrases such as “shall,” “must,” and “may not” in each section of the Code of Federal Regulations and to assign them to industries. RegData represents a vast improvement over a simple measure of page counts.h
Figure 5.8 shows that the decline in entry coincided with the rise of entry regulations, but this does not mean that regulations caused the decline in entry. There are two fundamental theories of regulations: public interest versus public choice. Following the work of English economist Arthur Cecil Pigou (1932), the public interest theory emphasizes corrective regulations to deal with externalities and protect consumers. On the other side are public choice theorists who are suspicious of Pigou’s ideas. Famed Chicago economist George Stigler (1971) argues that “as a rule, regulation is acquired by the industry and is designed and operated primarily for its benefit.” In our paper, Gutiérrez and I use industry and firm-level data to dig deeper into this issue. We find that regulations drive down the entry and growth of small firms relative to large ones, particularly in industries with high lobbying expenditures. This supports public choice theory over the Pigouvian public interest theory and brings another piece of evidence to support the decreasing domestic competition hypothesis.
Concentration from the Top Down and from the Bottom Up
Two main facts emerge when we consider the shifting demographics of US businesses. First, the entry rate of new businesses has declined. Businesses are now older and face fewer new competitors each year. This has led to concentration from the bottom up. Second, agencies and judges have allowed more frequent mergers among large businesses. This has led to concentration from the top down. Together, they account for the rise in concentration that we have observed.
Free entry is a critically important rebalancing mechanism at the heart of market economies. Unfortunately, free-entry rebalancing has diminished in the US economy over the past twenty years. It’s not just that fewer firms appear each year; it’s also that they do not enter in high-q industries as much as they used to. It also appears that lobbying and regulations explain much of the decline in entry rates over time and across industries. At this point, we have enough evidence to suggest that the rise in profits and concentration in the US reflects a significant increase in rents, but we do not have enough evidence to quantify the harm done to consumers and workers. It’s time to broaden our analysis and study what has been happening in the rest of the world.
Our focus so far has been almost exclusively on the US. I will now show you that much can be learned if we compare the US with other regions—Europe in particular.
* * *
a “Too much of a good thing,” Economist, March 26, 2016.
b René Stultz, “The shrinking universe of public firms,” NBER Reporter 2 (2018).
c The DoJ is an element of the executive branch, whereas the FTC is a commission made up of presidential appointees from both major political parties. Although their responsibilities overlap somewhat, they work cooperatively and tend to divide their attention in predictable ways. DoJ focuses on financial services, telecommunications, and agriculture. The FTC generally takes the lead in cases involving the defense industry, pharmaceuticals, and retail. State attorneys general and private lawsuits can also challenge potentially anticompetitive behavior.
d One of the most significant ways in which Section 7 of the Clayton Act changed US antitrust policy was by creating a lower standard of proof for anticompetitive effects than the Sherman Act required. Where the Sherman Act required proof that a company had been harmed by anticompetitive practices, Section 7 allowed the government to block mergers when “the trend to a lessening of competition in a line of commerce was still in its incipiency.” Congress was forced to amend Section 7 in the 1950s to close a loophole that had allowed companies to exploit the original law’s description of mergers as purchases of “stock.” The amendment made companies seeking to effectively merge through asset purchases subject to the law.
e To determine the relevant product market, imagine a “small but significant and non-transitory” increase in price. If such an increase wou
ld cause buyers to shift to other products and would thus be unprofitable for the monopolist, we expand the market to include the closest substitutes until we find products for which the increase would be profitable. We add next-best substitutes until there are no practicable substitutes to which the consumer may shift. At this point, the product market is defined. To define the area, if a buyer of the product could respond to the price increase by purchasing outside the area, then the area is too narrow. We expand it until we have an area where the seller could maximize profits by increasing the price.
To determine ease of entry, the agencies analyze the timeliness (less than two years to plan entry and have a significant market impact), likelihood (profitability under premerger prices), and sufficiency (adequate knowledge of the market and the financial resources to withstand supracompetitive pricing by a merged firm).
f Vita and Osinski (2018) offer a rebuttal, while Kwoka (2017a) maintains the validity of his original critique.
g B. R. Mayes, T. Mellnik, K. Rabinowitz, and S. Tan, “Trump’s trade war has started. Who’s been helped and who’s been hurt?” Washington Post, July 2018.
h Goldschlag and Tabarrok (2018) provide a detailed discussion of the database and its limitations, including several validation analyses that, for example, compare RegData’s measure of regulatory stringency to the size of relevant regulatory agencies and the employment share of lawyers in each industry. They conclude that “the relative values of the regulatory stringency index capture well the differences in regulation over time, across industries, and across agencies.”
[ TWO ]
THE EUROPEAN EXPERIENCE
We have analyzed the evolution of the US economy over the past twenty years. We have formulated and tested various theories. The theory of “star” firms argues that concentration reflects the increasing productivity of industry leaders. The intangible hypothesis argues that the accumulation of intangible assets explains the evolution of concentration, profits, and investment. The decreasing domestic competition theory argues that domestic competition has declined and that, in many industries, firms have been able to exploit their market power. Globalization, on the other hand, has brought foreign competition into some manufacturing industries.
Our detailed analysis of these theories has allowed us to refine our diagnosis of the US economy. We have found evidence of “star” effects during the 1990s, but not in the 2000s. The intangible hypothesis is clearly relevant for the retail and wholesale trade sectors, and globalization is a major force shaping the manufacturing sector.
The overall evidence, however, is that most industries have suffered from weak and declining domestic competition over the past twenty years. Decreasing competition has led to increasing concentration, increasing entrenchment of industry leaders, increasing profits and payouts to shareholders, decreasing investment, and decreasing productivity growth.
At this point, we would like to understand why and how this has happened. Is it because of technology? Is it because of changes in consumers’ preferences? Or is it because of regulations and policy choices?
We already have a hint that policy choices are important. Lobbying and regulations predict the decline in entry rates over time and across industries, but we do not have a perfect controlled experiment. The ideal experiment would be to compare similar industries in different regulatory environments. This ideal experiment does not exist, but it turns out that we can go a long way by comparing the United States and Europe. This is what the following chapters will do.
Europe provides a striking comparison to the US, but before diving into policy differences between the US and the EU, it is useful to get one issue out of the way. I do not claim that Europe as a whole is doing better than the US, or even that it is doing particularly well. The rise of populism and the growing distrust of established parties and institutions are similar in the two regions. The macroeconomic architecture of the euro area is still incomplete and much less stable than that of the US. European universities still lag behind American universities—there is a reason I am writing this book in New York and not in Paris. European financial markets do not offer the same growth opportunities to ambitious new companies as American markets do. Europe also lags behind the US and China in some new technologies, most notably in artificial intelligence.
If we take a step back, however, we see that the similarities outnumber the differences, especially as far as economic development is concerned. The two economies are roughly the same size. Consumers have roughly the same tastes and buy essentially the same products. Firms use similar technologies in most industries, and identical ones in many. American and European trade patterns are also similar. Similarities in these key dimensions make the comparison between the US and Europe particularly relevant.
Finally, it is not a coincidence that the comparison of Europe and the US will prove to be rich and informative. At least since the end of World War II, European policy makers and entrepreneurs have been inspired by what they saw (and admired) in the US. As we shall discuss, several European institutions were either modeled after the corresponding American institutions or at least heavily influenced by their design.
CHAPTER 6
Meanwhile, in Europe
IN THE PREVIOUS chapters we have shown that, since 2000, US industries have become more concentrated and American firms’ profit margins have increased. At the same time, investment has been weak, despite high profit margins and low funding costs. Is this a universal evolution? Is it an unavoidable consequence of globalization? Do we see the same trends in all countries?
I have just explained why there is much to learn by comparing Europe and the US. The two regions are similar enough to make the comparison meaningful but different enough to provide a fertile ground for testing our theories. So, how has Europe been doing?
Is Europe Growing?
The results that I am going to present are usually greeted with a fair amount of skepticism. Some of it is warranted—indeed, I was also skeptical at first. The US has traditionally had better economic policies than Europe. US markets were indeed the most competitive in the world for a long time. It is reasonable to remain attached to this prior assumption. We should ask for more evidence before we change our mind, and this is what we are doing in this book.
In discussing these results, however, I have noticed some unusually defensive reactions, suggesting that some people are simply unwilling to change their minds. Here’s one argument I hear frequently: “If you are right, why is the US growing faster than Europe?” Well, to start with, growth obviously depends on many confounding factors. It is possible for Europe to have at the same time good antitrust and data protection policies, inefficient macroeconomic policies, and weak and badly regulated universities. As it happens, I believe all these statements to be correct. But before we look at the details, let us pause for one second: is it even true that Europe is growing more slowly than the US?
Let us start by looking at the facts about Europe. Map 6.1 shows the groups of countries that make up Europe, including the euro area (EA19), which shares the common currency of the euro, and the European Union (EU28). The why-is-the-US-growing-faster-than-Europe argument seems to play on the idea that the EU is obviously growing more slowly than the US. Well, not so fast (pun intended). Of course, the US has faster population growth. But as we have explained in Chapter 1, what matters for our analysis is growth in living standards, or per-capita growth.
Figure 6.1 shows the cumulative growth of GDP per capita over the past eighteen years in the US, in Europe, and for selected European countries. In the Appendix I explain how real growth is computed. All countries’ GDP per capita in 1999 are normalized to 1 in the figure, so you can read the axis as showing exactly the cumulative growth in standards of living between 1999 and 2017.
US citizens are about 21 percent richer in 2017 than they were in 1999. Let me add an important caveat here: this is an average. It does not take inequality into account
, and therefore it may not represent the experience of the median American household. Still, this is the measure that people have in mind when they think about growth.
The euro area (EA19) has not done as well as the US: the average citizen in EA19 is only 19 percent richer in 2017 than in 1999. The European Union, however, has done a bit better than the US: the average citizen in EU28 is 25 percent richer than in 1999. Why? Because countries like Poland have been catching up, and countries like Sweden have done well. These averages hide a significant amount of heterogeneity inside Europe. Germany is about 25 percent richer. France is only 15 percent richer. Sadly, Italy is slightly poorer.
If we step back and look at the big picture, however, we see that the US and Europe are growing at approximately the same rate on a per-capita basis. This is exactly what standard economic theory would predict. An important but somewhat subtle point is that the decreasing domestic competition hypothesis does not predict permanent differences in growth rates but rather temporary ones. Market power has a negative impact on the level of real GDP. As markups rise, growth will slow temporarily. If markups stay high in your country, you will be poorer than you would have been otherwise, but your country will eventually grow at about the same rate as before because long-run growth depends mostly on technological progress.
MAP 6.1 The euro area (EA19) began with eleven members in January 1999: Austria, Belgium, Finland, France, Germany, Ireland, Italy, Luxembourg, Netherlands, Portugal, and Spain. Later arrivals were Greece (2001), Slovenia (2007), Cyprus and Malta (2008), Slovakia (2009), Estonia (2011), Latvia (2014), and Lithuania (2015). Members of the European Union (EU28) share a common set of institutions (the European Commission, the European Parliament, a court of justice, and so on) and, most importantly for this book, the Single Market. Cyprus, an EA19 country, is not shown on this map. Brexit negotiations may change the UK’s membership status. Data source: https://d-maps.com/m/europa/europemax/europemax11.pdf