Jeff Bezos once said that “there are two kinds of companies, those that work to try to charge more and those that work to charge less. We will be the second.” And he was right: Amazon does not charge high prices to consumers. Amazon, however, uses the large scale of its operations to obtain rebates from its suppliers and delivery services, which increases its market power and makes it difficult for other firms to compete if they do not enjoy the same rebates. The macroeconomic implications of monopsony power are not so different from those of monopoly power. If a platform can push down the price at which it buys goods, the goods producers will have to reduce wages. Regulators should therefore make sure that rebates and other forms of monopsony rents do not become a barrier to entry in these markets.
Conclusion
Se vogliamo che tutto rimanga come è, bisogna che tutto cambi.
GIUSEPPE TOMASI DI LAMPEDUSA
“For things to remain the same, everything must change,” explains Tancredi, the nephew of Prince Fabrizio Salina, as General Garibaldi’s forces sweep through Sicily in Giuseppe di Lampedusa’s novel, Il Gattopardo [The Leopard].a And so it is with free markets. Most of the issues that we struggle with are not new, but for markets to remain free, we must constantly adapt.
I have spent hundreds of hours researching and writing this book. You might have spent a few hours reading it, or, as I tend to do myself, jumping from figures to tables and reading the text only when the data look interesting.
At the very least, after so much work, we should have learned something. That is to say, we should have changed our minds about a few things. And we should have a few ideas to improve the current state of affairs.
I will not make a long list of policy recommendations. Instead, I am going to tell you what I learned, what surprised me, and what I took away.
What Surprised Me
I was surprised by how fragile free markets really are. We take them for granted, but history demonstrates that they are more the exception than the rule. Free markets are supposed to discipline private companies, but today, many private companies have grown so dominant that they can get away with bad service, high prices, and deficient privacy safeguards. Only two decades ago, the United States was effectively the land of free markets and a leader in deregulation and antitrust policy. If America wants to lead once more in this realm, it must remember its own history and relearn the lessons it successfully taught the rest of the world.
Excessive concentration hurts individual consumers, true, but its impact spreads wider. Former defense secretary Robert Gates tells us in Duty: Memoirs of a Secretary at War (2014) that promoting competition among defense contractors was a real issue. In 2015 the Pentagon’s acquisitions chief, Frank Kendall, explained that “one can foresee a future in which the department has at most two or three very large suppliers for all the major weapon systems that we acquire … The department would not consider this to be a positive development and the American public should not either.” Rodrigo Carril and Mark Duggan (2018) study mergers among defense contractors to investigate how market structure impacts competition and costs. They find that market concentration in the defense industry has increased in the last three decades, leading to less competitive bidding and to more contracts that require the federal government to pay the contractor for all costs incurred, plus a markup.
I was surprised by the strength of the evidence in favor of competition. We have always known that competition lowers prices and increases real wages and standards of living. But the evidence also shows that competition is good for investment and suggests that it fosters innovation and productivity growth. In theory, however, competition can be excessive. We should expect, then, to find at least some industry at some time period in some country where competition has had a negative impact on innovation. But we find hardly any. Why is that? This is the story of the dog who did not bark. I have seen and discussed many papers showing that competition is good for growth (Buccirossi et al., 2013). I have not seen a paper that convincingly shows that protecting incumbents through patents, barriers to entry, or any other stratagem leads to an increase in productivity. It is highly unlikely that no researcher would have found this effect if it was in the data. The dog did not bark; that is the key piece of evidence. But why is it so difficult to find any such example? I think the answer is actually quite simple. The standard theory is incomplete because it ignores the political incentives of incumbents who lobby to weaken competition and erect barriers to entry. They often succeed for the wrong reasons, and this is why free markets are fragile. But in the rare instances in which their arguments have merit, their success is almost assured. As a result, the empirical distribution of industries is biased toward insufficient competition, and we almost never observe an industry on the wrong side of the innovation / competition curve.
I was surprised by the power and persistence of institutions beyond their original intent. It came to my attention when comparing Europe and the US. Having lived almost the same amount of time on each side of the Atlantic, I think that Europe faces deep challenges, probably deeper and more dangerous than the ones confronting the US. Moreover, European countries have not generally been at the forefront of good and innovative economic policies over the past thirty years. Yet EU competition policy has become stronger than US competition policy, and EU citizens are better off for it. It can all be traced back to the design of the Single Market, an institution that was heavily inspired by the best in class at the time. That’s the irony. The reason EU consumers are better off than American consumers today is because the EU has adopted the US playbook, which the US itself has abandoned.
I was surprised by the extent and intricacies of American lobbying and campaign finance. They are at the same time visible and hard to pin down. They are hidden in plain sight but with enough noise to maintain plausible deniability. Institutions and interest groups create a dynamic ecosystem. Institutions shape the games played by lobbyists and are themselves influenced by political decisions. Economic inefficiencies are often embedded in overly complex institutions and sustained by continuous lobbying. The health-care system is a case in point. As a patient and especially as a parent of young children, I had naturally noticed the complexities, costs, and glaring inefficiencies of the system. But I did not fully comprehend the scale of the problem until I started putting the data together.
I was surprised by what I discovered about the internet giants. Like (almost) everyone, I use some version of Word to write documents, I have fond memories of my first MacBook, I remember the first time I saw the slick design of Google’s home page, I order from Amazon, and I keep track of friends and family on Facebook, Instagram, and WhatsApp. Researching these firms, however, helped to clarify my thinking about exactly how they are special and how they are not. They are certainly fascinating companies but not nearly as influential in the economy as we might think. At least not yet. My big hope is that they can translate their ingenuity into things that matter more than tweets, targeted ads, and cute photos. But I am also convinced that they have become too entrenched and need a strong dose of competition.
Finally, I was surprised by the gap between economic research and policy. As economists we love to complain that if only politicians listened to us, economic policies would be more effective. There is some truth to that idea, but it is also too self-serving for my taste. First of all, as any economist who has had some policy experience will tell you, an economic adviser spends most of her time trying to kill obviously bad ideas and rarely gets the chance to advocate for good ones. Moreover, there is fairly strong evidence that economists fail to provide timely advice. The financial crisis of 2008 is an obvious example. Most financial economists failed to understand the risks before it was too late. The prevailing view at most academic conferences during the 2000s was that the finance industry was driving innovation and growth, and few economists questioned that narrative. Challenging the common wisdom is stimulating, but it is also risky. I wrote a paper in early 2008 to mea
sure the efficiency of finance and, much to my surprise, I found that finance had not actually become more efficient. We discussed that idea in Chapter 11. This is probably one of my better-known papers, but it took seven years—and many rejections—to publish it (see Philippon, 2015).
There was a similar failure by economists to study and understand the growing concentration of many industries as it was happening. Today, of course, there is a flourishing literature on markups, competition, and concentration. But, with a few exceptions that I have highlighted in this book, none was prescient, and none challenged the status quo at a time when that would have been controversial and therefore useful.
The same issues occur in other areas of economics. Trade policy provides a good example. There is a wonderful Trade Talks podcast with Paul Krugman that everyone should listen to.b Krugman explains how models of trade tended to lag behind developments in global trade. Before the 1980s most trade took place between rich nations and involved the buying and selling of items within the same industries. Standard models did not explain these patterns: they predicted trade of different goods between nations with different comparative advantages. New trade models based on specialization and returns to scale were developed to account for trade between rich countries … just in time for the rise of trade with poor countries that is well explained by the older, comparative advantage models. When the issue of trade and inequality became salient in the 1990s, trade models predicted small effects and economists argued there was not much to worry about … just in time for the entry of China into the World Trade Organization, which had a significant impact on employment and wages. With the China shock behind us, there is less to worry about as far as inequality is concerned, but this is not how the public sees it.
There is a lesson of humility here. It is not only because of misguided populism that economists have lost the trust of the public. It’s also because we have often failed to challenge the consensus and to provide timely advice.
Taking Stock
My main argument in this book is that competition has declined in most US industries over the past twenty years. Here I would like to answer the trillion-dollar question: how much does this matter? To be more precise, suppose we could roll back the barriers to entry, undo the bad mergers, and somehow return to the level of competition we had in the late 1990s. How much better off would we be?
We are going to use a relatively simple model of the economy to answer that question. When economists talk about a “model,” we mean a set of equations that represents how economic agents behave. Households work to make a living: they supply labor; they decide how much to save and what to buy, and they make consumption and saving decisions. Firms compete with each other to supply the goods and services that households and other businesses want to buy. They hire capital, labor, and intermediate inputs from other businesses. They understand that demand is elastic: they lose customers if they set their prices too high. All of these decisions can be written as mathematical objects. We can also incorporate the decisions of the government (taxes, spending, regulations) and the central bank (interest rates).
The virtue of a model is that we can compute the outcome of all these decisions. We call this outcome the macroeconomic equilibrium. The concept of equilibrium is important because the decisions are interdependent. Consider the labor market, for instance. Households supply labor, whereas firms hire labor. But firms hire labor because they expect to sell their products, which are bought by households using their labor income. Similarly, when we say that households save, we mean that they keep money in their bank accounts or that they invest in mutual funds. But banks and mutual funds are intermediaries, not end users. The savings eventually find their way into loans, bonds, and stocks. The return on these claims depends on the capital demand by firms. All these decisions are therefore interdependent. The practical implication is that if we want to understand the consequences of competition—or lack thereof—we need to keep track of what happens in all these markets at the same time. That’s why we need a model.
Once we have the model, the key question becomes this: how large was the change in competition? Let’s review the evidence. We saw that after-tax profits have increased by about 4 percentage points of GDP (Chapter 3). The labor share of income has decreased by about 6 percentage points of GDP (Chapter 6). When we compare the US with Europe, we find a relative markup increase of about 10 percent (Chapter 7). Some of it was due to EU markups going down in addition to US markups going up.
I am then going to feed into the model an experiment that is consistent with this evidence. Let us start from a situation that represents the 1990s. Markups are 5 percent over gross output, which means that firms add a 5 percent margin to the costs of labor, capital, and intermediate inputs. The economy has free entry, so these extra profits simply offset the cost of setting up and operating the businesses. I define the units so that GDP is $100 and total labor income is $65. The labor share is 0.65.
Now imagine that competition declines and that free entry is violated. Businesses can increase their margins from 5 percent to 10 percent. What happens then? The demand for capital, labor, and inputs decreases. Wages decrease too. The impact on employment depends on the willingness of households to continue working for lower wages. I use a conservative model in which households keep on working.c As a result, the main consequences of higher markups are lower wages, lower investment, and lower productivity, while employment stays about the same. Let us look at the numbers in more detail. Because of competition, GDP is only $95, that is, 5 percent lower. Labor income drops to $57. The new labor share is 57 / 95, which equals 0.6, and is in line with the evidence in Chapter 6. The capital stock decreases by 10 percent, which is also in line with the gap discussed in Chapter 4.
Let us put these numbers into perspective. US GDP is about $20 trillion. If we could make the economy as competitive as it was twenty years ago, the GDP would increase by 5 percent to $21 trillion. Employee compensation is about $11 trillion. In a competitive economy it would be 65 / 57 × 11, or $12.5 trillion. In other words, my calculations suggest that the lack of competition has deprived American workers of $1.5 trillion of income. This is more than the entire cumulative growth of real compensation between 2012 and 2018. The lack of competition has cost American workers six full years of growth. That is a large cost by any measure.
Another way to gauge the importance of the issue is to compare it to other policy proposals. As I am writing these lines, the US is entering the 2020 election cycle. I do not have the details of the proposals by the various candidates, but I would be surprised if there is one that tops that number.
Returning to a high-competition economy will not be easy. Those who benefit from the lack of competition will fight to protect their vested interests. Let me offer a few principles to help you navigate the policy debate.
A Few Economic Principles for the Twenty-First Century
Principle 1: Free Entry, Always and Everywhere
A market cannot be free if it does not exist, so a monopoly is certainly better than a “zeropoly.” Once a market exists, however, there is no reason to believe that the rents of the incumbent need to be actively protected, and many reasons to believe that they are usually too high. As a practical matter, fostering competition improves welfare. As I explain, the dog of excess competition did not bark.
If you believe in free markets, you need to be tough with everyone. American politicians and regulators have let regulations that restrict entry grow at the federal level and at the state level. We discussed the evidence in Chapters 5, 8, 9, and 15, but our data probably capture only the tip of the iceberg. Conservatives are right to argue that the US needs fewer regulations. I would simply qualify this idea as “regulations that hinder the entry or growth of small firms.”
We also need to be tougher with incumbents, even the ones that we like. Today’s monopolies are yesterday’s startups. Successful firms are always both wonderful and excessive. Successful entrepreneurs
are always part titan and part baron. Steve Jobs said he wanted to “put a ding in the universe.” And he did. But that should not prevent us from keeping Apple’s monopoly power in check. Competition and antitrust remedies are not punishments for moral wrongdoing—at least most of the time. They represent economic solutions that make the broader economic system more efficient. Firms have the right to try to beat their competitors and even to drive them out of business. Regulators have a duty to make sure they do not impede free markets. But no one should take it personally; it’s just business.
Corollary 1: In an Efficient Market, the Marginal Firm Is Nearly Bankrupt
In a competitive market, the marginal firm should be on the brink of bankruptcy. This is the definition of free entry. Firms keep entering until they are not sure they will break even. Financial distress is not a bad sign and should never be used as an excuse for lax antitrust decisions. Moreover, bankruptcy is not liquidation. Distressed firms are usually reorganized. Bad policies follow almost invariably when temporary financial distress is used as an excuse to limit competition. The airlines used their financial distress as an excuse for concentration, and American skies became a large oligopoly.
Principle 2: Governments Should Make Mistakes Too
We live in a world where we tolerate data breaches from our banks, credit card companies, social media companies, email servers, and credit scoring companies. But in this world the idea that a regulator might make a mistake is unacceptable. When their cases are not perfect, regulators are pilloried by the press. It’s no wonder they are shy. But they should not be. We are facing new issues that require new solutions. New solutions always involve mistakes, trial and error, and corrections. A zero-mistake approach to regulation is an approach to zero regulations. The courts should hold regulators to high standards, but overly laborious burdens of proof are not helpful.
The Great Reversal Page 31