The Great Reversal

by Thomas Philippon

  If we accept these two assumptions, then cases in the EU can be used to proxy for “potential” regulatory scrutiny that is unaffected by US lobbying, and these cases should be positively correlated with excess lobbying in the US. We do show this to be true, and the link is significant. Whenever we observe more cases against one industry in the EU, we find more lobbying by that same industry in the US. We can then use EU cases to estimate the size of the reverse causality bias in the US—that is, the increase in lobbying in response to potential scrutiny. We find that the bias is substantial. One extra case in Europe increases US lobbying expenditures by 6 to 10 percent. This is a large effect since the standard deviation of lobbying expenditures (after accounting for industry and time effects) is only 38 percent.

  Finally, now that we have an estimate of the magnitude of the reverse causality, we can de-bias the estimate of the impact of lobbying in the US. We find that the estimated effect more than doubles. The de-biased estimate implies that a doubling of lobbying expenditures to the DoJ and FTC reduces the number of cases in a given industry by 9 percent. This is a sizable effect, considering that such lobbying nearly tripled from 1998 to 2008. If our estimates are correct, increases in lobbying can thus account for most of the decrease in enforcement in the US.

  Our model also predicts that lobbyists are more likely to succeed in the US than in the EU. Christine Mahoney (2008) finds that this is indeed the case. She performs a large-scale comparative study of the two systems, researching the work of 150 lobbyists fighting over forty-seven different policy issues, half in the US and half in the EU. She concludes that “In the US, 89% of corporations and 53% of trade associations succeed, while the majority of those fighting for the broader good—60% of citizen groups and 63% of foundations—fail in their lobbying goals … In the EU, we see that industry often wins as well (the success rates are 57% for trade associations and 61% for lobbying firms) but citizen groups and foundations fighting for the public good win at equal rates (56% and 67%).” She argues that these differences exist because legislators in the US depend on wealthy interests for campaign contributions.

  The Impact of Lobbying

  While there may be much that we don’t know about the precise impact of lobbying, the data we’ve been able to assemble tell us two key things for our purposes here.

  First of all, lobbying efforts work. Whether it’s a grassroots movement to save a beloved television show or an under-the-radar push to scuttle a new regulation before it ever becomes public, lobbying has a very real impact on public policy.

  Second, lobbying expenditures are growing rapidly in the US, and the vast majority are undertaken by the business community, either directly or through trade associations. These groups’ primary aims seem to be to protect (or create) economic rents.

  In its January 19, 2019, edition, the Wall Street Journal reported the illuminating example of the fight to outlaw online gambling. Gambling in America is mostly regulated by the states, and until 2011 the view of the Justice Department was that online gambling was prohibited by the federal Wire Act. In 2011, the DoJ changed its mind and decided that the prohibition applied to online sports betting but not to other forms of gambling. Casino magnate and top Republican donor Sheldon Adelson was displeased. In April 2017, his lobbyists drafted a memo arguing that the DoJ’s 2011 decision was wrong. In January 2019, the DoJ made an unusual move: it reversed its opinion. The Wall Street Journal noticed that the statement by the DoJ contained the same legal language and arguments as the 2017 memo written by Adelson’s lobbyists.

  The comparison with lobbying in Europe is interesting. There are many similar cases in which lobbyists’ language has made its way into regulations in Europe as well. In fact, in the case of technically complex regulations, this is exactly what one would expect, and it is consistent with the expertise and information-sharing view of lobbying. On balance, however, it seems that lobbying is less prevalent in Europe. This suggests that lobbying may be one potential driver of the decline in competitive pressures in US markets compared to EU markets. An important factor is that legislators in the US depend on wealthy interests for campaign contributions.

  This not the end of the story, then. Our next task is to understand how political campaigns are financed.

  * * *

  a  This is a reference to an old joke: A police officer sees a drunk man searching for something under a streetlight and asks what the man has lost. He says, “I’ve lost my keys,” and they look under the streetlight together. After a few minutes the policeman asks, “Are you sure you lost them here?” The man replies, “No, I lost them in the park.” The policeman asks why he is searching here, and the drunk man replies, “This is where the light is.”

  b  The ideal experiment would involve a random change in prices not linked to demand, such as a supply chain problem that randomly increased the stock of milk in the store. Then the owner would lower the price to decrease the stock, and you would be able to estimate the response of consumers (their demand elasticity).

  c  Greenwood and Dreger (2013) estimated that 75 percent of businesses and 60 percent of NGOs active in engaging EU political institutions were in the register, and the number of registrants has increased by more than 50 percent since 2013. On the other hand, lobbying may be overestimated due to double counting: the data contain the corporations that employ lobbying intermediaries as well as the lobbying intermediaries themselves. There are also some measurement issues with small firms, and we follow LobbyFacts.eu in applying restrictions based on the number of European Parliament passes and European Commission meetings to mitigate these issues. In particular, we drop observations in the top 5 percent of lobbying expenditures by year for firms that have no European Parliament passes and no European Commission meetings. We also replace lobbying expenditures for “University College Dublin–National University of Ireland, Dublin” in 2015 with the prior year’s quantity because it is an extreme outlier. The totals after applying these restrictions roughly match those reported in media outlets such as the Guardian (May 8, 2014). Note also that most firms report ranges of lobbying expenditures rather than specific amounts. We take the midpoint of all ranges in our estimates. Annual totals for the EU are based on the complete register available through LobbyFacts.eu as of year end 2012, 2013, 2014, and 2015.

  CHAPTER 10

  Money and Politics

  Politics has got so expensive that it takes lots of money to even get beat with nowadays.

  WILL ROGERS

  HOW SUCCESSFUL would a candidate running for a seat in the US House of Representatives be if, during every campaign event, she said that she would take a part-time job that would claim thirty hours of every work week if she were elected? She’d likely be laughed right out of the race, right?

  The problem is, she would have accurately described the amount of time that actual members of Congress spend on fundraising every week, year in and year out, as long as they remain in office.

  Politics has become so expensive that new members of Congress are told in no uncertain terms that their first and highest priority in Washington is to raise the money they need to ensure their re-election. In 2016, Representative David Jolly, a Republican from Florida, told the news program 60 Minutes what his first days in Washington were like. He had arrived after winning a special election, which meant that he would face voters again only six months later.

  “We sat behind closed doors at one of the party headquarters back rooms in front of a white board where the equation was drawn out,” he said. “You have six months until the election. Break that down to having to raise $2 million in the next six months. And your job, new member of Congress, is to raise $18,000 a day. Your first responsibility is to make sure you hit $18,000 a day.”

  And it never stops. Democrats and Republicans alike are given quotas to meet every year, depending on their seniority and placement on different committees. A Democratic memo prepared for new House members in 2013 advised them to set aside a
t least four hours every day to call potential donors—“dialing for dollars,” it’s called. The same memo suggested they set aside only two hours for activity on the House floor or in committee meetings.

  FIGURE 10.1  Average direct spending by winning candidates. All spending is in 2014 dollars to neutralize the effect of inflation. Data source: Center for Responsive Politics

  Will Rogers, quoted in the epigraph to this chapter, noted the presence of too much money in electoral politics in 1931. The problem isn’t a new one. But the kind of money politicians are now expected to raise is orders of magnitude larger than it was back then.

  Figure 10.1 shows the average direct campaign spending of winning candidates for the Senate and for the House. The cost of winning a Senate race was around $4 million in 1986 (adjusting for inflation). In 2014 it cost $12 million. Winning a seat in the House is cheaper, but the cost has still doubled over the past thirty years, from $800,000 to $1.6 million.

  Moreover, Figure 10.1 does not tell the full story. It shows only the direct spending by the candidates’ campaigns. Political action committees (PACs), super PACs, and political nonprofits, which we will dissect in this chapter, have become increasingly important since 2010.

  TABLE 10.1

  Five Most Expensive Senate Races of 2014

  Total spending

  Campaign

  Outside groups

  North Carolina Senate

  $113,479,706

  $32,390,468

  $81,089,238

  Colorado Senate

  $97,285,589

  $27,887,734

  $69,397,855

  Iowa Senate

  $85,364,286

  $23,452,451

  $61,911,835

  Kentucky Senate

  $78,231,062

  $44,838,119

  $33,392,943

  Georgia Senate

  $66,136,490

  $39,579,101

  $26,557,389

  Data source: Center for Responsive Politics

  Table 10.1 shows the expenditure of the top five Senate races in 2014. The most expensive race was the one pitting an incumbent Democratic senator, Kay Hagan, against a Republican challenger, Thom Tillis, for the North Carolina Senate. What is interesting is that the two candidates spent just $32 million of the total $113 million spent in the election. Most of the spending came from outside groups.

  Naturally, campaign contributions and election costs go hand in hand. Figure 10.2a shows the contributions by various groups. Total contributions have increased by about $6 billion. By far the largest donors are business lobbies. It is also the group that has increased its contributions the most, from $2 to $6 billion, thus accounting for two-thirds of the overall increase. The other big increase comes from very wealthy individuals, the top 1 percent.

  Figure 10.2b shows the concentration of donors, pooling all groups together, and breaking them down by largest contributors. The top 1 percent of donors contribute about three-quarters of the total. As we discussed in the previous chapter, political contributions, just like lobbying expenses, are extremely skewed. It’s really only the big donors that matter.

  Among individual contributions, the skewness is even more impressive. Consider the top 0.01 percent. One way of thinking about this is that a person in the top 0.01 percent is the richest among 10,000. This segment has increased its share of national income, and today earns about 5 percent of total income, as we know from the famous work of Thomas Piketty and Emanuel Saez. That is a very skewed distribution. But it’s nothing compared to the distribution of campaign contributions. The top 0.01 percent of donors contribute an astounding 40 percent of all contributions.

  FIGURE 10.2   (a) Political expenditures by groups; (b) The concentration of contributions (both in 2014 dollars). Data source: Center for Responsive Politics

  TABLE 10.2

  2016 Election Donations ($MM)

  Hillary Clinton

  Donald Trump

  Raised by candidate

  $973

  $564

  Raised by super PACs

  $217

  $82

  Total

  $1,190

  $646

  This does not mean that the biggest spender always wins. Donald Trump won the presidential election in November 2016 despite raising and spending much less than Hillary Clinton (see Table 10.2). A few months earlier, he beat better-funded candidates in the Republican primaries. In fact, Donald Trump raised less outside money than any major party presidential nominee since John McCain in 2008. Yet he dominated the airwaves and won the election.

  On average, however, the largest spender usually wins. For instance, analysis of the 2014 elections by the Center for Responsive Politics shows the candidate who spent the most prevailed 94.2 percent of the time in House races and 81.8 percent of the time in Senate races. But that does not mean that money is the reason that they win. The fundamental problem of endogeneity that we have discussed in Chapter 9 is back with a vengeance in the case of political campaigns. We observe a correlation, but we do not know which way the causality runs.

  Perhaps incumbents would win even if they did not outspend their opponents. Perhaps people simply like to give money to the winning team: this is called reverse causality. Or perhaps it the same underlying qualities that make some people better candidates and better fundraisers at the same time: this is called an omitted variable bias.

  Our task in this chapter is to understand how money and politics interact and to tease out the causal effects of one on the other. Like lobbying, this is difficult because spending decisions are strategic choices, not random decisions. The curse of endogeneity is still with us. Businesses might strategically bid on winners, but that does not mean that money is necessary to win. On the other hand, it does suggest that businesses expect some returns and that politicians value their contributions.

  Campaign Finance Laws

  Campaign finance controversies are as old as the Republic, and their history is full of irony, misleading laws, and half-baked solutions. Theodore Roosevelt received more than $2 million in corporate contributions for his 1904 campaign. After his election, the donations created a controversy, and Roosevelt himself called for the prohibition of corporate contributions. In his annual address to Congress in 1905, he said: “All contributions by corporations to any political committee or for any political purpose should be forbidden by law; directors should not be permitted to use stockholders’ money for such purposes; and, moreover, a prohibition of this kind would be, as far as it went, an effective method of stopping the evils aimed at in corrupt practices acts.”

  Roosevelt signed the Tillman Act in 1907. It was the first legislation prohibiting monetary corporate contributions to national political campaigns. But the act did not include an effective enforcement mechanism. It included penalties, but nobody to apply them. There was no Federal Election Commission and no disclosure requirements for candidates. In addition, the act did not apply to primary elections. In many places (much of the Democratic South, for instance), the general elections were not really contested, so the primary election was the most important one. And finally, a corporation could ask its officers or directors to make personal contributions to a candidate and reimburse them later with bonuses or other perks.

  Disclosure requirements and extensions to primary elections came in 1910 and 1911. There were also some spending limitations, but these were later struck down by the US Supreme Court.a Political action committees wouldn’t appear for a few more decades, and the way they did is quite illuminating. Congress tried to ban labor union contributions to candidates through the Smith–Connally and Taft–Hartley Acts of the mid-1940s. The Smith–Connally Act forbade unions from contributing to federal candidates. In response, the Congress of Industrial Organizations (CIO) formed the first PAC in 1944 to raise money for the re-election of President Franklin D. Roosevelt. The PAC’s money came from voluntary contributions from union members rather than union treasuries,
so it did not violate the Smith–Connally Act. PACs were unfettered by the political advertising and spending laws that applied to the candidates, so they could spend as much as they wanted to independently promote specific candidates and issues among their membership and to the general public.

  As Thomas Stratmann (2019) explains, “the modern era of campaign finance laws at the national level began in the 1970s, and these laws have changed a few times since in subsequent legislation and court rulings.” The Federal Election Campaign Act (FECA) of 1971 initially proposed spending caps but not the resources to enforce them. FECA was refined in 1974 to create the bipartisan Federal Election Commission (FEC)—the body that enforces campaign finance law to this day. In 1976, Sen. James Buckley, a Republican from New York, argued in front of the Supreme Court that FECA limits on campaign spending violated free speech rights. The Supreme Court agreed, and its ruling in Buckley v. Valeo was ultimately interpreted as allowing unlimited spending by political candidates.

  In 2002, the Bipartisan Campaign Reform Act—also called the McCain–Feingold Act—attempted to limit spending and ban political ads within thirty days of the election. The law was quickly challenged, and the Supreme Court struck down several of its provisions.