One of the principles I taught my economics students the first day of class and then applied incessantly thereafter was the importance of thinking on the margin. Garett Jones, an economics professor at George Mason University, has written a whole book in which he does just that. Jones considers what would happen if we made highly democratic countries less democratic and entrusted certain political decisions more to unelected officials. If you think he’s attacking democracy, you’re missing his point. He is calling for slightly less democracy. In short, he is thinking on the margin.

In what will probably be one of the most important books of 2020, Jones argues that if we made the United States and many other countries slightly less democratic, we would get slightly more freedom and slightly better policies. He makes his case by examining the details of central bank policy on inflation, appointed versus elected judges, restrictions on who can vote, the effects of the European Union, and the extreme case of Singapore.

Refreshingly, he knows how to coin humorous but informative lines. For example, in discussing William F. Buckley Jr.‘s famous statement that he would rather be governed by the first 2,000 names in the Boston phone book than by the Harvard University faculty, Jones proposes an alternative: “Rather than being governed by the masses of Boston or by the professors of Harvard, I’d far rather be governed by the engineering faculty of MIT.”

I started reading his book as someone who hates having important decisions made by “faceless bureaucrats.” I ended up thinking, along with Jones, that that might be better. He makes clear that basic democracy is good for a simple, important reason: “democracies don’t engage in widespread slaughter of their own citizens.” Thus, he calls for only slightly less, rather than a lot less, democracy.

The evidence / Exhibit A for Jones’s case is the effect of central bank independence on the inflation rate. He references a path-breaking 1993 article by Harvard economists Alberto Alesino and Lawrence Summers that looked at central banks around the world and used simple graphs to show that the more independent a bank was from the political system, the lower the inflation rate. If you dislike high inflation, as most people and most economists do, Alesino and Summers’s article argues for central bank independence.

Evidence from New Zealand was quite striking. In the 1970s and 1980s, when it had one of the least independent central banks, its average inflation rate exceeded 10%. But in 1989 the politicians gave New Zealand’s central bank much more independence. The result: by 1991 New Zealand’s inflation rate was down to 1%.

Jones also considers the effects of having elected versus appointed judges. He quotes from a 1988 book by elected judge Richard Neely of the West Virginia Supreme Court of Appeals:

As long as I am allowed to redistribute wealth from out-of-state defendants to insured in-state plaintiffs, I shall continue to do so. Not only is my sleep enhanced when I give someone else’s money away, but so is my job security, because the in-state plaintiffs, their families, and their friends will reelect me.

If this were just an anecdote, it wouldn’t mean much. But Jones, digging into the academic literature, shows that damage awards granted by elected judges are systematically higher than those granted by appointed judges. Specifically, “the average award paid by an out-of-state defendant was about $140,000 higher when the judge was elected rather than appointed.” He notes that averages can be distorted by a few top-end awards, but the median award for a partisan judge (one running on a political party’s ticket) was $38,000 higher than that for nonpartisan judges (those who were appointed or elected in nonpartisan races). And the difference at the 75th percentile was a whopping $304,000. Jones, a careful analyst, points out that to judge the effects of partisanship negatively, one must assume that—all other things equal—higher payouts are worse. Given the incentives that Neely identified in his quote and based on the literature on liability that I’ve read, I do.

Jones, who reveals in his introduction that he was once an aide to Sen. Orrin Hatch of Utah, also finds a systematic difference in voting between those senators who were in the first four years of their six-year term and those who were “in cycle,” that is, within two years of standing for reelection. He writes, “When a senator is in cycle, she’s 10 percentage points less likely to vote for a trade deal.” Since most trade deals move in the direction of free trade, this means that senators are more pro–free trade when they aren’t as worried about reelection.

Notice, by the way, Jones’s use of “she” even though the vast majority of senators in the sample studied are male. He consistently uses the feminine gender to talk about politicians, people at high levels in business, and judges. I know that publishers are pushing authors to do this, but Jones goes overboard. The one place where he uses the feminine gender where it fits an actual majority is in his discussion of Social Security beneficiaries: women outlive men on average.

People power / On voting, Jones reveals that there is a systematic difference between the knowledge and understanding of formally educated voters and those who have less education. He discusses the work of Italian public opinion researcher Vincenzo Memoli. Memoli measured basic political knowledge using survey questions such as who the Italian prime minister was and which political parties were on the left and which on the right. In a 2001 sample in which he used a scale of 1–9 to measure respondents’ knowledge, with 9 being the most informed, respondents who had a college degree had an average score of 7.7, those who had completed only high school averaged 6.7, middle school 5.3, and elementary school 4.7.

Jones quotes other work showing that more formally educated people are less likely to believe in implausible conspiracy theories such as the idea that there was never a moon landing. Of course, one can be more formally educated and better informed and still believe in bad policies. But Jones cites his George Mason colleague Bryan Caplan’s 2007 book The Myth of the Rational Voter, which finds that the least educated are the most likely to support, among other things, higher tariffs, rent control, and regulations that make it harder for employers to fire workers. Those are three policies that economists broadly oppose because of their harm to general welfare: tariffs hurt consumers; rent control destroys housing and reduces the incentive to build; and laws against firing workers make employers less likely to hire.

Jones cleverly titles the chapter in which he discusses voting “This Chapter Does Not Apply to Your Country.” It’s his way of trying to get us to drop our biases and imagine that we are setting voting policy for another country. Not one to avoid tough issues, he applies his understanding of voting to an issue that has become prominent in the United States: whether to allow felons who have been released from prison to vote. His discussion made me realize that I had regarded this as a moral issue: since these people had “served their time,” I thought they should be allowed to vote. Jones grants that people can have moral views different from his own, but he sees this as a practical issue: what would be the consequences of letting felons vote? He argues against letting them vote because they are on average less educated than non-felons. He quotes a striking finding by Caroline Wolf Harlow, a statistician with the Bureau of Justice Statistics:

About 41% of inmates in the Nation’s State and Federal prisons and local jails in 1997 and 31% of probationers had not completed high school or its equivalent. In comparison, 18% of the general population age 18 or older had not finished the 12th grade.

I found myself agreeing with Jones. (I should note that, for my first 14 years as an adult in the United States, from age 21 to 35, I was unable to vote because I wasn’t a U.S. citizen. When I finally got to vote, in 1986, it was a disappointment.)

Whatever your views on who should be allowed to vote, I would bet you’ll find some of his thoughts stretching your mind and reducing your biases. One of his ideas is to turn the U.S. Senate into a Sapientum, a council of the wise. The path to doing that, he writes, would be to change who gets to vote for senators. For instance, these voters could be limited to those who have a certain level of education, those who have been in a craft trade for at least five years, and those who are military veterans. These would require a certain level of intellectual maturity.

When Republicans took control of the U.S. House of Representatives in the 2010 elections, they fulfilled their pledge to eliminate earmarks—specific spending added to legislation to help constituents in various members’ districts. Jones predicted at the time that the reform would make it harder to rein in federal spending. Earmarks, he notes, were never more than 1% of federal spending, but were useful for persuading various members to vote for bills that reined in spending overall. He admits that he was “wildly incorrect” at first. Congress, along with President Obama, did start to get discretionary federal spending under control. But, argues Jones, in the long run—where we are now—he has been proven right: “In the U.S. Congress, with the decline of earmarks, the biggest games left in town are ideological outrage, social media grandstanding, and other behaviors that make individual politicians more willing to stand up to party leaders.”

Jones also weighs the pros and cons of Brexit in an interesting manner. He notes that Britain has the longest tradition of freedom of any country in the EU. He compares Britain to the fittest member of the gym, using that analogy to argue that Britain has the least to gain and the most to lose from the EU.

He also considers the paradoxical case of Singapore, a country with far less democracy than many other developed countries but, nevertheless, a great deal of economic freedom. He argues that two reasons Singapore works well are that it has a very high-IQ population and an independent judiciary, but he warns against going too far in Singapore’s direction.

There is one discordant note in an otherwise great book. In his introduction, Jones writes admiringly about how his Senate boss, Hatch, would give Democrat Ted Kennedy a big bear hug when they saw each other in the hallway. Jones uses this story to talk about the importance of bipartisanship. Certainly Kennedy should get many points for two things: his huge work on the Immigration and Naturalization Act of1965, which ended up allowing many more immigrants into the United States, and his initiative on airline deregulation in the late 1970s. But I can never think of Kennedy without remembering Chappaquiddick and that when he issued an apology for the incident, it was to his family and not the family of the young woman who drowned.

But just as we are better off judging a politician’s political moves separate from his character, we should judge Jones’s book separate from this discordant note. I hope 10% Less Democracy gets a lot of attention.