If you are not an economist, social scientist, or serious policy wonk, it can be difficult to know how to decide which of two conflicting sources to trust, if both have a bunch of credentials and present a lot of charts and have a lot of endnotes. A left-leaning think tank scholar tells me raising the minimum wage does not have substantial effects on employment rates. A right-leaning think tank scholar tells me it absolutely does. Unless I am going to dig into every study, and try to resolve a complicated empirical question on my own, how am I to know what to believe?

My training is as a lawyer, not an economist, so when I see a dispute over whether, say, some proposed regulation will or won’t “kill jobs,” I don’t always know how to settle the question to my own satisfaction. Fortunately, there are things I can do to increase my confidence in an answer. I look carefully at what each side is saying, and see if any parts of it that I do understand are wrong, irrational, unfair, or manipulative. Then I can at least identify untrustworthy sources, those who I should mostly ignore because I have evidence that, on matters I don’t have expertise in, they are willing to pull the wool over my eyes.

The Cato Institute is a libertarian think tank, often rated one of the most important and influential think tanks in the world. About 10 Nobel laureates in economics have been affiliated with it, and it is home to many PhD economists who have extremely long CVs. I do not have a PhD, and my CV has more children’s books than academic articles. And yet I take almost nothing from the Cato Institute seriously. Why? Am I simply a partisan, refusing to listen to the arguments of smart people who disagree with me?

Perhaps. But mostly I don’t trust them because every time I do read something from them, it’s untrustworthy. The Cato Institute is ideologically committed to free markets. They think the market knows best, and regulation is harmful. But, like many libertarians, it means that they are reluctant to admit any facts that contradict this case. If you read a Cato publication on almost any issue, you will find that—surprise!—privatization and deregulation are the solution, and government is getting in the way. As a general rule, you should be wary of people who always know the solutions before they have looked at the facts.

Let me show you two examples of what I mean. Here is a recent Cato paper by Vanessa Brown Calder called “Parental leave: Is there a case for government action?” (It’s just the first paper I came across.) Many on the left believe that the United States government should introduce paid family leave policies similar to the ones found in essentially every other industrialized country. The argument we lefties make is that working mothers need to be guaranteed time to spend with their newborn child during the critical first months of its life, without having to worry that they can’t afford to be with their baby. Not only do the vast majority of women believe new mothers should be allowed to take paid leave, but the countries that have required paid leave by law have achieved success in extending the time mothers get to bond with their babies.

Calder’s paper argues that the federal government should not require paid family leave. To address the problems faced by working mothers, Calder proposes libertarian measures like deregulating childcare to eliminate legal requirements that care providers have a high school diploma. I don’t think that’s any kind of solution at all, and think it illustrates just how few actual helpful policy proposals libertarians are capable of offering, but here I’m more interested in examining the way Calder makes the case that paid family leave is bad policy.

You hear people accusing each other of “cherry-picking” evidence a lot, the accusation being so common that it has lost its force. I don’t like the phrase, actually, because I don’t think it actually sounds bad: Cherries are delicious, why wouldn’t you pick the cherries? What’s wrong with presenting the best possible case? But the actual actions “cherry-picking” refers to are a form of serious intellectual dishonesty. If you pick only the data that support the position you already hold, and ignore or bury the data that undermine that position, you are not a scholar. You are violating the core mission of social science, which is to find the truth. You’re just a propagandist, someone who produces whatever arguments they think will be convincing without regard for whether those arguments are correct.  

Vanessa Brown Calder’s paper is exactly this kind of unscholarly propaganda. She wants to show us that paid family leave is a bad policy, so she goes rifling through academic research and presents all the pieces that suggest paid family leave is bad, ignoring the parts that she doesn’t like because they don’t comport with the Cato Institute’s unscientific libertarian dogmas.

Here is how Brown Calder presents her research:

Proponents of these measures often argue the private market cannot sufficiently provide paid leave due to collective action problems. They claim that government-supported leave would markedly improve workers’ lives by improving labor-market outcomes and reducing gender inequality. In this paper, I provide economic research and federal data that suggest otherwise.

Her first section is titled “The Private Market Provides Paid Family Leave.” She aims to show that, contrary to the claims of paid leave advocates, the market is doing an adequate job. She says that advocates of paid leave distort the facts by citing a number from the Bureau of Labor Statistics, which says that only 15 percent of workers have paid maternity leave. Calder argues that this number is misleading, because it only counts those workers who have paid maternity leave. Or rather, it counts those who have a maternity benefit “in addition to any sick leave, vacation, personal leave, or short-term disability leave that is available to the employee.” Calder says that this means a worker who used her sick leave or vacation leave as baby time would not be counted, and that this makes the statistic inaccurate.

Now, you might think the fact that women are cobbling together their maternity leave from their vacation days and sick leave actually shows why paid maternity leave is necessary. But look also at the way she proves that the market is providing adequately. She cites five other surveys. They found, respectively:

• “63% of employed mothers said their employer provided paid maternity leave benefits.”
• “63% of Americans who took time off from work in the past two years for parental, family, or medical reasons report that they received at least some pay during this time.”
• 50.8% of mothers report using paid leave “of some kind” before or after childbirth and 60% report using paid leave or disability
• “45% of working women who took parental leave received some pay”

For Calder, the upshot of this is that “between voluntary expansions in paid leave and more mothers in the workforce, U.S. companies are already spending substantial resources on paid leave benefits” and “the functional consequence of firms’ policies is that workers can take time off for family matters.” You will, I am sure, have already seen why much of this is fishy. “Some pay.” How much pay? How much time? In the first survey, it turns out that only ⅓ of the 63 percent provided full pay during the leave period. In two of these surveys it’s unclear whether it was specifically maternity leave, and in none of these findings do we know anything about what the benefits are! Perhaps the benefits are substantial. But for Calder, once we’ve proved that there is any kind of benefit that is used by new mothers, we’ve shown that the market is doing its work.

Another few things to note: First, California, by far the largest state in the union, has a mandatory paid leave law, which one might consider important when discussing what the “market” has done. (New York also has one too, though it was enacted after most of these studies.) Second, Calder has chosen to cite only the studies that yield the highest numbers of people taking paid leave. The Council of Economic Advisers says “only 39 percent report being able to take some type of paid family leave for the birth of a child,” and even one of the studies Calder cites elsewhere contains findings that “35 percent of worksites offered paid maternity leave to all or most employees” and “39 percent of workers reported having paid leave coverage for reasons related to the birth or adoption of a child.” If the high numbers make it onto your chart, and the lower numbers are left off, are you an honest scholar? (You are not.) But wait a second, there’s another thing: Let’s assume it’s 63 percent. And let’s assume those 63 percent all get decent benefits for a generous time and aren’t having to use up their other sick leave. What about the rest of them? Presumably the ones with the benefits are going to disproportionately be in professional jobs. Are a huge number of working-class moms just screwed? “The Private Market Provides Paid Family Leave.” It doesn’t seem to, actually! It seems to provide some leave to some people. Note Calder’s phrasing: “U.S. companies are already spending substantial resources.” Substantial! Enough? Who knows!

Throughout the rest of Calder’s paper, the dishonesty really escalates. She wants to show that paid family leave laws don’t help women, and might actually hurt them. (It is an absolute truth of the libertarian universe that every regulation must Hurt The Very People It’s Trying To Help.) Consider this paragraph:

Government-mandated leave has similar effects internationally. A study of 16 European countries over a period of around 20 years found that “parental leave is associated … with reductions in [women’s] relative wages at extended durations.”15 Other researchers have noted that “work-family policies … have also contributed to … lower wage-levels for women relative to men.”16

“I bet if I actually look at those footnotes, I will find that Calder is taking the parts of the papers that support her argument and ignoring the parts that undercut it,” I said to myself upon reading this. And, well, here’s the conclusion from the first study she cites:

Rights to paid time off work raise the percentage of women employed, with a substantial effect observed for even short durations (i.e. less than 3 months) of guaranteed work absence. [L]eave legislation increases the employment-to-population ratio of all females by around 4% and of women of childbearing age by approximately 9%… Short periods of leave are found to either have no effect on or to raise female earnings. Conversely, lengthier paid entitlements are associated with substantial wage reductions, with predicted decreases in the range of 1.5% to 3% for durations of six months or more… To summarize, the employment of women appears to be increased by even relatively short (less than 6 months) durations of paid leave, whereas their relative wages may fall with more extended entitlements. This suggests that the work absences currently guaranteed in many European countries may be so lengthy as to reduce the earnings of females. Conversely, there is little indication that substantial costs would be incurred, in the U.S., if the period of parental leave were modestly extended or if payment were provided.

So, the study author found that: 1. Paid leave policies increase female employment 2. For anything up to six months, there is either no effect or an increase in female earnings. 3. If it’s over six months wages drop slightly. 4. THERE IS LITTLE INDICATION THAT INTRODUCING A PAID PARENTAL LEAVE POLICY WOULD INTRODUCE SUBSTANTIAL COSTS. What does Calder cite? The bit about wages dropping during long leave periods, and nothing else. The next footnote—more of the same? Yes, to a slightly lesser degree. It does indeed present evidence that existing work-family policies contribute to gender inequality in the workplace. But it argues that the solution is to fix the policies to promote gender equality; it doesn’t even contemplate getting rid of paid leave policies altogether:

As we see more generous and paternity leave and “daddy month” measures – as well as men’s take up of these policies – we hope to see more egalitarian outcomes. Yet, our analyses also suggest the importance of designing effective leave policies – that promote parental attachment to the labor force rather than “cooling out” parents, and public provisioning for high quality, employment-enabling childcare.

Okay let’s do one more, real quick:

A recent review of the literature suggests “employers who find leave-taking costly may discriminate against female employees by being less likely to hire them.”21

Does this recent review of the literature say anything else? Oh look, it does:

Maternity and family leave policies play an important role in helping working parents— especially mothers—navigate the challenges of balancing job and family responsibilities upon the arrival of a new child. This review of the existing research provides insights into key questions regarding the take-up of leave, impacts on women’s labor market experiences, consequences for child well-being, and effects on employers. The current evidence demonstrates that the introduction of new leave programs and extensions of existing programs increase leave-taking rates among parents. Leave policies that are less than one year in length can increase labor supply among women immediately after childbirth as well as several years later; longer leaves can have negative consequences on women’s careers. Moreover, while extensions in existing paid leave programs have no impacts on child welfare, the introduction of short programs can improve infant health and children’s long-run outcomes like education and earnings. With regard to employers, the current research finds little evidence of adverse impacts on self-reported productivity, employee morale, turnover rates or the total wage bill.

Those who study the impact of paid family leave policies have generally found that there are benefits and trade-offs, and they vary based on the kind of policy you introduce, but in general short-term paid leave policies produce a lot of good outcomes and only limited downsides. If you are an honest scholar, you would present both the benefits and the trade-offs. Most economists who study this are honest, and that’s exactly what they do. If you are a Cato Institute “scholar,” on the other hand, you present the parts that make it seem as if all regulation is bad and the downsides of mandates always necessarily outweigh the upsides and lefties are economic illiterates who will Hurt The Very People They’re Trying To Help.

Now, Calder does not find space to present the benefits of paid leave policies described in the literature. But she does find space for an extremely stupid chart. Have a look at this:

Although descriptive data do not constitute strong standalone evidence, female unemployment is lower in the United States compared to non-U.S. OECD (Organisation for Economic Co-operation and Development) countries with government-supported leave (Figure 4). Labor-force participation is higher for U.S. women than it is in the OECD overall. This pattern is consistent with the theory that government-supported leave increases discrimination against women in the labor market.

Ok, so the other OECD countries provide leave and have worse female unemployment. But should we really be lumping those countries together? Is it necessarily true that if your country has more generous leave policies than the U.S., your female unemployment rate will be worse?

From the source she cites, here’s the U.S. unemployment rate for the 26 years from 1990 to 2016:

But what about, say, the Nordic countries? Here’s Norway:

Consistently lower. Iceland?

Same. Denmark?

Sometimes higher, sometimes lower. Sweden?

Usually higher. Note, though, that it’s not much different for men and women, and that women tend to have lower unemployment rates. 

So why present this chart? What does it tell us? Calder even concedes that it “does not constitute strong standalone evidence,” but that’s something of an understatement. Thinking you can draw conclusions about paid family leave solely from the wildly varying unemployment rates of all other OECD countries compared to the United States… well, I think this chart is solely designed to impress the people who are impressed when they see charts. Now as for Calder’s contention that U.S. women’s higher labor-force participation “is consistent with the theory that government-supported leave increases discrimination against women”: yes, it’s consistent with that theory. It would also be consistent with a ton of other theories, like the theory that U.S. women work more because they have to. And Calder ignores any study showing that paid family leave policies increase women’s participation in the workforce.

I have spent a long time on this particular paper. I think the very fact that the Cato Institute put it out tells you all you need to know about their standards of rigor and their lack of interest in genuinely answering the empirical question of whether the free market is working or not. I’ll bet you I could go through a dozen of their other papers and find the exact same tendencies. (I really, really don’t want to have to do this though.)

But I do want to give you another example of Cato Institute “policy analysis.” Here is an op-ed about Democrats and Obamacare from this morning’s Wall Street Journal, by Cato’s director of health policy studies, Michael F. Cannon. I’ll give it to you without much context so you can have the same “Huh?” experience I had while reading: 

By the narrowest of margins, the U.S. Senate rejected legislation Wednesday that would have subjected patients with expensive illnesses to soaring premiums, canceled coverage and medical bankruptcy. You might expect such legislation to have been introduced by Republicans and defeated by Democrats, but you’d be wrong. Democrats sought to deny care to the sick. Republicans stopped them. The legislation would have rescinded new Trump administration rules that expand consumer protections in short-term health-insurance plans. Exempt from ObamaCare’s hidden taxes, mandates and regulations, these plans often cost 70% less and offer a broader choice of health-care providers. The new rules, which took effect Oct. 2, will expand coverage to an estimated two million Americans who would otherwise go uninsured… All 49 Senate Democrats plus Maine Republican Susan Collins voted to rescind these consumer protections and subject sick patients to canceled coverage and denied care by reverting to the Obama rule. The Democrats’ claim that short-term plans hurt patients with pre-existing conditions “doesn’t exactly make sense,” the Washington Post reports, because ObamaCare remains an option for those who want it. The Democrats’ real motivation was fear that ObamaCare would prove unpopular in a head-to-head contest with free-market health insurance. So they tried to sabotage the competition. If protecting their big-government scheme meant stripping coverage from two million Americans and denying care to sick patients, Democrats said so be it.

Now, I’ll admit, I hadn’t actually followed the tussle over this piece of legislation and didn’t know about it, in part because Democrats are generally horrible at actually publicizing their agenda. And I didn’t know much about the issue of short-term health insurance plans. Cannon implies that Democrats want to deny care to the sick merely because it would make their disastrous Big Government policies look good. That sounds very bad indeed. I checked into it.

Cannon has to know that he has not given a fair statement of the Democrats’ “real motivation.” The dispute over short-term plans is this: These are cheapo plans that do not meet the Affordable Care Act’s definition of minimal coverage. The ACA prevented insurance companies from establishing benefit caps, prohibited them from turning down people with pre-existing conditions, and required that they cover prescription drugs, mental health, prenatal care, etc. The short-term plans, which have been derisively called “junk” plans, don’t have to meet these requirements. The Trump Administration is trying to extend the length that people can use the “short-term” plans. The argument made by them, and libertarians like Cannon, is that people should get the free choice to purchase cheap, limited-benefit plans if they want to. Cannon’s “two million” figure is the number of uninsured people who would be expected to buy cheap, limited coverage plans if they could.

The reasons Democrats oppose this are quite obvious. First, we lefties are quite skeptical of the libertarian concept of “free” consumer choice, because we know how aggressively and misleadingly companies often market their products. People often don’t understand the “fine print” and insurance companies have a strong financial interest in tricking people into thinking they have more coverage than they do and then screwing them when they make a claim. Here, for example, is a Consumer Reports story about one woman’s experience with one of the short-term, ACA-exempt plans:

When Jeanne Balvin had emergency surgery for diverticulitis in June 2017, her short-term health insurance plan—a policy she bought instead of more comprehensive insurance—covered most of the bills after she paid a $2,500 deductible. But when she landed back in the hospital with an abdominal infection a few weeks later, she says her insurance company, UnitedHealthcare, wouldn’t cover the charges—and then canceled the three-month policy she had just renewed. UnitedHealthcare said the infection was a pre-existing condition related to the diverticulitis and wouldn’t be covered under terms of the contract. And when Balvin, 61, was hospitalized a third time at the end of July—this time for a blood clot probably caused by inactivity following the hospitalizations—she had no insurance at all, leaving her with $97,000 in hospital bills.

This is precisely the situation ObamaCare was passed in order to prevent. And the more you expand the use of these plans, the more of this will happen. Democrats also oppose the plans because they threaten to destroy the entire structure of the ACA insurance market: Young healthy people will buy cheapo plans, the group of people with ACA-compliant plans will be disproportionately sicker, those prices will rise, and the whole ACA scheme will collapse. “Exactly,” Cannon would say. “Like I say, they want to protect the ACA even if it means denying people the ability to choose cheaper plans.” But there is an actual reason they want to protect the ACA, and it’s not just because they get wet when they think about Big Government. It’s because the “free market” situation that existed before the ACA, and that Cannon/Trump would like to return to, produces inhumane outcomes. People get denied for pre-existing conditions, go deep into debt when they exceed their benefit maximum, and have significant gaps in their coverage.

That’s why the American Medical Association strongly opposes the Trump plan. Here’s an excerpt from a brief filed by the AMA, The American College of Physicians, The American Osteopathic Association, The American Academy of Family Physicians, The American Academy of Pediatrics, The American College of Obstetricians and Gynecologists, the HIV Medicine Association, and the Medical Society of the District of Columbia:

Although [we] represent a variety of different specialties, [we] all share the goal of improving healthcare in the United States. A key part of this mission is providing as many of [our] members’ patients as possible with affordable, meaningful health coverage. As courts have recognized again and again, this is the same goal as underpins the Affordable Care Act itself. The 2018 Short-Term, Limited Duration Insurance (STLDI) Rule is antithetical to this shared goal. The Rule will be devastating to the health, well-being, and pocketbooks of millions of Americans—and disproportionately so for women, children, and the chronically ill. To understand why, this Court need look no further than the comments Defendants received during the rulemaking process. Approximately 12,000 commenters submitted responses to the proposed rule. Remarkably “[n]ot a single group representing patients, physicians, nurses or hospitals voiced support” for the proposal… [T]he coverage gains of the past decade should be maintained. Central to this principle is ensuring meaningful coverage, assisting individuals with low-incomes or unusually high medical costs in obtaining health insurance coverage and meeting cost-sharing obligations, and ensuring the continuation of essential health benefit (EHB) categories and their associated protections against annual and lifetime limits and out-of-pocket expenses. Affordability is also critical, as is stabilizing and strengthening the individual health insurance market, maintaining key insurance market reforms under current law, and expanding choice of health insurance coverage to best meet individual needs. The proposed rule fails to comply with these important principles, and in fact, would reverse progress that has been made in expanding meaningful coverage to millions of previously uninsured Americans.

Michael F. Cannon did not mention the AMA or other organizations in his op-ed. I suppose he thinks they are just so ideologically committed to Big Government for the sake of Big Government that they don’t mind if they make millions of their patients sick. But let’s consider for a moment the possibility that they know what they’re talking about.

At the very least, a think tank’s “director of health policy studies” should not bury strong counter-arguments from experts. But even in Cannon’s longer discussions of this issue, he at no point takes the objections seriously. He says that the short-term plans “free consumers to avoid coverage they do not want” and cites studies showing that millions would choose the short-term plans, saying that this “would produce significant social-welfare gains” (because the more people’s market preferences are satisfied, the better). But the whole argument of the critics is that the gains of “increased choice” and “letting health people opt to have less coverage” are outweighed by the costs of destroying the whole framework by which the Affordable Care Act has succeeded in getting so many people into plans with strong basic coverage. That’s the AMA’s argument, and if you’re going to be a serious public commentator on health policy, you can’t pretend to readers of the Wall Street Journal that the argument doesn’t exist.

So, the Cato Institute: Never trust them. Because these sorts of people are just not going to represent the issues fairly. They’re not going to report the conclusions of experts accurately (Cannon cites the National Association of Insurance Commissioners to show that industry professionals agree with him, and ignores the many medical practitioners who don’t). They’re going to pick quotes from studies that sound good, and unless you have academic journal access and can check for yourself, you’ll never know whether the studies contained other important findings. In an academic setting, this would border on malpractice. These people may have charts and statistics and degrees, but they are just not scholars. They already think that laissez-faire capitalism is correct, and they think it because it’s what their economic theory tells them must be true, and so any empirical evidence to the contrary will be ignored. (Again, if you truly won’t be satisfied until I go through a dozen more of their publications and show all the same tendencies, I can do it. But just try it for yourself.)

Honestly, I’m just exhausted by the existence of these people. It’s so hard to figure out the truth in a complicated world. I wish I didn’t have to deal with a whole army of propagandists who couldn’t care less about neutral scholarly inquiry. They might be right, or they might be wrong, but I’m always going to have to investigate for myself since their publications are useless. So I generally don’t bother responding to the Cato Institute because it’s always going to require such a battle: They will do everything in their power, they will strain themselves to their limit, to prove that regulation is bad and laissez-faire is good. (This is, by the way, why Charles Koch has given them so much money, and the Institute used to be called the Charles Koch Foundation.) All these “economic liberty” places are the same. The Free Enterprise institutes, and the Economic Education funds. You could pile up mountainous evidence of market failures or regulatory success, and they’d die before they conceded the point.

I know everybody on the left will greet my conclusion here with a yawn: Breaking news—the Cato Institute care about their free market ideology more than they care about facts! But it’s still so unbelievable to me that this can be an influential think tank that anyone takes seriously. It’s worth remembering that the actual profession of economics is far less rigidly committed to laissez-faire than the “economistic” commentators who serve as the field’s public face. You can see the difference by looking at how Cato writes about these issues versus the scholars Cato actually cites. The real papers have nuance. They are scholarship. This is bullshit, and can never be trusted.

If you appreciated this article, please consider making a donation, purchasing a subscription, or supporting our podcast on Patreon. Current Affairs is not for profit and carries no outside advertising. We are an independent media institution funded entirely by subscribers and small donors, and we depend on you in order to continue to produce high-quality work.