Minimum wage, maximum ideology

Last week, Seattle voters approved a $15 minimum wage, the highest in the country. This rang the bell for another round of the debate about whether the minimum wage is a Good Thing. Simply stated, the question is whether increased wages for some workers will cause other workers to lose their jobs.

It’s hard to understand these arguments without a bit of economic theory. I’m strictly an amateur here, but the principles are simple and can be stated without graphs or equations in three simple rules:

1  A business will sell fewer units of whatever it produces as the unit price goes up (and vice-versa); that’s why Fords outnumber Ferraris.

2  As the business increases the number of units produced, the per unit cost will generally fall (“increasing returns to scale”).

3  At some point, however, per unit costs will start to rise again as the business’s plant becomes too small for its output or the cost of inputs (such as labor) rise to meet the increased demand.

All of these pressures on the business balance out at a sweet spot where the business’s cost of producing one more unit just equals the revenue from that unit. At that optimum output the business maximizes its profit: Make one less unit and revenue will fall by more than cost (lower profits), make one more and costs will rise by more than revenue (again, lower profits).

Finding that optimum output is largely a trial-and-error process, but most businesses get it nearly right for the simple reason that the ones that didn’t are no longer around. It’s a jungle out there.

Raising the minimum wage may force a business to pay some of its workers more. If you pay your workers more, however, then it will cost you more to produce each unit. Which means that the sweet spot—the output level that maximizes profit—will move to a lower output. But as the business reduces output to maximize profit, it will need fewer workers for the reduced output. (The business may also invest in labor-saving machinery that wouldn’t have made sense when wages were lower.)

The theory, unobjectionable in itself, doesn’t tell you the size of the effects. Opponents of the minimum wage argue that the main effect will be lower employment, proponents that the main effect will be lower profits. Economists have done studies—lots of studies—on the effects of a minimum wage, with varying conclusions. Each side gets to cite the studies it likes.

Economists don’t have the luxury of being able to set up experiments where all the variables can be controlled. Occasionally they get “natural experiments,” such as when two bordering states set different minimum wages. But neighboring states can differ in all sorts of ways. And the economists don’t get to compare a $1/hr minimum wage with a $20/hr minimum wage. More like $9 vs $10. With such small differences, general statistical noise is likely to mask any unemployment effects.

Nonetheless, I think most of us can agree that at some level a minimum wage must increase unemployment; if Seattle had raised the minimum wage to $50/hr (about $100,000 a year), a considerable number of Seattleites would probably find themselves unemployed. Sensible minimum-wage advocates understand this, and see the problem as where to set the number. They want it high enough to help people (in the way a $1 minimum won’t) but low enough so that it won’t put substantial numbers of people out of work (as a $50 minimum would). I’ve seen this phrased as finding the “tipping point” below which there is little or no increase in unemployment and above which the increase is substantial. It seems unlikely, however, that there is such a break point, just a smooth rise or fall. Any rise (or fall) in the minimum wage is likely to increase (or decrease) unemployment to some extent; it’s all a question of how much.

Economic theory tells us that any attempt to set prices—including the price of low-skilled labor—will lessen economic efficiency; that is, the value of the goods and services the economy produces will be less than if the prices were set purely by market forces. But even a well-run society will accept market interventions if it judges that the non-efficiency benefits will outweigh the efficiency costs. Consider the income tax: The economic damage caused by people shifting activities to improve their tax position is huge. But we have to finance the government somehow, and income taxes seem the most practicable way. (Actually, a poll tax—that is, a tax that is the same amount for every adult regardless of income—would not affect economic efficiency, but such a tax is a political non-starter.)

That a minimum wage may hurt the economy doesn’t settle the issue; it all depends on the balance of costs and benefits. If Seattle’s adopting a $15 minimum wage will only increase unemployment slightly, then we might favor it. But if it causes lots of low-wage workers to lose their jobs, then it may not be such a good thing. Unfortunately, no one knows which it will be; the effects will be hard to measure after the fact, let alone in advance. Moreover, the effects on one Seattle business may be completely different from the effects on the business across the street, and the effect of a Seattle minimum wage different from the effect of a Tacoma, Topeka, or Tallahassee minimum wage.

Unfortunately, the minimum wage has become a test of ideological purity. Liberals, who are happy to intervene in the market, are all for it, while conservatives, who like market solutions, are solidly against. No one wants to hear that “it all depends.” Nor are purists troubled by the occasional inconsistency: Some of the same people who deny that a minimum wage can throw a substantial number of people out of work will simultaneously argue that low-wage foreign workers are causing American job losses. You can’t have it both ways. If lower labor costs can pull a factory from Massachusetts to Alabama (or Alma Ata), then a rise in the minimum wage can push it there. Similarly, conservatives can’t argue that a minimum wage will substantially increase unemployment, then turn around and claim that jobs don’t migrate to low-wage jurisdictions.

Where do I come out on all of this? Generally, I oppose the minimum wage. I can understand someone supporting it if (a big “if”) the effect on wages is large and the effect on unemployment small. But we seldom know what those effects will be, so supporting a minimum wage seems like a kind of feel-good politics—you support a policy to show the purity of your principles without asking if it really does more harm than good.

I would like to do something for the poorest in our society, but it has to be something that is likely to work. Some public policy wonks argue that an expanded earned income credit (EIC) can raise the compensation of low-wage workers without increasing unemployment. I’m still trying to figure out how I feel about the EIC; If I do, I’ll let you know in a later post.

—Stan

A first rough draft of marijuana legalization

Almost everyone I’ve known who’s been involved in events important enough to make the newspapers has been disappointed, even angered, by the superficiality of the coverage. It’s inevitable: Information is being conveyed to a general public that knows less about the subject than the reporters, who in turn know less than the experts and those directly involved.

Reporters like to say that “News is only the first rough draft of history” (attributed to the journalist Alan Barth, writing in 1943). They like the connection to history, but I tend to focus on the “first rough draft” part; generally, it’s silly to complain about the inadequacies of newspaper stories. Still, there are better and worse first drafts. Consider the story in the June 1 New York Times under the headline “After 5 Months of Sales, Colorado Sees the Downside of a Legal High.”

I paused to read the article because I favor marijuana legalization and because my younger boy now lives in Boulder (but doesn’t inhale). I found the article frustrating, not because of what it said about marijuana but because of what it said about journalism.

To begin with, when the headline says “sees the downside,” then there should be a real downside. “Sees” is one of those words that implies the truth of what’s seen. But the article presents little hard evidence of a downside and—after a time—admits as much: The subhead on the jump reads “Anecdotes, but scarce data, about the effect of a marijuana law.”

After an opening statement about “a series of recent problems as cautionary lessons for other states flirting with loosening marijuana laws,” the reporter moves to a sensational incident: Richard Kirk,, after eating marijuana-laced cookies, began raving about the end of the world, grabbed a handgun from a family safe, and killed his wife, Kristine. There’s a photo of Mrs. Kirk’s coffin being loaded into a hearse. (You later read that Mr. Kirk may also have consumed prescription medication for back pain.)

The article continues: “Some hospital officials say they are seeing growing numbers of children and adults sickened by potent does of edible marijuana. Sheriffs in neighboring states complain about stoned drivers streaming out of Colorado and through their towns.”

This is followed by a quote that “by any measure, the experience of Colorado has not been a good one unless you’re in the marijuana business,” which you then discover comes from the executive director of a group opposing legalization. I understand that reporters have to go to spokespersons, but it would be better if they noted the source before the quote, so that readers would know that what they are about to read is, as evidence, worthless. (There are opposing statements by legalization proponents later on, but many readers may not get that far.)

You reach the truly important fact in paragraph five: “Despite such anecdotes, there is scant hard data. Because of the lag in reporting many health statistics, it may take years to know legal marijuana’s effect—if any—on teenage drug use, school expulsions or the number of fatal car crashes.”

The article does contain some valuable information: Since January, when the State Police began tracking the number of stoned drivers, “marijuana-impaired drivers have made up about 1.5% of all citations for driving under the influence of drugs or alcohol.” Also, “crime in Denver is down by about 10 percent, though it is impossible to say whether changes to marijuana laws played any role in that decline.” The head of the Drug Enforcement Agency, Michele Leonhart, told a U.S. Senate panel that Kansas officials “had tallied a 61 percent increase in seizures of marijuana that could be traced to Colorado.” However, the article notes that “according to the Kansas Highway Patrol, total marijuana seizures fell to 1,090 pounds from 2,790 pounds during the first four months of the year, a 61% decline.” (Could Ms. Leonhart have mixed up her “increase”s and “decline”s?) Finally, “so far this year, nine children have ended up at Children’s Hospital Colorado in Aurora [a Denver suburb] after consuming marijuana, six of whom got critically sick. In all of 2013, the hospital treated only eight such cases.” (It’s not clear whether the “such cases” refers to merely consuming marijuana or getting critically sick.)

One implication of the article should give even legalization advocates pause: Apparently it’s possible to pack quite a wallop into a marijuana cookie. As a result, the article informs us, Colorado has tightened labeling and packaging rules for edible marijuana, and is considering limiting the amount of THC (the psychoactive component of marijuana) that can be packed into a single cookie. “Even supporters of legalization … say Colorado needs to pass stricter rules about edible marijuana.”

I’m not saying the story was bad. It had some useful information, especially if you read through to the end. And I’m not opposed to anecdotes; as the saying goes, the plural of anecdote is data. (Sometimes turned into “the plural of anecdote isn’t data”; I prefer the first version.) But to create a “story,” the reporter (and whoever supplied the headline) had to create an interesting narrative, and “Anecdotes, but scarce data” doesn’t make an enticing come-on. So we have an opening that strongly implies that there are real problems with legalization, followed by attempts to walk that conclusion back.

Even the anecdotes seem weak. The domestic shooting (and the report later in the article of a young man who leapt to his death off a hotel balcony after eating a marijuana cookie) are examples of a logical fallacy so ancient it even has a Latin name: post hoc ergo propter hoc (“after this, therefore because of this”). And there are lots of questions: How much marijuana, and what else, had the people ingested? (No toxicological evidence is presented.) Were there any pre-existing psychological conditions? (Again, no information.) And most important, but most difficult to judge, the counterfactual: If they hadn’t taken marijuana, what other drug(s) might they have consumed, and with what consequences? People are going to get high, so the important question is: Will more get high, and to worse effect, if marijuana is substituted for other psychoactive media?

Anything that enters your body, from penicillin to peanuts, can harm you. It all depends on the particular interaction. I don’t want to play down the dangers, so here’s an anecdote from my own experience: A million years ago, when I was in college, some of my friends heard that you could get high on a common type of flower seed (I’ve forgotten which one). They bought seed packets at a garden store, chowed down, and for the most part had an enjoyable experience. But one of my friends suffered a terrifying psychotic episode—his first, as far as we know—that led to his temporary hospitalization and that affected him for some time.

There are dangers to psychoactive drugs. No one should be surprised by that. We’ve had enough experience (not merely anecdotal, although the anecdotes will do) with alcohol to have a good idea of its downsides, but have decided that legalization is better than prohibition. But weighing the pros and cons for marijuana, and most of the other stuff we might eat or inhale, requires going beyond the merely anecdotal. That’s not something you’re likely to get from a first rough draft of history.

—Stan

Lance and PEDs

I took up bicycling in my early 50s, and in my early 60s began to compete in seniors races. I’m not competitive nationally but I’ve had some local success. That’s enough for my non-cycling friends to anoint me an expert on performance enhancing drugs (PEDs).

One might have considered doping (including reinfusing one’s own blood and other interventions that don’t technically involve “drugs”) to be just another training method, like weight- or high-altitude-training or special diets. But that’s not the way many people see it.

Much of the anger about PEDs is fueled by our general societal demonization of recreational drugs (other than alcohol). I’m on the other side: I favor legalizing marijuana and any other drug that is no more harmful than alcohol—a low threshold given alcohol’s well-documented dangers.. So I tend to be more tolerant about PEDs.

Once one puts aside the anti-drug hysteria, there seem to be two main arguments for banning PEDs. First, they are a form of cheating, in that athletes not using PEDs can’t effectively compete against dopers. This seems to be true for professional cycling. (The effects of PEDs on other sports are not as clear. Ollie, who has forgotten more about baseball than I’ll ever know, has at times seemed dubious about their effect on baseball performance.)

It’s likely that in the 1990s, almost all the cyclists competing in the Tour de France were taking some form of illegal PEDs because without the PEDs they would have been left behind. But if everyone is taking PEDs it hardly seems like cheating. Taking PEDs doesn’t give you an unfair advantage; it’s a level playing field.

When Lance Armstrong was stripped of his seven Tour de France titles, the Tour’s promoters did not elevate any of the other riders, probably because it would have been too embarrassing. Many of the riders who placed second or third behind Lance (and the two preceding winners) have also been suspended for PED use.

The other argument is that riders were pressured into taking PEDs that may ultimately prove dangerous to their health. There certainly was pressure, but it came mainly from the demands of the sport: If you didn’t take PEDs you weren’t going to have a career in professional cycling. But you could say as much about many other aspects of cycling, such as attacking a dangerous downhill: If you don’t take the risks, you’re left behind.

There are lots of different PEDs, and it seems likely that they present different risks; steroids may be more dangerous, EPO less. But because PEDs are illegal, there isn’t much reliable research on their long-term health effects. I’m willing to believe that at least some of them are dangerous—lots of things are—but are they more dangerous than cigarettes or alcohol? We’ve had a lot of experience with public panic over the dangers of new drugs (Ecstasy, for example) that are less dangerous than alcohol. It’s hard to separate the bits of honest evidence from the hysteria. (On drugs and drug hysteria generally, take a look at David Nutt, Drugs Without the Hot Air: Minimising the Harms of Legal and IllegalDrugs.)

In any case, to worry about the safety of PEDs is to strain at a gnat. Even without PEDs, cycling is extremely dangerous when compared to baseball, basketball, or football. It may even be more dangerous than Formula 1 auto racing, where the last fatality was Ayrton Senna in 1994. Since then there have been at least two deaths in the three annual Grand Tour cycling races, plus deaths in other races and in training, where cyclists have to mix it up with automobiles. Safety improvements have been slow in coming. Helmets only became mandatory in European racing in 2003 (over the objections of the riders themselves). Race promoters design courses with fast curvy downhills, final mass sprints around sharp corners and down narrow streets, and other challenges to life and limb. And races don’t stop for rain, which makes the roads slick and vision sketchy. A typical day will see multiple crashes, with riders—who wear no protective clothing other than a helmet—being dumped on hard pavement and unyielding curb stones at up to 45 miles an hour. Most professional racers have broken their collar bone at least once.

What’s more, the ban on PEDs has made things more, not less, dangerous. Because they are banned, PEDs are administered by people who are not highly qualified. Blood transfusions are given in hotel rooms, and when things go wrong riders cannot go to a hospital or consult a non-team doctor. (There are vivid descriptions of these problems, and cycling’s drug culture, in Tyler Hamilton’s TheSecret Race: Inside the Hidden World of the Tour de France.)

People rightly recoil at the hypocrisy of it all. Until a few years ago nearly every professional cyclist used PEDs, and just about everyone involved in cycling knew it. Armstrong was particularly vociferous in his denials and nasty to his accusers, but everyone lied. For in the unlikely event that a rider was caught by the hit-or-miss doping tests, fessing up would have seen him barred from a job he loved, typically for two years. A cyclist’s professional life is short, and most cyclists, especially in Europe, are blue-collar types without a lot of attractive career opportunities. In that milieu, lying becomes just another professional skill, appreciated by your co-workers whatever the general public might think.

All this may be changing as drug testing improves. If PEDs can be reliably detected, then use will decline because riders will learn that you can have a career even if you don’t take drugs (and probably not have a career if you do). In this changed environment, being drug-free becomes the new job requirement, like wearing a helmet. Anyone who then uses PEDs and somehow gets through the net would be both very lucky and clearly cheating.

If that’s the current situation, it’s only happened quite recently. Before the last few years there was little reprehensible about PED use—certainly, users were not taking advantage of non-users, of whom there were precious few. But during the brief following interregnum when drug tests gradually improved, the rules and the biking culture were in flux, with no clear moral rights and wrongs and little to guide individual riders. It seems mean-spirited for us to now go all sanctimonious and condemn a cyclist for mistiming his adjustment to an evolving situation.

Armstrong, like Barry Bonds or Tiger Woods, was an extraordinarily gifted athlete and, by all reports, a somewhat flawed human being. There are a lot more of the latter than the former.

—Stan