Category Archives: Misunderstood

Three Misunderstood Things 7-24-2017

This week: census, environmental regulations, coal jobs


I. The census

What’s misunderstood about it: How can counting people be a partisan issue?

What more people should know: A lot rides on the census. The Census Bureau knows it gets the answers wrong, but Republicans have a partisan interest in not letting it do better. In 2020, it’s being set up to fail.

*

When the Founders wrote the Constitution, they knew the country was changing fast. New people were pouring into America — some coming by choice and others by force. If Congress was going to represent these people into the distant future, it would have to change as the country changed. So somebody would have to keep track of how the country was changing. That’s why Article I, Section 2 says:

The actual Enumeration shall be made within three Years after the first Meeting of the Congress of the United States, and within every subsequent Term of ten Years, in such Manner as they shall by Law direct.

Congress has implemented that clause by setting up the Census Bureau, which tries to count everyone in America in each year that ends in a zero. You can look at this as a rolling peaceful revolution: Via the census, states like Virginia and Massachusetts have gradually surrendered their founding-era power to new states like California and Texas.

No doubt you learned in grade school that counting is an objective process that produces a correct answer — the same one for everybody who knows how to count. But in practice, when a bunch of people count to 325 million, agreement starts to break down. Now imagine that you’re counting a field full of 325 million cats, most running around and jumping over each other, and a few actively hiding from you. How do you come up with an answer you have faith in?

That’s the Census Bureau’s fundamental problem: Americans won’t stand still long enough to be counted, and some are actively suspicious of anybody from the government who comes around asking questions. Inevitably, then, not everybody gets counted, and some people get counted more than once. This is not a secret; the Census Bureau admits that it gets the wrong answer.

That might not be so bad if the errors were random, but they’re not. Basically, the more stable your life is, the more likely you are to be counted correctly. If, for example, you’re still living in the same house with the same people that a census worker counted ten years ago, they’re going to count you again. But if you’re sleeping on your friend’s couch for a few weeks while you’re waiting for a job to turn up, and thinking about moving back in with Mom if you can’t find one, then you might get missed.

Stability isn’t a randomly distributed quality. The LA Times spells it out:

The last census was considered successful — that is, the 2010 results were considered to be within an acceptable margin of error. But by the Census Bureau’s own estimates, it omitted 2.1% of African Americans, 1.5% of Latinos and nearly 5% of reservation-dwelling American Indians, while non-Latino whites were overcounted by almost 1%. The census missed about 7% of African American and Latino children 4 or younger, a rate twice as high as the overall average for young children.

But that raises an epistemological question: How do you know your count is wrong if you don’t have a correct count to compare it to? And if you have that correct count, why not just use it?

The answer to the first question is statistics. Imagine, for example, that you’re trying to count all the species that live in your back yard. You go out one day and count 50. Then you go out longer with a bigger magnifying glass and find 10 more. Then the next couple of times you don’t find anything new. But then you find two. Are you confident that’s all of them now? What’s your best guess about how many are really out there?

Now extend that to every yard in the neighborhood. Imagine that after each household does its own count, you all converge on one yard for a more intensive search than you’d be willing to do on every yard. That search finds even more new species. Now how many do you think you missed in the other yards?

Statisticians have thought long and hard about questions like that, and have a variety of well-tested ways to estimate the number of things that haven’t been found yet. If you apply those techniques to the census, you get more accurate estimates of the total.

So why not just use those estimates? Two reasons:

  • It sounds bad: Ivory-tower eggheads are using a bunch of mumbo-jumbo Real Americans can’t understand to invent a bunch of blacks and Hispanics that nobody has ever seen.
  • Republicans have a partisan interest in keeping the count the way it is.

The Census determines two very important things: how many representatives (and electoral votes) each state gets, and how hundreds of billions of dollars in federal money for programs like Medicaid and highway-building get distributed among the states. The miscount gives more power and money to mostly white (and Republican) states like Wyoming and Kansas, and less to a majority non-white (and Democratic) state like California. Within a state, Republican gerrymandering works by crowding Democratic-leaning urban minorities into a few districts, leaving a bunch of safely Republican rural and suburban districts. That minority-packing is even easier to do if a chunk of those people were never counted to begin with.

The 2020 census is already headed for trouble. The Census Bureau is being underfunded, taking no account of the fact that it has more people to count than last time. Plans to modernize its technology went badly. And it is currently leaderless: The bureau chief resigned at the end of June, and Trump has nominated no one to replace him.

So we’re set up for an even bigger uncount of minorities this year. And that’s got to make Paul Ryan happy.

II. Environmental regulations

What’s misunderstand about it: Many people believe that a clean environment is a costly luxury.

What more people should understand: Externalities. That’s how well-designed environmental regulations can save more money than they cost.

*

Nobody should come out of Econ 101 without an understanding externalities — real economic costs that the market doesn’t see because they aren’t borne by either the buyer or the seller.

Pollution is the classic example: Suppose I run a paper mill, and I use large quantities of chlorine to make my paper nice and white. At the end of the process I dump the chlorine into my local river, because that’s the cheapest way for me to get rid of it. Because I use such an inexpensive (for me) disposal process, I can keep my prices low. That makes me happy and my customers happy, so the market is happy too. Any of my competitors who doesn’t dump his chlorine in the river is going to be at a disadvantage.

The problems in this process only accrue to people who live downstream, especially fishermen and anybody who wants to swim or eat fish. They suffer real economic losses — losses that are probably much bigger than what I save. But since their loss is invisible to the paper market, nothing will change without the some outside-the-market action — like a government regulation, a court order, or a mob of fishermen coming to burn down my mill.

Now suppose the government tells me I have to stop dumping chlorine. I have to find either some environmentally friendly paper-whitening technique or a way to treat my chlorine-tainted wastewater until it’s safe to put back into the river. Either solution will cost me money, and I will have no trouble calculating exactly how much. So you can bet there will be an article in my local newspaper (which now has to pay more for the newsprint it buys from me) about how many millions of dollars these new regulations cost. The corresponding gains by fishermen, riverfront resort owners whose properties no longer stink, and downstream towns that don’t have to get the chlorine out of their drinking water — that’s all much more diffuse and hard to quantify. So the newspaper won’t have any precise number to weigh my cost against. Chances are its readers will see the issue as money vs. quality of life. They won’t realize that the regulations also make sense in purely economic terms.

That’s an abstract and somewhat dated example, but similar issues — and similar news stories — appear all the time. The costs of new regulations are borne by specific industries who can calculate them exactly, while the benefits — though very real — are more diffuse, and may accrue to people who don’t even realize they’re benefiting. (Companies are very aware of what they’ll have to spend to take carcinogens out of their products, but nobody ever knows about the cancers they don’t get.) But that doesn’t mean that the benefits aren’t bigger than the costs, even in dollar terms.

The best example from my lifetime is getting the lead out of gasoline. If you were alive at the time, you probably remember that the new unleaded gasoline cost a few cents more per gallon. Spread over the whole economy, that amounted to billions and billions. What we got out of that, though, was far more than just the vague satisfaction of breathing cleaner air. Without so much lead in their bloodstreams, our children are smarter, less violent, and less impulsive. The gains — even in purely material terms — have been overwhelmingly positive.

III. Coal jobs.

What’s misunderstood about it: What happened to them? Environmentalists are often blamed for destroying these jobs.

What more people should know: No doubt environmentalists would kill the coal industry if they could. But the real destroyers of coal jobs are automation and competition from other fuels.

*

Coal miners are the heroes of one of the classic success stories of the 20th century. Mining was originally a job for the desperate and expendable, but miners were among the first American workers to see the benefits of unionization. Year after year, coal mining became safer [1], less debilitating, and better paying, until by the 1960s a miner no longer “owed his soul to the company store“, but could be the breadwinner of a middle-class family, owning a home, driving a nice car or truck, and even sending his children to college. Sons and daughters of miners could become doctors, lawyers, or business executives. Or if they wanted to follow their fathers into the mines, that promised to be a good life too.

However, the total number of coal-mining jobs in the United States peaked in 1923.

Was that because Americans stopped using coal? Not at all. Coal production kept going up for the next 85 years.

The difference was automation. Mines employed three-quarters of a million men in the pick-and-shovel days, but better tools allow 21st-century mines to produce more coal with far fewer workers.

If you take a closer look at that employment graph, you’ll notice a hump in the 1970s, when coal employment staged a brief comeback. That corresponded to the Arab Oil Embargo of 1973 and the increased oil prices of the OPEC era. For decades after that, coal was the cheaper, more reliable energy source. Americans who dreamed of energy independence dreamed of coal. In a 1980 presidential debate, candidate Ronald Reagan said:

This nation has been portrayed for too long a time to the people as being energy-poor, when it is energy-rich. The coal that the President [Carter] mentioned — yes, we have it, and yet 1/8th of our total coal resources is not being utilized at all right now. The mines are closed down. There are 22,000 miners out of work. Most of this is due to regulation.

However, all that changed with the fracking boom. Natural gas is now the cheaper fuel. Meanwhile, the price-per-watt of renewable energy is falling fast, and is now competitive with coal for some applications. So if a utility started building a new coal-fueled plant now, by the time it came on line a renewable source might be more economical — even without considering possible carbon taxes or environmental regulations.

The dirtiness of coal is a huge externality (see misunderstanding II, above), so regulations disadvantaging it make good economic sense. Looking at the full cost to society, coal is the most expensive fuel we have, and should be phased out as soon as possible.

Statements like that make good fodder for politicians (like Trump or Reagan) who want to scapegoat environmental regulations for killing the coal industry. However, dirty coal is like the obnoxious murder victim in an Agatha Christie novel: Environmentalists are only one of the many who wanted it dead, and other suspects actually killed it.


[1] The number of coal-mining deaths peaked at 3,242 in 1907. In 2016 that number was down to 8.

Three Misunderstood Things, 7-17-2017

This week: healthcare costs, the “Biblical” view of abortion, and sanctuary cities.


I. Cutting healthcare costs.

What’s misunderstood about it: Liberals and conservatives both talk about cutting costs, but mean different things. Liberals are usually talking about cutting the cost to society as a whole, while conservatives focus on cutting the cost to the federal government. Either side might be talking about cutting the cost to certain individuals.

The right follow-up question: When a proposals claims to “cut healthcare costs”, are those costs going away, or just being passed on to someone else? Or did that money pay for some needed care that someone is now going to do without?

*

Nearly everyone agrees that American healthcare is too expensive. Americans spend far more on healthcare, both per capita and as a percentage of GDP, than any other country in the world. That might be fine if our extra money bought us better health, but in fact the reverse is true: Our life expectancy at birth is similar to much poorer countries like Costa Rica and Cuba, and on average Americans die four years sooner than the Japanese or the Swiss.

So cutting healthcare costs is a sure applause line for a politician. But what does it mean?

An win/win outcome would be to deliver the same or better care more efficiently and effectively: Hospitals make fewer mistakes and produce fewer unnecessary complications. Treatment gets targeted better, so that no one has to suffer through (or pay for) a test or treatment had no chance of helping. Drugs replace surgeries, and diet/exercise regimens replace drugs. Preventive care catches conditions before they become serious (and expensive). Environmental and job-safety and product-safety standards expose people to fewer health-threatening risks.

Admittedly, it’s not always obvious how to get those win/win results. ObamaCare made some small steps in this direction, but we really don’t know yet whether they’re working, and those changes may not survive the ObamaCare repeal process.

So most cost-cutting proposals are not about those win/win solutions. Liberals often try to offer the same treatment for less money by squeezing providers: cutting insurance companies out of the loop via single-payer plans, capping the prices that drug companies or hospitals can charge, or paying doctors less. Those are great ideas unless you’re an insurance company, a drug company, a hospital, a doctor, or a lobbyist for one of those powerful vested interests.

Conservatives often cut costs by getting somebody to do without healthcare they would otherwise want, usually rationing by cost: Everything is available if you can pay, but you might “choose” not to pursue some treatment that would bankrupt your family. Perhaps Americans (especially poor and working-class Americans) really do seek massive amounts of unnecessary treatments, and they would stop if only they had more “skin in the game“, but I haven’t seen that in my own life. What I have seen is my wife taking monstrously expensive drugs to keep her cancer from coming back. If we were poor and had to pay for them ourselves, it would be really tempting to cross our fingers and hope.

And finally, both sides talk about cutting costs by transferring those costs to somebody else. For liberals, “somebody else” is usually the government, or (passing the buck one step further) the taxpayer. For conservatives, it’s the individual — especially if he or she is unhealthy. Capping what the government is willing to put into Medicare or Medicaid, for example, may help the government control its budget deficit, but it doesn’t do anything to lower the need for treatment or the cost of providing it.

Similarly, letting individuals design their own (cheaper) health insurance — letting people opt out of insurance for care they won’t need, like prenatal care for men or geriatric care for young people — may lower some people’s individual expenses, but the total number of pregnancies and old people hasn’t changed. The cost of caring for them hasn’t gone away, it has just shifted to somebody else.

II. Christianity and abortion.

What’s misunderstand about it: The belief that a newly fertilized ovum has the full moral worth of a baby (or an adult) is often described as the “Christian” or even “Biblical” position.

What more people should know: The Bible says nothing about conception, and what it does say about fetuses and souls points in a different direction. The current ensoulment-at-conception dogma didn’t solidify among conservative Protestants until well after Roe v Wade.

*

Religiously, the question of whether abortion is murder comes down to when the fetus acquires a soul. Souls, after all, are the difference between murder and what butchers have done for millenia. (If you believe a chicken has an immortal soul, you really should be a vegetarian.) Anti-abortion arguments often (and usually inaccurately) point to the number of weeks before a fetus has a heartbeat or can feel pain, but such physical traits are just placeholders for a metaphysical trait that can’t be recognized in a secular setting like a legislature or a courtroom: the presence of a soul.

Unfortunately, microscopes and ultrasound machines didn’t exist when the Bible was being written, so scripture never mentions the miraculous moment when a sperm enters an ovum, nor gives a detailed description of fetal development. The observable sequence at the time was: sex, the woman shows signs of pregnancy, the fetus begins to move on its own, and birth. No one knew how to break the process down much finer than that, and apparently God never whispered His superior knowledge into anybody’s ear.

But anti-abortion Christians really, really want Biblical support for their position, so they thrust an enormous amount of interpretation onto a handful of texts that are either vague or really about something else. For example, Jeremiah 1:5, which you will occasionally see on billboards: “Before I formed you in the womb I knew you.” That might be a lyrical way of saying that God had been planning Jeremiah’s mission for a long time, or it might more literally say that Jeremiah’s soul existed before his conception, but it actually doesn’t say anything about precisely when that soul entered the body that was forming in his mother’s womb.

Which is not to say that the Bible is silent about souls entering bodies. There is a text — I believe it’s the only one — that quite explicitly describes a soul entering a body. But it doesn’t say what anti-abortion folks want to believe, so it seldom gets mentioned in abortion arguments. I’m talking about Genesis 2:7, which describes the creation of Adam.

And the Lord God formed man of the dust of the ground, and breathed into his nostrils the breath of life; and man became a living soul.

In other words: God formed Adam’s body completely, and then his soul entered that body with the breath. The obvious implication is that a fetus is soulless until it breathes. The Christian Left blog does a more detailed discussion of how this view aligns with other places where the Bible mentions pregnancy and miscarriage.

From the early Christian era through the Middle Ages, many Christian thinkers identified the soul with motion, and so held that it entered the fetus at the quickening, which was variously identified at anywhere from 40 to 120 days.

The Catholic Church has been against abortion in any form at least since the 1600s, when it began hoping for Catholic families to outpopulate Protestant families. But Protestant opinion varied widely, even among theological conservatives, until after abortion became a unifying conservative political issue in the late 1970s: The theology appears to have followed the politics, rather than leading it. The history of this discussion has been completely written over in the ensuing years. Slacktivist characterizes this process with a line from George Orwell’s 1984: “We have always been at war with Eastasia.”

As for why this corruption of church history and biblical interpretation is necessary, I believe the root issue is female promiscuity. Pregnancy is a great blessing to families that are ready to raise children, but traditionally it has also been the ultimate comeuppance for unmarried women who think they can have sex without consequences. When abortion is freely available, pregnancy becomes a much less effective threat for keeping women in line. That’s what social conservatives are really worried about, and why they don’t see effective birth control as a solution to the abortion problem.

III. Sanctuary cities

What’s misunderstood about them: What they are. In no American city, whether it identifies as a “sanctuary city” or not, do local officials actively prevent federal immigration officials from detaining or deporting undocumented immigrants. The issue is entirely about the extent to which local officials help ICE.

What more people should understand: Federalism. Under the Constitution, state or local government officials can’t block federal agents from enforcing federal laws, but they don’t have to help.

*

The word sanctuary evokes the idea that once you get there, you’re safe. That’s certainly how it worked for Esmeralda in The Hunchback of Notre Dame.

No city in the United States is a sanctuary for undocumented immigrants in that sense. Federal agents of Immigration and Customs Enforcement (ICE) can go anywhere in the country and arrest anyone they believe has broken federal immigration laws. Local officials can’t stop them, any more than they could stop the FBI from arresting terrorists or the Secret Service from arresting counterfeiters. (Legally, churches aren’t sanctuaries either, even though many of them — including the one I belong to (that’s me in the back, under the chandelier) — are supporting a sanctuary movement. So far ICE hasn’t been willing to break down church doors to haul somebody away, but fear of public opinion is all that stops them.)

However, unlike in some other countries, American state and local governments are not divisions or departments of the national government. The system we know as federalism prevents the national government from simply issuing orders to state and local officials. In particular, cooperation between various levels of law enforcement is mostly voluntary. (This is not an entirely liberal or conservative thing; conservatives want local police not to cooperate with federal gun laws.)

Vox has a pretty clear video explaining the situation.

Whenever local police arrest somebody, fingerprints are taken and submitted to the FBI, which then shares them with ICE. If ICE recognizes those fingerprints as belonging to someone they want to deport, they can send the local police a request to hold the person for an additional 48 hours, which gives ICE time to send out its own agents to make an arrest. But local police don’t have to comply.

Depending on where you live, local police might respond on a case-by-case basis, or the local government might establish a policy. The extent to which that policy refuses cooperation is what defines a sanctuary city.

A separate issue is whether the national government can cut off funds to uncooperative cities. (Again, this is a not a strictly liberal/conservative issue. The Affordable Care Act said that states that didn’t expand Medicaid in the way the law described would lose all federal Medicaid money. But the Supreme Court ruled against that kind of strong-arming.) In January, Trump issued an executive order threatening to pull federal funding from sanctuary cities, but, a judge blocked the enforcement of this order, writing:

Federal funding that bears no meaningful relationship to immigration enforcement cannot be threatened merely because a jurisdiction chooses an immigration-enforcement strategy of which the president disapproves.

In May, the Trump administration appeared to back off. Attorney General Sessions issued a definition of sanctuary cities that applied to very few places, and the restricted funds were only law enforcement grants from the Departments of Justice or Homeland Security.

[BTW: If you show a Trump supporter the Vox video, they’ll likely respond with this video from 1791L. However, that video does not actually identify any mistakes in the Vox video.]

Three Misunderstood Things

This week: the anti-gay baker, why the Senate can’t move on, and whether raising the minimum wage kills jobs.


I. The Masterpiece Cakeshop case (which the Supreme Court will hear in the fall).

What’s misunderstood about it: People think it has free-speech implications.

What more people should know: The baker objected to the whole idea of making a wedding cake for two men, and cut off the conversation before the design of the cake was ever discussed. That makes it a discrimination case, not a freedom-of-speech case.

*

Defenders of Masterpiece Cakeshop owner Jack Phillips frequently portray him as a martyr not just to so-called “traditional marriage”, but to the freedom of tradespeople not to say things they object to. For example, one conservative Christian tried to demonstrate a double standard like this:

Marjorie Silva, owner of Azucar Bakery in Denver, said she told the man, Bill Jack of the Denver suburb of Castle Rock, that she wouldn’t fill his order last March for two cakes in the shape of the Bible, to be decorated with phrases like “God hates gays” and an image of two men holding hands with an “X” on top.

Is this cake gay or straight?

But the Colorado Civil Rights Commission ruled against Jack, because the two cases are very different: Silva objected to the message Jack wanted on the cake, not to anything about Jack himself or the situation in which the cake would be served. If the government had demanded that Silva make that cake, it would have been an example of forced speech, which there is already a long legal history against.

Do conservatives also have a right to refuse forced speech? Yes. A Kentucky court recently ruled in favor of a print-shop that refused to make t-shirts for a gay-pride festival.

So liberals must have howled in rage, right? Not me, and not philosopher John Corvino, who defended the Kentucky decision on the liberal news site Slate:

the print shop owners are not merely being asked to provide something that they normally sell (T-shirts; cakes), but also to write a message that they reject. We should defend their right to refuse on free-speech grounds, even while we support anti-discrimination laws as applied to cases like Masterpiece Cakeshop. … Free speech includes the freedom to express wrong and even morally repugnant beliefs; it also includes the freedom for the rest of us not to assist with such expression.

The reason the baker has lost at every stage so far — the administrative court and state appeals court ruled against him, and the Colorado Supreme Court refused to hear his appeal, letting the lower court ruling stand — is that he wasn’t objecting to putting some particular message or symbol on the cake, like a marriage-equality slogan or a rainbow flag. For all he knew when he refused, the men might have wanted a cake identical to one he had already made for some opposite-sex couple. In short, he objected to them, not to the cake they wanted.

Corvino explains:

One might object that Masterpiece Cakeshop is similar: “Same-sex wedding cakes” are simply not something they sell. But wedding cakes are not differentiated that way; a “gay wedding cake” is not a thing. Same-sex wedding cakes are generally chosen from the same catalogs as “straight” wedding cakes, with the same options for designs, frosting, fillings and so forth. It might be different if Masterpiece had said “We won’t provide a cake with two brides or two grooms on top; we don’t sell those to anyone.” But what they said, in fact, was that they wouldn’t sell any cakes for same-sex weddings. That’s sexual orientation discrimination.

II. Mitch McConnell’s agenda.

What’s misunderstand about it: If the Senate is stuck on its ObamaCare replacement, why can’t it move on to the next items on the Republican agenda: tax reform and the budget?

What more people should know: McConnell is trying to exploit a loophole in Senate rules. As soon as a new budget resolution passes, his ability to pass both TrumpCare and tax reform goes away — unless he changes the proposals to get Democratic votes.

*

During the Obama years, we often heard that “it takes 60 votes to get anything done in the Senate”, as if filibusters that can only be broken with 60-vote cloture motions were in the Constitution somewhere, and the minority party had always filibustered everything. (That’s why even the weakest gun-control bills failed, despite 54-46 votes in their favor.) But the Senate recognized a long time ago that budgets have to get passed somehow, and so the Budget Control Act of 1974 established an arcane process called “reconciliation” that circumvents the filibuster in very limited circumstances.

That’s how the Senate’s 52 Republicans can hope to pass bills without talking to the Democrats at all. But there’s a problem: Reconciliation is a once-a-year silver bullet. Fox Business explains:

Reconciliation allows Congress to consider just three items per fiscal year, whether they pertain to one bill or multiple. Those items are spending, revenue and debt limit. Since the GOP also wants to pass its tax reform agenda using reconciliation, it cannot statutorily do that under this budget blueprint because the two policy measures overlap.

And NPR elaborates:

The budget resolution for the current fiscal year dictates that any reconciliation measure must reduce the deficit, which the GOP’s Obamacare repeal was designed to do. Republicans then could draft a new budget resolution for the upcoming fiscal year with easier deficit targets, allowing for more aggressive tax cuts.

Under the most commonly accepted interpretation of the reconciliation rules, as soon as Congress passes a budget resolution for Fiscal Year 2018 (which begins this October), the window for passing TrumpCare under the FY 2017 resolution closes. So the only way to get them both done before facing another election campaign is to do them in the right order: first TrumpCare, then a new budget resolution, then tax reform.

Otherwise, McConnell’s options become less appealing: He can get rid of the filibuster completely, which several Republican senators don’t support. He can scrap either TrumpCare or tax reform for the foreseeable future. Or he can start envisioning the kinds of proposals that might get eight Democratic votes, plus a few to make up for Republican defections.

III. The minimum wage.

What’s misunderstood about it: Both supporters and critics of an much-higher minimum wage think they know what effect it will have on jobs.

What more people should understand: The effect of a minimum-wage increase on jobs is an empirical issue, not something you can deduce from first principles. And the data we have only covers small increases.

*

There is a certain kind of conservative who thinks he learned everything he needs to know about this issue in Econ 101: Every commodity, including unskilled labor, has a demand curve; if you raise its price, demand for it falls.

The right response to that analysis is maybe. Imagine that you own a shop with one machine, run by your sole employee. The machine produces some high-profit item. To make things simple, let’s ignore counterfeiting laws and imagine that the machine prints money. Cheap paper and ink go in, $100 bills come out.

Obviously, you could afford to pay your employee a lot more than the $7.25-per-hour federal minimum wage. But you don’t, because the machine is simple to operate and you could easily replace him, so he doesn’t have any bargaining leverage.

Now what happens if the minimum wage goes up to $15? Do you fire your guy and shut the machine down? Do you abandon your plan to buy another machine and hire a second worker? No, of course not.

Admittedly, that’s an extreme example, but it points out the right issues: Whether an increase in the minimum wage causes you to employ fewer people depends on how much you’re making off those people’s work. If you have a razor-thin profit margin, maybe a higher wage makes the whole operation unprofitable and you lay workers off. But if you could actually afford the higher wage, and the only reason you don’t pay it already is that your workers lack bargaining leverage, then you don’t.

In fact, if a minimum-wage increase gives your customers more money to spend on whatever you make, then you might have to hire more people to meet the demand.

Which situation is more typical? One reason to think the second situation is, is that sometime in the 1970s wages stopped tracking productivity: Workers have been producing more, but not getting comparable pay raises, presumably because they lack the bargaining power to demand them.

During the same era, the minimum wage has not kept pace with inflation. An increase to around $11 would just get it back to where it was in 1968. If it wasn’t causing massive unemployment then, why would it now?

Supporters of a higher minimum wage also point to studies of past increases, which don’t show big job losses.

But there’s a problem on that side, too: Past hikes haven’t been nearly as big as the proposal to go from $7.25 to $15. I was a minimum-wage worker myself in the 1970s when it increased from $1.60 to $1.80. I suspect my employer was not greatly inconvenienced. But larger increases might have a shock value that makes an employer say, “We can’t afford all these workers.”

That’s why the new data coming in from Seattle is so important: Seattle was one of the first cities to adopt a much-higher minimum wage, so we’re just beginning to see the results of that. The headlines on that initial study were that the higher wage is costing jobs, but that early conclusion is still debatable.

So in spite of my own preference for a higher minimum wage, I find myself in agreement with minimum-wage skeptic economist Adam Ozimek: This is an empirical question, and both sides should maintain more humility until we see more definitive data.