Empirical Evidence, Sexual Harassment, and the Economic Way of Thinking

by on July 11, 2012 at 7:30 am in Economics, Law | Permalink

An improvement in a district’s schools will increase house prices in that district. A burger and fries will cost more than a burger. All else the same, cars that get more miles per gallon will sell for higher prices than cars that get fewer miles per gallon. I trust that most people would not find these statements objectionable.

All else equal, an improvement in workplace conditions will reduce wages. Now that looks slightly different than “A burger and fries costs more than a burger” but it’s really not. Think of the worker as buying better working conditions with a lower wage, thus better working conditions (burger plus fries) cost more (lower wage) than worse working conditions (burger alone).

Conceptually the burger story and the workplace story are very similar but the latter causes Henry Farrell to object:

…where you will find me objecting, and quite vociferously too, is to the suggestion that we ought to employ simple econ 101 style analysis in order to figure out the tradeoffs, and the appropriate solutions. This style of analysis has an awful lot of presuppositions built into it, and these presuppositions are politically loaded.

Let’s take the case of labour relations. If, as Matt argues, you ought to start with a model of firms, each of which has a cost function such that the total compensation they can offer is fixed, and any increase in costly rights (such as not having your body searched to stop you smuggling widgets out between your arse-cheeks) is inevitably associated with a proportionate decline in wages, then, indeed, you will end up concluding that compulsory rules forbidding body searches will lower overall choice without really benefitting anyone. But you only get to that result because of what you are assuming in the first place. You are assuming that there isn’t any real distributional action happening within the firm – and in particular, that the firm’s owners don’t have any surplus (that they are able to extract because e.g. of workers’ weaker bargaining position) that could be reallocated through regulation. In other words, you are only ‘refuting’ the people you disagree with, because you assume away the problem that they are worried about.

(FYI, Henry is responding to a smart and devastating post by Matt Yglesias that no doubt rankled because Matt is a liberal.)

Henry gets the economics wrong (the model does not require perfect competition) but let’s ignore that and look to his larger point:

The lesson here is straightforward. Simple economic models can be quite useful, but they should be employed with very considerable caution. In particular, one should always think carefully about whether the assumptions of your model blind you to factors that are important to the debate that you are applying them to. As a secondary matter, you should also look to the empirical support for your model – intuitively appealing models are frequently wrong.

Now. what are you expecting to follow this paragraph? Surely, some empirical evidence that the model is wrong. None follows. Henry seems to think that economists have never thought about their assumptions or tested their models. Nothing could be further from the truth so let me provide some empirical evidence about the model of compensating differentials:

Here is Jonathan Gruber and Alan Krueger (n.b. both of them have worked for Obama) on workers’ compensation:

This paper empirically examines the incidence of the workers’ compensation program to infer the likely consequences of mandated health insurance proposals. In certain industries, such as trucking and carpentry, workers’ compensation insurance costs are quite large, and vary tremendously within states over time, and across states at a moment in time. This variation is used to identify the incidence of the program. Empirical analysis of two data sets suggest that changes in employers’ costs of workers’ compensation insurance are largely shifted to employees in the form of lower wages.

Here is Gruber on mandated maternity leave:

I consider the labor-market effects of mandates which raise the costs of employing a demographically identifiable group. The efficiency of these policies will be largely dependent on the extent to which their costs are shifted to group-specific wages. I study several state and federal mandates which stipulated that childbirth be covered comprehensively in health insurance plans, raising the relative cost of insuring women of childbearing age. I find substantial shifting of the costs of these mandates to the wages of the targeted group. Correspondingly, I find little effect on total labor input for that group.

And, dare I mention it, here is Joni Hersch in a recent paper in the AER on sexual harrasment:

Workplace sexual harassment is illegal, but many workers report that they have been sexually harassed. Exposure to the risk of sexual harassment may decrease productivity, which would reduce wages. Alternatively, workers may receive a compensating differential for exposure to sexual harassment, which would increase wages. Data on claims of sexual harassment filed with the Equal Employment Opportunity Commission are used to calculate the first measures of sexual harassment risks by industry, age group, and sex. Female workers face far higher sexual harassment risks. On balance, workers receive a compensating wage differential for exposure to the risk of sexual harassment.

Henry goes wrong because he doesn’t want to conclude that sexual harassment is ok but he thinks that the only way to deny that conclusion is to deny that wages are higher so he rejects the model, he rejects the assumptions that he thinks (incorrectly!) are driving the model and he assumes without looking for any evidence that wages are in fact not higher. (Talk about being blinded by assumptions!).

Matt Yglesias, by the way, is much smarter on this point, he doesn’t reject the model but gives several extensions, such as externalities, that may change the conclusion.

Here is another way to think about sexual harassment and the model. What the theory and the empirical results are saying is that people exposed to a higher risk of sexual harassment are paid more, just as people exposed to a higher risk of death are paid more. In the case of risk, however, the firm’s owners (shareholders) are paying higher wages but also getting the benefits of risky work. But in the case of sexual harassment the shareholders are paying higher wages but not getting the benefits of sexual harassment. In other words, from the firm’s point of view sexual harassment is a bit like employee theft – with the stealing being done by the harassers. (I alluded to this point in my original post and Miles Kimball made it as well.) Thus, shareholders may benefit if the government can reduce sexual harassment at low cost, precisely because they would then be able to pay lower wages without losing productive workers.

Now this analysis is not complete or the only way to frame the issue (again see Yglesias for other approaches). It may bother some people that in this version the biggest gainers from sexual harassment law may well be the firm owners rather than the previously harassed (but does it bother you that the among the biggest gainers from better schools may be property owners without children?) Nevertheless, the principle here is clear, the way to think about these issues is not to throw out economic reasoning but to apply the reasoning ever more deeply.

1 Master of None July 11, 2012 at 8:12 am

A possible extension:

All else equal, an improvement in status will reduce wages.

Agree or disagree? I can think of several examples that support, but there are many that don’t.

2 Neal July 11, 2012 at 10:12 am

All else cannot be equal because status and wages are entangled.

3 Neal July 11, 2012 at 10:12 am

That is to say, “agree, but this model will never make a useful prediction.”

4 Kyle July 11, 2012 at 9:53 pm

Prediction: if finance became lower status because of Occupy Wall Street fewer kids from top schools would enter into it and pay would go up for the ones that did.

5 Kinanik July 11, 2012 at 10:21 am

All else being equal, that’s true. There seem to be some conflating effects though, as jobs with high status are usually so because the risk is higher – so firemen may receive a pay cut for their high status, but the reason for their high status is the increased danger of their job, which comes with a pay increase.

Of course, where status is determined by wage, it would become very hard to ferret out causal effects, but if status came only from wage, you would see a flatter distribution of income than when it’s not.

There also seem to be a fixed number of positions for a lot of high status jobs. Successful actors are high status and wealthy, but there is a huge group of people seeking to become successful actors. If you average the wage of everyone who tries to become an actor, what then? Same with sports – there’s a lot of unpaid effort that goes into trying to become a professional athlete.

6 Barry July 11, 2012 at 11:40 am

Something I had read once was that convenience store clerks were hour-for-hour far more likely to be shot on the job than were the police (working by neighborhood, of course).

7 Dick King July 11, 2012 at 12:46 pm

The probability of a President of the United States being shot is about 2% per year. The pay isn’t that good considering the responsibilities and the low job security, either.

Somehow we don’t run short of applicants.


8 msgkings July 11, 2012 at 1:05 pm

Don’t you have to factor the increased lifetime earnings after you leave the job? I’d say it pays pretty good. And if you included non-monetary rewards it’s about the highest paying job there is.

9 The Original D July 11, 2012 at 1:59 pm

If you count the total $$ spent to elect him to office, the president’s paymasters probably disagree.

10 JoeDog July 11, 2012 at 8:13 am

He lost me at “arse-cheeks.” I find few things more grating than Americans who use the British spelling.

11 Joey July 11, 2012 at 8:45 am

He’s Irish.

12 Cranky Brit July 11, 2012 at 11:44 am

That’s even worse.

13 Andrew' July 11, 2012 at 12:23 pm

“He lost me at “arse-cheeks.””

Try using both hands.

14 JWatts July 11, 2012 at 12:34 pm


15 msgkings July 11, 2012 at 1:03 pm

Outstanding, LOLed.

16 mrpinto July 12, 2012 at 5:49 pm

He’s Irish, and you’re a no-good hater. Few things are more grating than a no-good hater.

Seriously. If you can’t find something more annoying than regional variations in word-spellings, you need to expand your horizons!

17 asdf July 11, 2012 at 8:19 am

“In other words, from the firm’s point of view sexual harassment is a bit like employee theft – with the stealing being done by the harassers.”

Couldn’t you say that the opportunity to harass people is an improvement in workplace conditions for the harasser, and thus the company can pay the harasser lower wages because he is able to harass as compensation.

We can play this game all day.

18 Ritwik July 11, 2012 at 8:31 am


The initial distribution of property rights, as always, matters. One can, in fact, reformulate every single liberal/progressive argument into some version of *I disagree with the current distribution of property/ property rights*

Alex/ libertarians could take the Coasian way out here, but they will have to assume away not just transaction costs, but also any pre-conceived social notions of justice. Tyler would not approve.

19 John Thacker July 11, 2012 at 10:57 am

One can, in fact, reformulate every single liberal/progressive argument into some version of *I disagree with the current distribution of property/ property rights*

Except that most liberals and progressives then rule out any bargain or selling one’s property rights, and rule out the possibility of diversity. I certainly agree that transaction costs exist, but that’s only a reason to be concerned about the initial distribution of property rights, as you note.

It’s not a reason to ban people from choosing an alternative working arrangement. Compare the policy of changing the default to no-smoking, to a policy of outlawing even specific bars set up for smokers. Compare changing the default to automatic investment in a 401(k) plan to not allowing people to choose something different.

If liberals and progressives in fact reformulated their arguments and limited them to disagreeing with the current default allocation of rights, I think that Alex, Tyler, and other libertarians would be much happier.

20 Brian Donohue July 11, 2012 at 11:08 am


21 Yancey Ward July 11, 2012 at 11:45 am

Great comment, John.

22 k July 11, 2012 at 12:13 pm

Umm….if transaction costs exist, selling property rights itself becomes costly.

23 JWatts July 11, 2012 at 1:46 pm

“Umm….if transaction costs exist, selling property rights itself becomes costly.”

Just because they exist doesn’t mean they are costly. Or even that people will agree on what costly means.

24 k July 11, 2012 at 3:57 pm

either costs are costly or they are not worth talking about :)

25 John Thacker July 11, 2012 at 5:58 pm

Of course. That’s why the initial allocation matters.

But again, that doesn’t mean selling them should be banned. Some people will put different value on the rights; some people will have higher or lower transaction costs.

You’re already putting the minority of people with different values and views worse off by legislating the majority view. Why ban even the possibility of coming to a different arrangement?

If it’s really so difficult, then no different arrangements will be made, and what’s the harm?

26 Tom West July 11, 2012 at 12:27 pm

There are no inalienable rights?

It might be interesting to see what selling your lifetime right to vote would go for.

More seriously, I think liberals (at least this liberal) considers that there are considerable externalities to trading away certain rights, and because the rules of the game are *never* fixed in stone, such externalities are important. As an example, I would consider it likely that if people were allowed to trade away the right not to work in a smoky environment and many people took it, it’s quite likely that the rules would change to remove that right altogether. When we’re dealing with human beings rather than homo economus, perceptions matter – a right that can be traded away is no longer a right, it becomes a government granted privilege, and government granted privileges can or should be stripped away.

27 msgkings July 11, 2012 at 1:09 pm


Not every right of a human being can be defined as transactable property.

Whether you agree or not can’t really be reasoned, it’s a matter of first principles. Either you think you can trade away all of your rights, or just some of them. Either abortion is a medical procedure, or murder. Either there is a Creator, or there isn’t.

28 John Thacker July 11, 2012 at 6:13 pm

I agree that some rights should be inalienable. I think that the vast majority, including of libertarians, think that. But the list of inalienable rights should be limited.

Indeed, the standard libertarian argument about the distinction of negative versus positive rights is in large part an argument that positive rights result in contradictory rights. I’d argue that modern liberals tend to have fewer inalienable rights than conservatives, because when you multiple the list of rights you can’t avoid having them contradict each other.

An inalienable right to not be offended interferes with an inalienable right to free speech. Regulations that try to consider each and every scenario and possibility will not only reach conclusions that don’t please many people (and wouldn’t match the agreement that they’d negotiate between themselves), but will themselves have transaction costs, legislative costs, and maintenance costs, often exceeding that of making a Coasian bargain.

I, and other libertarians, thus use principles. such as the distinction between negative and positive rights, to decide what is and what is not an inalienable right. Obviously people will differ, and one can be informed by advances in science. Libertarians are generally okay with mandatory vaccinations, for example, for airborne diseases, because the alternative seems to require even worse impositions on liberty.

As an example, I would consider it likely that if people were allowed to trade away the right not to work in a smoky environment and many people took it, it’s quite likely that the rules would change to remove that right altogether.

Largely agree– but if so many people were willing to sell that right (and people who wanted to smoke were willing to pay a lot), wouldn’t that also mean that a democratic society would be willing to change the rules anyway? After all, the rules that a democratic society comes up with are roughly majority rule, altered by how strongly people feel about it.

Even if the default allocation were changed, then there would still, under my system, be a diversity of workplaces. It strikes me that you and your fellow liberals believe either that their should be only one outcome, or that a uniform outcome is the inevitable result, so we must regulate in favor of what uniform outcome we want. However, I don’t think that is necessary or desired.

Transaction costs surely exist in the market. It is not perfect. But neither is the outcome of the political process, which has its own inefficiencies and transaction costs. If there must be a default, let there be the default that satisfies the majority. But let the minority negotiate safe spaces where it can indulge its different preferences. Don’t insist on uniformity.

29 Tom West July 12, 2012 at 1:02 am

A cogent reply, as usual, Mr. Thacker.

Largely agree– but if so many people were willing to sell that right (and people who wanted to smoke were willing to pay a lot), wouldn’t that also mean that a democratic society would be willing to change the rules anyway?

So now *everyone* ends up exposed to smoke *and* no-one gets a premium for it.

But let the minority negotiate safe spaces where it can indulge its different preferences. Don’t insist on uniformity.

The trouble is that a lot of workplace rights are a sort of prisoner’s dilemma. As long as everybody demands a non-smoking environment, everybody gets it. But as soon as one (okay a number) are willing to divest themselves of the right, for whatever reason, the inevitable pressure is that everyone loses it.

Government legislation (when it works – there are certainly failures there as well) essentially forces everyone into co-operate mode, even for those who would prefer to defect. It’s a loss of freedom in exchange for an equilibrium that makes most better off.

And yes, I will admit to being an elitist. Human beings are well-known to be terrible at estimating risk, either over or under-estimating it. Now, a Libertarian can look at someone with emphysema caused by poor decisions that a young person made to work in smoky conditions to earn a few bucks more a month and say “tough luck, you were a responsible adult, you now live with the consequences.” I, however, am not. Since I know that I’ll be pressing for society to cover costly medical care for the rest of his life, I have little trouble denying him the freedom to make that bad choice.

It’s slippery slope and all that, and I don’t claim to be making such a decision with divine right behind me, but if I’m voting that we’ll be paying his medical bill, then I’m voting that he not have the opportunity to make particularly egregious trade-offs of health for money. May the side with the most voters behind them win…

30 John Thacker July 12, 2012 at 11:27 am

Our fundamental disagreement is that I simply don’t think that it’s the case that in the absence of legislated uniformity, every workplace would still be the same. You apparently do believe that.

The situation of smoking is not actually a prisoner’s dilemma, unless you’re talking about a situation where everyone wants to smoke but prefers that their co-workers don’t, but they’re all better off if no one smokes. Smoking is a case of people having quite different preferences; the smokers are worse off if nobody can smoke. It’s thus a case of bargaining rather than a prisoner’s dilemma.

A better example of a prisoner’s dilemma is the case of vaccination– and note that there, the vast majority of libertarians do favor government action. Exception that proves the rule.

I understand your argument that without legal uniformity everywhere will allow smoking. I don’t think that’s true, but if it were, I think that would demonstrate that smoking would be more valuable to the smokers than having them not smoke to the non-smokers. I can see all sorts of arguments about how non-smoking friends wouldn’t be willing to insist upon going to a non-smoking restaurant or bar, etc.

However, I think that reality shows that non-smoking restaurants, bars, and workplaces were growing in number before the passing of any ban. The ban came after the decline in smoking and tolerance for the smoking of others; it was a classic case of a majority consolidating its gains by eliminating the preferences of the minority.

31 Pithlord July 12, 2012 at 3:33 pm

“If liberals and progressives in fact reformulated their arguments and limited them to disagreeing with the current default allocation of rights, I think that Alex, Tyler, and other libertarians would be much happier.”

This has already been empirically disconfirmed, at least with respect to “other libertarians.” Sunstein and Thaler did just that — and libertarians in the blogosphere became unhappier than ever.

32 Anonymous July 11, 2012 at 8:19 am

JoeDog, do you find it grating when Irish people do it? Because he’s Irish.

33 asdf July 11, 2012 at 8:36 am

Yglesias post is good despite his trying to make it bad. It’s funny watching people sort of get it right but then alter or omit certain things to protect their ideology, this place is a great example too.

Anyway, the problem with Yglesias post is it’s true, but that doesn’t necessarily support his cause. As a commentator points out no regulation is the result of, “smart technocrats with all the necessary data and analysis making clean calcuations and then implementing the right regulations.” Instead it’s, “political coalitions with their own beefs to pick duking it out and some legislative stew coming out at the end.” Thus, there is some value in assuming a base level of prefered “freedom” since you can’t trust the legislative process to “get it right” even if getting it right is theoretically possible.

34 The Original D July 11, 2012 at 2:30 pm

Eh, I think it’s a middle ground. Legislation rarely spells out detailed regulation. Rather, it creates or designates and agency to do that. You can argue about the competence or partisan leanings of the bureaucrats though

35 dead serious July 11, 2012 at 8:47 am

“All else equal, an improvement in workplace conditions will reduce wages.”

I’m not sure why you need the “all else equal” qualifier in that axiom, but it’s false. It can be true *sometimes*, but that statement in and of itself is false.

Casual Fridays in summer months, for example, don’t cost the employer in terms of expenditures nor do they impact revenues; thus they don’t impact employee wages.

36 Anon July 11, 2012 at 9:08 am

No, you’re wrong. And a bit rude to boot.

If an employee values casual Fridays (this is considered a workplace IMPROVEMENT? Good grief…), then he will accept lower monetary compensation in exchange for a more “enjoyable” workplace.

Imagine you like racquetball. Would you accept a lower wage to work somewhere which had a racquetball court available during breaks?

Repeat this glorious experiment, but switch casual Fridays in for racquetball.

37 dead serious July 11, 2012 at 10:25 am

A racquetball court costs something to build/buy/maintain/staff.

Casual Fridays cost nothing. Whether *you* consider them an improvement is not germane: many people do. I don’t change my attire, so it doesn’t have any bearing on me.

Some “improvements” cost money or resources and so I would agree with the original statement. However some do not.

38 John Thacker July 11, 2012 at 11:01 am

You’re dead wrong, dead serious, and the literature and empirical evidence bears that out.

You’re making a common but wrong assumption, that wages are determined by costs and resources. They are not. Compensation is determined by productivity, and non wage things that increase happiness decrease wages.

Try turning it around. Would you insist on being paid more to have to work at a place with an unpleasant work environment? Then it doesn’t matter to you whether or not it would be free for the employer to improve the environment.

I know people who have worked for terrible bosses who are just jerks. The boss isn’t gaining anything through being a jerk. Most people try to leave there as soon as possible; some people stay because they get paid more than they would working for a good boss, and they know that.

It has nothing to do with costs to the employer. Neither theory nor empirical evidence would predict that.

39 John Thacker July 11, 2012 at 11:02 am

I guess I should say that the boss may psychologically be gaining something by being a jerk, but it doesn’t save the company money or impact expenses.

40 Tom West July 11, 2012 at 2:11 pm

Compensation is *limited* by productivity. It’s actually *determined* by supply/demand, which for most jobs is usually considerably lower.

It’s why minimum wages may cost a few jobs on the margin, but usually result in more money going to the lowest classes of workers in aggregate and why its labor shortages, not rising productivity that most contributes to wage increases (in the short term).

41 Tom West July 11, 2012 at 2:13 pm

*sigh* its labor shortages -> it’s labor shortages

42 The Original D July 11, 2012 at 2:32 pm

If the boss’ jerkiness results in higher turnover, it definitely raises the employer’s costs.

43 Pithlord July 12, 2012 at 3:39 pm

Wouldn’t basic price theory predict that if there is a way employers can improve employee happiness without cost to them, a competitive labour market will require all employers to do it and wages will stay the same?

If the market’s sticky, then this might take longer, but I’m not sure there is any robust reason to think wages would drop. I would have though nominal money wages would be a whole hell of a lot stickier than any non-wage psychological compensation.

44 Erick July 11, 2012 at 10:46 am

Some simple evidence of this is the fact that “jean passes” exist where employees can donate to some sort of fund, and in exchange are allowed a certain numbe of days where they can dress below the dress code.

45 James Vonder Haar July 11, 2012 at 3:55 pm

I’m not a trained economist, but isn’t it more accurrate to state that a policy that employees prefer that costs the employers nothing will create a surplus, which will be distributed among both agents according to their bargaining power?

46 Isaac July 11, 2012 at 9:16 am

Apparently you didn’t actually read the blog post, because it presents exactly the reasons why Casual Fridays would reduce employee wages (or, from the other perspective, the existence of Casual Fridays allows the employer to offer less compensation elsewhere, for the same result). It may not be a very big effect, though I’m sure someone has measured it, but it’s consistent with the other examples presented.

47 Pithlord July 12, 2012 at 3:40 pm

On the assumption casual fridays don’t actually cost the employer anything, I don’t think this would follow at all. See my comment upthread.

48 zbicyclist July 11, 2012 at 10:51 am

I think you need to separate out the cost to the employer from the benefit to the employee. They are correlated, but are NOT the same thing.

In an ideal world, the employer finds a way to provide a benefit for $1 that the employee values at $2.

Flextime is a good example: relatively low cost to the employer, but valued by many employees who will therefore accept lower wages / not take a higher paying job because the employer provides scheduling that aligns better with their childcare needs (etc.)

49 derek July 11, 2012 at 11:29 am

It also works the other way. In my industry we are significantly driven by events outside of our control. Our customers expect to pay us well so that when something breaks, we show up at all hours fully resourced to get things running. The penalty for them not doing it is loss of product in the tens of thousands.

For someone to do what I do, you have to give up the hottest months of summer (not too long around here), be willing to work weekends and evenings. The pay is good. It has to be or you don’t have anyone working for you.

50 Zachary July 11, 2012 at 11:54 am

Too Oft ignored point.

51 lemmy caution July 11, 2012 at 2:20 pm

When times are good and employees have a lot of other job opportunities, employers come up with things like “casual fridays” and other perks. When times are bad and employees don’t have a lot of other job opportunities, employers load up the employees with extra work and yank perks.

If wages and perks/burdens were really easily interchangeable, that wouldn’t really be the case. The wages and perks/burdens mix would be adjusted to to maximize the utility of the employer and worker. However, wages are sticky and employers can switch how they treat employees at any time.

52 ohwilleke July 11, 2012 at 7:48 pm

The “all else equal” condition is necessary because, of course, everyday experience makes absolutely clear that the workplaces with the worst workplace conditions also happen to be the ones that pay the lowest wages.

Wages are much more strongly driven by the size of the pool of people who are qualified to do the job than they are by the desirability of the job to people who are qualified to do it relative to other jobs that the same people are qualified to do. Jobs anyone can do without an education or skill don’t pay well (more educated or skilled people can do those jobs too, but improve their wage by chasing jobs that fewer people can do and that hence pay more).

In real life, when differences in worker quality are visible to the employer when hiring (or sometimes firing or non-renewing a job), working conditions can be used as a lure to get better workers. But, typically an employer who cares about worker quality from a business perspective will use both higher wages and better working conditions to lure better workers, while employers who have a business model that is less picky about worker quality will try to cut both worker pay and working conditions and will hire people at the bottom of the barrel that nobody else will hire (the bottom feeders fare especially well in markets with non-work related discrimination because then they can get quality workers that no one else will hire despite low pay and poor working conditions).

The high pay for poor working conditions employer is largely a myth that economic theory says should exist, but doesn’t, because its models are too crude.

Incidentally, the more realistic example of casual wear in the summer is of employers allowing casual wear at work in order to make less intense air conditioning bearable (something that works in part because not that hot in the summer Britain set standards for formal wear in much of the Anglo-American world), so the employer benefits from decreased utility bills. Utilities facing the need to build more powerplants to meet demand, government employers, big businesses, and progressive smaller employers have been implementating this strategy on a wide scale for almost a decade now.

53 Nathan July 11, 2012 at 9:07 am

Makes sense. I’d be curious to see the argument that the biggest beneficiaries of better schools are people without children in more detail, though. I mean, sure, your property values goes up, but so do the taxes you have to pay to keep possession of the place every year. And people with children are taking advantage of of schooling that would otherwise be very expensive to purchase. I would think the big winners are people with more than your average number of children, and that childless homes simply break even.

54 david July 11, 2012 at 9:14 am

The argument is quite simple: you cash out and sell to people with children. Since the supply of houses near the improved school is generally steeply inelastic, you can capture the entire improvement in monetized form. The group you describe – people with more than the average number of children – are the ones forking out the cash to you. People already nearby with more than the average number of children also benefit, but they are statistically unlikely to be the individuals in the entire housing market who gain the most subjective benefit from the school improvement.

55 Nathan July 11, 2012 at 9:35 am

Got it, thanks.

56 byomtov July 11, 2012 at 11:19 am

Yes but.

It depends on housing preferences. If housing is a purely economic decision and you have no non-economic reason for not wanting to sell then this might be true. (Even in that case, note that transaction costs – real estate commissions, moving costs, general hassle – are quite high.)

But there are many reasons people might not want to sell, this alleged windfall notwithstanding. Maybe they like the area, find the location convenient and pleasant, have friends and family nearby, have the house fixed up to their tastes (and maybe not all those costs are recoverable), have lived there for many years, etc.

And is the price effect really unambiguous? Families with children will pay more, but others will pay less. Besides, what if it’s a small house?

In other words, your logic leaves out way too much.

57 david July 11, 2012 at 5:26 pm

There is no reason to believe that people without children are disproportionately more likely to like the area, find the location convenient and pleasant, have friends or family nearby, have the house fixed up to their tastes, or have lived there for many years. Indeed the converse is more plausible. So it is still the case that people without children receive the disproportionate benefit of school improvement, under the factors you have identified.

58 byomtov July 11, 2012 at 6:37 pm

Sorry. I don’t understand.

If an increase in the value of my house is not important to me – or does not compensate me for the disutility of selling and moving – and my property taxes have increased, how have I benefitted?

Your claim was that the childless couple could get the benefit of the improved schools by selling. But if they don’t want to sell for subjective or idiosyncratic reasons then where is the benefit?

59 byomtov July 11, 2012 at 6:43 pm

Almost forgot. Your response also does not address the considerable transaction costs involved in your hypothetical sale.

60 david July 11, 2012 at 8:10 pm

The claim is not that you are likely to benefit; the claim is that you are most likely to be larger beneficiaries if you are childless. Everyone else who has children also faces the costs of moving and property taxes.

61 byomtov July 11, 2012 at 9:30 pm

There is a difference between Alex’s claim and yours. He says,

among the biggest gainers from better schools may be property owners without children

That’s OK. It’s tentative, and suggests only that some people without children might profit nicely by in effect selling their access to improved schools.

But Nathan says,

I’d be curious to see the argument that the biggest beneficiaries of better schools are people without children in more detail, though.

That claim, which you seem to be defending, is a lot stronger and broader, and does not hold. There will be childless households that will lose rather than gain from the new arrangement. In fact, I would guess that childless households are more likely to lose than gain, because the sale option will often be unattractive.

62 Thomas Sewell July 11, 2012 at 1:22 pm

The fun corollary is that families with children who home school or send them to private school are able to move into a lower quality school district area and get more for their money in terms of housing than they would in a better quality school district, all else being equal.

Thus some of us end up looking for bigger houses in poor quality school districts in order to get a good deal. :)

63 Mike July 11, 2012 at 2:55 pm

Though where I am, most of the available private schooling has become so expensive, it appears to be cheaper to live in the neighborhood with the good public school. Perhaps people who home-school can exploit the advantage, though I’d assume for many they could make more money by sending the kids to public school and having the schooling parent work. non-economic reasons for wanting to home school or go to private school aside, of course.

In my experience, the people who benefit from this are the ones who have no kids and don’t intend to have any. They can live wherever the hell they want, and benefit from the reduced demand on neighborhoods with lousy schools. I say this with bitterness, having just paid through the nose for a house in a good school neighborhood!

64 Benny Lava July 11, 2012 at 9:10 am

All else the same, cars that get more miles per gallon will sell for higher prices than cars that get fewer miles per gallon.

Wait, no that is not true at all. I paid more for the V6 model which got less miles per gallon than the 4 cylinder DOC. Have you ever bought a car before in your life? When you start with ignorant assumptions you end with ignorant conclusions.

All else equal, an improvement in workplace conditions will reduce wages

Depends on the labor market now doesn’t it?

65 Isaac July 11, 2012 at 9:18 am

One would venture to suggest that “all else equal” means “equal labor market” also.

66 FP July 11, 2012 at 9:46 am

V6 engine has more power. What part of “All else the same” is not clear to you?

67 Benny Lava July 11, 2012 at 6:58 pm

Does it have more power? How would any American driver ever know? As they say performance is for rednecks. I got the bigger engine merely for status, not performance.

What part of all else equal isn’t clear to you? You seem to be making a circular argument here.

68 ohwilleke July 11, 2012 at 8:02 pm

“All else the same, cars that get more miles per gallon will sell for higher prices”

This is a proposition that is deceptive because the laws of physics and the basic principles of automotive engineering get in the way, in a manner not independent of income inequality between consumers.

In a car driven by an internal combustion engine power and miles per gallon are inextricably intertwined, as is the sturdiness of the car’s construction (which drives weight) and miles per gallon. Since power, sturdiness of construction and miles per gallon all matter to consumers and are inextricably intertwined with each other, it is never the case, even statistically, that you can imagine an “all else is the same” situation. The best that you can do is observe that people who have a greater ability to pay not only for the car but for the gasoline that fuels it, tend to favor more power and more sturdy construction, while those who are less of an ability to pay for both the car and the gasoline that fuels it, are more concerned about fuel efficiency.

Electric cars and the availability of expensive low weight/high strength materials have skewed this analysis somewhat. But, I would argue strenuously that people who are buying cars like the Tesla at $100,000 each are not motivated to do so to any material extent by their ability to reduce their gasoline bill. Indeed, it is almost always the case that unsubsidized hybrids sell at a price premium that can’t be justified by the fuel savings.

For example, the hybrid Toyota Highlander is identical in almost every respect except fuel efficiency with its conventional engine Toyota Highlander twin. But, the hybrid sells at a price premium relative to the conventional vehicle that the fuel savings ove the life of the vehicle cannot possible justify in dollars and cents alone, even if one employs some option pricing analysis because buying a more fuel efficient vehicle purchase is one way to hedge against future fuel price increases. Anyway you cut it, a significant part of the price premium involved in buying a super fuel efficient vehicle comes from the cultural and symbolic message it send, not from the accountant’s analysis of the pros and cons.

69 KLO July 11, 2012 at 9:55 am

Was the V6 engine otherwise equal in every way to the 4 cylinder engine? Did it have the same power and torque? If so, were the power and torque curves identical? If so, did the two engines weigh the same amount? If so, were cars with the V6 sold with the same transmission as cars with the 4 cylinder? If so, did the engines sound identical in operation?

The problem with this argument is the “all else being equal” bit. In the real world, all else is not equal, as your example of cars with different engine options shows.

70 John Thacker July 11, 2012 at 11:04 am

In the actual car market, engines have become much more efficient, to the delight of customers. However, that efficiency has mostly been “spent” in the name of higher power and torque, rather than used to maintain the old power numbers with higher gas mileage.

71 JWatts July 11, 2012 at 2:17 pm

“However, that efficiency has mostly been “spent” in the name of higher power and torque, rather than used to maintain the old power numbers with higher gas mileage.”

The reports I’ve read indicate otherwise. What I’ve read indicates that IC gasoline engines have improved on average several percent per year over the last 30 years. This gain has been distributed out between, more power, lower emissions and greater miles per gallon over that time frame in the US. All new cars with significant sales in the US produce substantially less emissions than similar cars from 30 years ago. (Lowering emissions does generally involve increasing power. Pushing exhaust through catalytic converters does decrease net horsepower.) In addition, the overall fleet mpg average has also increased substantially during the same time period. The remainder of improvements have gone into performance improvements.

For example:
Passenger Car MPG US average:
1980 24
1990 28
2000 28.5
2010 33.9

Passenger Light Truck/SUV MPG US average:
1980 18.5
1990 20.8
2000 21.3
2010 25.2

72 John Thacker July 11, 2012 at 6:17 pm

Yes, you’re right that I should have mentioned emissions reductions. In my haste to correct the assumption that engines have not gotten more efficient I made an incorrect statement. A more correct statement would have been to say that “engines have gotten quite a bit more efficient. However, the efficiency has not all gone to increasing gas mileage at the same power or emissions; rather, much of it has gone to increasing power and emissions.”

Thank you for the correction.

73 Benny Lava July 11, 2012 at 7:03 pm

We can either approximate all else equal as two similar car models where one is more fuel efficient, or with two identical models but one having a trim package with a more efficient engine. Either way the argument put forth is verifiably false. I find it telling that the libertarians on this blog insist on a childlike faith in authority and eschew empirical evidence. Alex says it so it must be true.

74 John Thacker July 11, 2012 at 7:45 pm

I don’t think you understand what the word “verifiably” means.

You may not care about performance, but many people do. Higher performance gives higher status, which causes people like you to follow the lead of the “rednecks” and select the bigger engine.

If you look from model to model, or model year to model year, car magazines and consumers will always hail and prefer an equivalent (or slightly more power, more cylinder) engine with higher gas mileage for the higher gas mileage. Your claim is empirically false, and you deny all evidence in favor of your childlike delusions.

75 Careless July 11, 2012 at 12:09 pm


76 RPLong July 11, 2012 at 9:17 am

The guy lost me when he started using the word “presupposition.”

I’m not saying it’s not a word. Likewise, I would never suggest that “betwixt” is not a word. It’s just…

77 Neal July 11, 2012 at 10:14 am

Ever heard of “presuppositionalism”?

78 wasd July 11, 2012 at 9:21 am

Come on, Cowen. Dare I mention it, Joni Hersch’s paper is in P&P, not the real AER. Have you read it? It’s weak.

Alternative explanation: people who earn higher wages are more likely to report sexual harassment. It’s not like the author does anything to help convince anyone her results are causal, or refute any of countless other explanations for the statistical regularity she’s found.

79 Bartman July 11, 2012 at 9:57 am

Why are you referring to Alex as “Cowen”?

80 David C July 11, 2012 at 9:36 am

I get it that wages are higher in riskier professions. What I can’t agree, is that they are ENOUGH higher to compensate the employees for the risks they take. If we were in a long-term full-employment job market, where employees had some measure of market power, maybe. But not even then for low-skill occupations.

My grandfather worked in coal mines in Ohio between the world wars. He earned the proverbial dollar a day, and was in fact coerced to shop in the overpriced company store. He was lucky enough to escape death and major injury, but witnessed many coworkers who weren’t so lucky. I can tell you that he had no market power, no realistic way to quit and find another job. Its in situations like these that regulations can help workers.

81 John Thacker July 11, 2012 at 11:07 am

Monopsony situations do exist, especially in company towns (and thankfully are rarer than they used to be), and they cause what you describe.

However, I was just talking to my grandfather about his work history. He worked at mills, at truck driving, at fixing cars, and a host of other jobs, some for the dollar a day (or dollar a week during the Depression.) His history is different; he actually quite a few times quit a job and found another. At least one time he quit because of a new foreman imposing bad work rules and being a jerk; the result of this was to get the mill owner to fire the foreman and come to my grandfather’s house to ask him to come back to work.

82 dearieme July 11, 2012 at 11:18 am

“was in fact coerced to shop in the overpriced company store”: why didn’t the company save itself a lot of bother, close the store, and pay the workforce less?

83 Skye Winspur July 11, 2012 at 2:22 pm

I’m guessing the company found a lot of benefit in being able to send a spokesman before Congress and plausibly assure the nation that Our Firm is guaranteeing a high quality of life for this Ohio town (chances are the company was getting some government contracts).

84 Wonks Anonymous July 11, 2012 at 11:32 am

I believe in compensating differentials generally speaking, but I’d like to hear what economists think of Peter Dorman’s empirical work arguing against the standard econ view on workplace risk. Maybe employees are really terrible assessors of risk.

85 msgkings July 11, 2012 at 1:03 pm

Considering most individuals are terrible assessors of many kinds of risk, this seems very plausible.

86 Seth D July 11, 2012 at 9:53 am

David C is on the right track here. The fact that we observe lower wages as a result of greater regulation is a function of the “distributional” decisions made within a firm, as Henry argues, which is in turn a function of the power dynamics within the firm and the labor market. Considering the risks of coal mining, why weren’t miners like David C’s grandfather paid more? Surely economic theory would predict that no one would take so little to do such dangerous work. The answer is that the workers had little bargaining power and the owners of coal mining operations decided to pocket the surplus to enrich themselves instead of paying the workers more. So when confronted with new workplace safety regulations, the owners can choose to lower wages further in order to maintain their current level of profit or they can keep their wages constant and earn less money for themselves. This is the distributional decision. Just because we observe that firms tend to choose to cut wages doesn’t mean they were forced to do so. They could have taken less money for themselves and had happier and healthier (and better paid and likely more productive) workers as a result. The assumptions of the economics 101 model don’t allow for such decisions. This is the heart of the Henry’s critique.

87 improbable July 11, 2012 at 10:21 am

But wouldn’t kindler gentler more generous bosses give more of their surplus to the workers anyway? Even those not in dangerous jobs? When people invoke this it seems like a deliberate way to muddle two issues in a way that is counter-productive to understanding what’s going on.

Politically we often try to create packages like that, taxing something and at the same time throwing a bone to someone, to make the whole package appealing. That doesn’t mean you shouldn’t try to understand the effects of each half.

As for coal miners, they are paid quite a substantial premium for the risk. (First page of google: “You can come right out of high school and make $70,000 a year”, I know a lot of PhDs making half that!) I’d be very surprised to hear that there wasn’t a decent risk premium 100 years ago too… although of course a higher proportion of jobs had some risk, and we were all much poorer.

88 Careless July 11, 2012 at 12:18 pm

So you’re saying that David’s grandfather wouldn’t have accepted very slightly less pay for a safe work environment, or are you another one reading out the “all else equal”?

89 improbable July 11, 2012 at 2:28 pm

I’m sure he would have “accepted very slightly less pay for a safe work environment” but I’m also sure that wasn’t on offer. How much can an unskilled guy straight out of school make in the sort of comfy job where the biggest risk is that lunch might give him diabetes? Not 70k that’s for sure.

And I have no idea where you were going with “all else equal”… did you read my comment? I’m arguing _against_ Seth and David’s suggestion that you should, at the same time, change two things. It ain’t a free lunch just because you took up a paper round too!

90 improbable July 11, 2012 at 2:30 pm

Sorry “Careless” on reflection perhaps you were replying to Seth not to me.

91 byomtov July 11, 2012 at 4:36 pm

I’m sure he would have “accepted very slightly less pay for a safe work environment” but I’m also sure that wasn’t on offer.

Probably not.
The notion that all sorts of combinations of pay and working conditions are available to the worker is plainly false. There are simply too many rigidities and it is inevitable that in most cases choices are highly restricted. All these assumptions that assume each worker negotiates an individual contract are foolish.

92 Seth D July 11, 2012 at 10:29 am

I agree improbable, but it illustrates the principle.

93 Tony July 11, 2012 at 10:35 am

I don’t believe for a second that employees realistically estimate the cost of risk exposure in, say, heavy industry. The employer, on the other hand, knows exactly how much safety measures cost, and I think they know that the employee would take a smaller cash premium pretty much every time in lieu of providing a safer workplace. This is a market failure driven by a universal cognitive bias – “it won’t happen to me”.

The observation that employees have SOME wage benefits in the face of various risks doesn’t mean that the wage benefit would be ENOUGH to compensate for the risk in a laissez-faire environment. Some times I think that working-class men, in particular, are practically suicidal when it comes to confronting workplace hazards. (Revealed preference? No, just testosterone poisoning.)

I’ll say it again: in contrast to libertarianism, liberalism is about defending dumb people against clever people, and I think that’s the right thing to do.

94 improbable July 11, 2012 at 11:04 am

This isn’t my field, but whenever I have seen numbers posted about this sort of thing (mostly here!) they always seemed to say that workers were fairly consistent about counting one life saved as being worth N years salary, in terms of the risk premium demanded, at various occupations, places, and times. So I’m less sure than you that they are dumb, in aggregate. (Especially for things like coal mining where there is a long history to integrate mentally, and a community to do this. The macho young coal miner might, if his parents had moved to safer jobs elsewhere, take his machismo to the football field instead.)

Anyway so the empirical question is to ask what risk premium workers do demand. And whether there are occupations which are badly out of line, which may indicate dumb people being taken advantage of.

The other question is whether the general risk premium is high enough, perhaps we as a whole society would like to tip the balance… fewer orphans, more child poverty? This is much harder. But make no mistake there is no free lunch here, that is Alex’s point.

95 Brian Donohue July 11, 2012 at 11:11 am

“liberalism is about defending dumb people against clever people, and I think that’s the right thing to do.”

Well put and refreshingly candid. The elitist core of modern liberalism.

96 MD July 11, 2012 at 1:11 pm

Conservativism is about defending clever people against dumb people, which is also a bit elitist.

97 Paul Zrimsek July 11, 2012 at 2:00 pm

To the (limited) extent that it’s about any one thing, conservatism is about defending the cleverness of the people elites like to mistake for dumb people.

98 Brian Donohue July 11, 2012 at 2:51 pm

but dumb people are conservative. c/f Mill/Disraeli

What I’m about is being a really clever person trying to convince dumb people not to fall for the tripe being dished up by sort of clever people.

Does this make me an elitist?

99 MD July 11, 2012 at 3:25 pm

Why do you say dumb people are conservative? I would think that higher education and higher income people tend to be more conservative. I mean conservative like modern day American conservative, not conservative v. classical liberal of Mill’s time.

If you think you are a really clever person, and that others are merely sort of clever or even dumb, yes, you are sort of an elitist.

And anyway, I’m not sure what is wrong about being elitist. I like to make rye manhattans with Punt e Mes and homemade maraschino cherries rather than open up a can of Bud Light. I’m pretty sure that makes me sort of elitist, even if I am not an elite.

100 GiT July 11, 2012 at 3:39 pm
101 MD July 11, 2012 at 3:53 pm

The highest, but also the lowest. That paper also said:

Traditionally, the Republican Party has been made up of the elite, while the Democratic Party has received the votes of common people.

Despite the patterns we see for education levels, voting by income remains strongly patterned along traditional lines, with Democrats continuing to win the vast majority of low-income voters and Republicans doing best among the top 10%–20%.

102 Brian Donohue July 11, 2012 at 4:00 pm

I find the high education/ low income quadrant to be the source of much indignation.

103 John Thacker July 11, 2012 at 6:45 pm

And also, as Gelman’s book has noted, within each state, there is a positive correlation between income and education and voting Republican. However, the wealthier states are more Democratic, both with a change in the slope and intercept of the correlation.

A small amount of that is possibly a reverse causation, with Democratic-favored policies driving up land prices and cost of living, though perhaps that is driven by geography. Hard to say.

104 derek July 11, 2012 at 9:27 pm

Smart conservatives don’t waste their time in government. They run businesses, make money. Conservative politicians usually are second rate used car salesmen who saw an opportunity and can talk.

105 dearieme July 11, 2012 at 11:20 am

“liberalism is about defending dumb people against clever people”: in which category do you place yourself?

106 Tony July 11, 2012 at 11:46 am

I would say the dividing line is approximated by “has a spreadsheet and knows how to use it”. People who write down numbers and use basic mathematics to make real-world decisions using objective information vs. those who rely on their “gut” and make mistakes they later regret, like borrowing against their house to pay for a tropical vacation. It’s not really a matter of smart vs. dumb (I was being intentionally provocative) but more a matter of attitude towards life. I bet it’s also correlated pretty well with the Stanford marshmallow test.

I’m definitely in the analytic category and I’m not actually disdainful of people that aren’t. I just see the way they are systematically stripped of their money and it makes me cringe. I don’t think orienting laws to temper this effect is “subsidizing dumbness”, and I don’t think it will create more of it. What it will do is make life somewhat less unpleasant for people who are never going to prosper in a libertarian, technocratic society.

107 The Original D July 11, 2012 at 3:02 pm

+100 If I had a nickel for everyone who said “I’m not good at math”…

As it happens, con men have lots of nickels from those same people. How many of Madoff’s victims bothered to calculate the risks?

108 John Thacker July 11, 2012 at 6:21 pm

What makes you think that those same people are going to prosper in a liberal, legalistic, bureaucratic society? Substitute the complexity of modern, contradictory, changing government rules and regulations for the modern, changing market conditions. Will they necessarily gain? Are they necessarily as good as organizing, pressuring government, not being misled by their spokespeople, figuring out which school district to live in, figuring out how to get approval for their additions to their house, figuring out how to minimize their taxes?

109 John Thacker July 11, 2012 at 6:24 pm

As it happens, con men have lots of nickels from those same people.

And some of those con men are politicians who told people that they were trying to help them.

If someone can’t properly keep track of their own business, and their own profits and loss, and their own preferences, and what’s best for themselves, how can they possibly judge what’s best for the country, and whether politicians are telling the truth?

110 The Original D July 11, 2012 at 7:02 pm

Where is they learn how to do this again?

111 John Thacker July 11, 2012 at 7:51 pm

“Example is the school of mankind, and they will learn at no other.”

I’m not the one saying that people can’t be trusted to run their own lives, yet can be trusted to run others.

My view is that people know their own circumstances best, and they know their own circumstances better than they know that of others’. So I don’t see how you can be so paternalistic and favor universal suffrage. And I don’t see how people who can’t be trusted to figure out the complex world of the market could be trusted to figure out the equally complex world of the government bureaucracy. Each and every one of these paternalistic, irrational flaws applies equally well to someone facing decisions as a voter and as a person depending on a government.

I think that, once again, people are trying to compare the imperfect reality of the market to a hypothetical, perfect government that does not exist.

112 The Original D July 11, 2012 at 11:02 pm

There is such a thing as continuum. To suggest that perhaps some kinds of activities should be regulated to protect people from themselves is not the same as confiscating all their assets.

Conversely, people should not be incented or encouraged to do things that are a higher risk than they might realize. Like the housing bubble.

113 The Original D July 11, 2012 at 11:03 pm

And my point was that the same people who think we should “leave people alone” also frequently are the same ones who trash our school system. That is, they say that a lot of people are ignorant. Which is it?

114 John Thacker July 12, 2012 at 11:40 am

And my point was that the same people who think we should “leave people alone” also frequently are the same ones who trash our school system. That is, they say that a lot of people are ignorant. Which is it?

I’m fairly certain that there is broad agreement among all parts of the political spectrum that inner city schools are not as good as they should be. The difference is the recommended way to deal with it. Are you suggesting that people who call for more funding, paying teachers more, smaller classes, etc., are “trash[ing] our school system?”

By your argument, aren’t you saying that people who call for slashing funding to teachers or schools are praising our school system, by saying that it’s doing such a great job and people are so well-educated that we can afford to cut back?

Yes, people are more ignorant than they should be, but I do argue that people are capable of making choices– and note that it’s the fans of government here who are arguing that they can’t. However, it doesn’t matter, since they’ll have to make choices regardless. Better that they should make choices most closely related to their own lives and expertise than make decisions about what’s good for people far away whom they don’t know. In both cases they’ll have the opportunity to outsource some of these judgments and decisions, but they still must choose among various proxies, agents, and advisers.

For every choice in the private sector, there is a parallel in the public sector. For every way to outsource decisions in the public sector, there is an equivalent in the private sector. There are reasons to distinguish the two, and there are cases one can make for the public sector, but the capability of people to make decisions and judgements does not strike me as a distinguishing factor.

115 John Thacker July 11, 2012 at 6:43 pm

To put it another way, you don’t believe that people are capable of running their own lives, but you trust them to run yours?” Or do you favor a restricted franchise, or do you always expect to be one of the rulers, not the ruled, and not have the rules apply to you?

If all you do is favor a set of defaults, but allow people to bargain and make their own choices, knowing that some people will make mistakes, then we’re not very far apart.

116 The Original D July 11, 2012 at 11:06 pm

I outsource a lot of my decisions to other people. I trust the government to keep my drinking water clean instead of checking every time whether it actually is clean.

Some day — say, after a benzene spill in my neighborhood that I don’t know about — that may turn out to have been a bad decision. But government has eliminated a lot of the risk, so on a risk-adjusted basis I’m better off focusing my attention elsewhere.

117 John Thacker July 12, 2012 at 11:30 am

In the private sector, people also outsource decisions and expertise to other people. The market, through reputation and other factors, also eliminates risk and allows outsourcing decisions.

So again, it’s not a fundamental difference here. If you don’t trust people to outsource decisions to the private sector, why trust them to outsource decisions to politicians?

118 Pithlord July 12, 2012 at 3:47 pm

The basic social-democratic insight is, “In some categories of decision making, voice gives rise to better results than exit. In others, not.”

You are oblged by your ideology to deny that, just like the marxist you so resemble.But it is actually a complicated mix of empirical and normative issues to figure out, and you can’t just do it by pure thought alone combined with a few cynical remarks about politicians.

119 Yancey Ward July 11, 2012 at 11:58 am

All liberals live in Lake Wobegone.

120 JWatts July 11, 2012 at 3:08 pm

“All liberals live in Lake Wobegone.”

And are all above average, even among their own population.

121 Sam July 11, 2012 at 10:42 am

A compensating differential only raises wages if the workers have alternatives. Consider a model in which every firm has the risk of sexual harassment. In this case, the workers’ only options are to work at a firm with sexual harassment or to not work at all, and I’ll assume that the number of women whose marginal decision to enter the workforce would be changed by sexual harassment is small.

In that case, if the sexual harassment doesn’t induce workers to quit, then the firm doesn’t have to pay the workers more, because all of the workers’ alternatives also include sexual harassment. It might be possible for firms to attract women by having a harassment-free environment, but maybe it’s actually really hard to enforce rules about sexual harassment.

In reality, almost all firms have some risk of sexual harassment, there are substantial costs to moving jobs every time a worker is sexually harassed, and jobs are not transparent about their levels of sexual harassment when workers sign contracts. I suspect wasd’s explanation is correct, that women who earn more money are in a better position to report sexual harassment.

122 otto July 11, 2012 at 10:55 am

“Consider a model in which every firm has the risk of sexual harassment.”

You really need a model in which every firm has equal risk of sexual harassment. Which seems mighty unlikely.

123 Sam July 11, 2012 at 12:59 pm

Ok, valid point. But I actually think that what you really need is for every firm to have the same expected risk of sexual harassment, and I think it’s really hard to judge a firm’s level of sexual harassment before you work there.

Think about whatever field you work in, and then think of 10 businesses in that field. Can you rank them by their levels of sexual harassment? I sure can’t, and I can’t think of any reliable way to figure it out either.

124 The Original D July 11, 2012 at 3:03 pm

Ratio of male to female employees might be a good proxy, but I don’t how how you would get data on that.

125 GiT July 11, 2012 at 3:10 pm

Workers don’t pick their jobs from the set ‘every firm.’

126 ohwilleke July 11, 2012 at 9:18 pm

Your point is a good one that illustrates a lot about how sexual harassment and non-discrimination laws actually played out in the workplace.

These laws were not put in a place at a time when non-harassment and non-discrimination were consensus wildly held values, and even those who held to those values had trouble enforcing them at the time that these laws were enacted because sexual harassment and discrimination were sufficiently widely held norms that they were hard for managers to restrain. Some work places may have been better than others, but they were all universally bad, relatively to many of the work places that are successfully sanctioned by juries for breaking these laws today. (Contrary to otto’s comment below, sexual harassment or its absence are sufficiently binary components of the social norms that make it possible that a model in which every firm has equal risk of sexual harassment actually isn’t necessary, particularly in the case of phenomena which play out in individual cases in unpredictable ways that have more to do with the particular individuals involved than the overall firm unless strong social norms against it are in place.)

These laws had their effect mostly by articulating a clear moral standard and starting a virtuous cycle of peer pressure rolling. They were about changing social norms about how to run an organization across an entire marketplace all at once by changing the norms of the managerial class, where few firms strongly held those norms before, more than they were about punishing individual firms for being particularly egregious.

The vast majority of employment discrimination logically should happen in the hiring process and not the promotion or employee termination process, yet only a tiny portion of all discrimination enforcement actions related to the hiring process and successfully resisting a hiring enforcement suit even when you are breaking this law is pretty easy so long as one lies and purports to adhere to the law’s direction. Even though employment discrimination laws have almost no enforceable legal bite in the workplace, the power that the law has had to force discriminating employers to be covert has been sufficient to dramatically reduce discrimination in hiring relative to the pre-enactment situation, and sexual harassment laws have had a quite similar effect in having a corporate culture changing influence far out of proportion to the magnitude of the sanctions that have been successfully imposed for breaching them.

Now that those laws have done their magic, a generation of managers later, I suspect that the change in norms that they have wrought would be quite robust. They would not revert to the old pattern even if they could now that the new norms have been accepted in our society. For example, all but a tiny handful of firms that are small, religious operated, family run, or for some other reason effectively exempt from legal scope of sexual harassment laws and non-discrimination laws choose not to exercise their legal right to harass employees or discriminate with impunity. The changes in how we view social relationships in the workplace that these laws have brought about are so dramatic that only the most consciously historically accurate popular fiction set in periods before these laws were in effect depict worlds where sexual harassment and discrimination are widely accepted as a worker’s lot.

127 ohwilleke July 11, 2012 at 9:26 pm

“and sexual harassment laws have had a quite similar effect”

To really be precise, I should be clear that sexual harassment laws are in fact really just a particular way that laws against gender discrimination in the workplace are enforced and not a genuinely separate and distinct body of law. But, it is easy to miss that point in daily life because the fact patterns we talk about each set of laws addressing are so different from each other, and the origins of sexual harassment laws in anti-discrimination laws is as much a creature of historical accident as it is a logical necessity.

128 Enrique July 11, 2012 at 10:45 am

there is also a strategic element to these behaviors:


129 Paul Seabright July 11, 2012 at 10:57 am

I’m surprised how little weight is given in these discussions to the fact that many aspects of workplace harassment, sexual or otherwise, are non-contractible. No employer can make a credible promise to a future employee about the extent and nature of harassment a particular job will entail. This is a well known source of inefficiency: harassers will choose a higher level of harassment than the level on which they and the person harassed would have agreed if credible promises were available. The observation that, if such promises could be made, rational employees might choose higher wages in return for accepting a risk of workplace harassment (and that, in such a context, it might be right and fair to honor those contracts) does not detract from the fact that, in the world as it is, restrictions on workplace harassment may raise worker’s welfare. Whether they do is an empirical question, of course, but the hypothesis that they might is not some leftist fantasy that ignores basic economics. By now the hold-up problem IS basic economics.

130 John Thacker July 11, 2012 at 11:15 am

OTOH, a legal based approach also does not result in the work rules that everyone would choose if credible promises were available. The problem of the promises being non-credible affects regulation as well, because it makes it difficult to craft regulation. Both are inefficient.

Take harassment. Because of the difficulty of standards of proof, legal protections inevitably result in company policies (because of threat of lawsuit, etc.) that ban entirely voluntary actions, including voluntary relationships. Some of these bans are justified at least somewhat by theory (such as that a voluntary relationship is not voluntary when between people of different power, or that the relationship damages those who don’t get such a relationship). However, it’s not clear that liberals and progressives respect those theories, as we saw during Clinton’s Presidency.

To take another example, consider the case of the worker who got in trouble for donating their boss a kidney that the CT folks got upset about that. I can imagine them equally objecting that allowing the donation of a kidney to a boss could lead to coercion.

131 Paul Seabright July 11, 2012 at 11:32 am

Certainly, restrictions on workplace harassment face credibility problems of their own: the legal system does not have available to it all, or even most, of the mutually attractive bargains that individuals might be willing to strike, though it may have some that they would find preferable to those available in the absence of legal restrictions. This is now exactly where the debate should be: on an empirical footing comparing alternative inefficient outcomes. It strikes me as likely, for instance, that France polices sexual harassment in the workplace too lightly while the US polices it too strictly (the value of consensual workplace relationships, which is a real value, is probably exaggerated in France relative to the damage done by workplace harassment, while it is probably undervalued in the US). But that is my hunch, and I don’t pretend to have the systematic evidence that such a comparison would need.

132 John Thacker July 11, 2012 at 6:28 pm

This is now exactly where the debate should be: on an empirical footing comparing alternative inefficient outcomes.

Agreed. Which is why the libertarian and conservative emphasis on federalism and a diversity of situations provides better empirical tests and data than legislating a common, uniform, national standard than all must follow, no?

133 Paul Seabright July 11, 2012 at 7:12 pm

Yes indeed; I’m very sympathetic to that. Though it’s an open question how much diversity will in fact be chosen under federalism; this is likely to depend on the extent of geographical diversity in preferences and the political expression of those preferences.

134 John Thacker July 11, 2012 at 7:52 pm

Oh, well, I’d prefer alternatives to federalism that don’t depend on pure geography too. I simply haven’t heard of a more convincing arrangement, but I’m open to suggestions.

135 byomtov July 11, 2012 at 9:03 pm

If we were completely ignorant that might be true.

But at some point you have to look at the data you have and make a decision. Otherwise you’re forever arguing that we can’t do anything because we need still more data, ignoring the fact that delay is also a decision.

And, BTW, is efficiency as measured by GDP or something really the only standard you want to apply?

136 Barry July 11, 2012 at 11:43 am

I don’t think that that’s an accident; the ‘libertarian’ position uses ‘contract’ and ‘get another job!'[1] to get out of dealing with this.

[1] unless it’s dealing with a union, in which case getting another job is a cost/highly difficult, etc.

137 JWatts July 11, 2012 at 3:14 pm

“I don’t think that that’s an accident; the ‘libertarian’ position uses ‘contract’ and ‘get another job!’ to get out of dealing with this.”

No, the libertarian position is that using contracts and a mobile, free labor force is the appropriate response to this type of problem. Granted, it probably will not work in every case, but on the other hand neither does the current heavy regulatory approach work in every case.

138 Justin July 11, 2012 at 11:23 am

Don’t forget the public choice perspective, which I think suggests that the primary benefit of most workplace regulations doesn’t go to shareholders, management, or employees. They primarily benefit lawyers. (It is of course plaintiffs’ lawyers who lobby for these regulations, but management-side lawyers [like me] benefit at least as much.)

139 Barry July 11, 2012 at 11:41 am
140 sherparick July 11, 2012 at 11:58 am

For the full response by Peter Dorman to Alex’s and Tyler’s posts, see: http://economistsview.typepad.com/economistsview/2012/07/do-workers-in-fairer-or-safer-jobs-make-less-money.html or Peter’s own web site, Econospeak, http://econospeak.blogspot.com/2012/07/pay-for-oppression-do-workers-in-fairer.html.

“… Now it happens that this is a topic I know something about. In fact, I wrote the book on it. (OK, a book.) For the full story, read the book. Here I will make a few brief comments about the evidence.

I agree that fringe benefits that are essentially monetary in character, such as pensions and insurance, are subject to this sort of process. Workers really do trade money in one form for money in another, and unions bargain explicitly over this. It is also true that public insurance, like workers comp, is largely financed out of wages too, no matter how the laws are written.

It is not true that nonmonetary costs and benefits of work are compensated—not fully at any rate, and sometimes not at all. For an empirical demonstration, see this old study I wrote with Paul Hagstrom that has, to my knowledge, never been rebutted. Tyler links to a slightly less old Viscusi/Aldy lit review that cherry-picks shamelessly, not only in its selection of what counts as a valid study, but (especially) in which results of the authors they choose to report.

For those who don’t want to go to the sources, here are the two fundamental empirical problems with studies that claim to show wage compensation for things like occupational safety: (1) They use industry-based measures of which jobs are risky, but they ignore all the other industry-level determinants of wages, like concentration, capital intensity and percent unionized (usually). (2) They show signs of potential publication bias, where specifications are selected that yield the “right” result. What signs? They don’t provide summary data on the full range of specifications tested (“taking the con out of econometrics”), so that the reader can determine whether the reported specifications are outliers or not, and the results are hardly ever tested on subsamples. Think for a moment about this last point: if there are compensating wage differentials for women exposed to sexual harassment, this should apply to black women and white women, unionized women and nonunion women, women in blue collar jobs, white collar and pink collar, and other plausible breakouts. If not, you would need to have a story to explain why some get it and not others. (Or why some subsamples even have exacerbating differentials, which I found with occupational safety—the “sweatshop effect”.) There is a small literature on subsamples in wage-risk regressions, and they are all over the map.

Note that the Joni Hersch study Tyler links to is vulnerable on all these counts. It doesn’t consider interindustry differentials. There is only one specification reported. No subsamples. Not convincing.”

I would also say that even if employees overall have decrease in wages for protection from “Black Lung” disease (http://www.npr.org/2012/07/09/155978300/as-mine-protections-fail-black-lung-cases-surge), to use one safety issue in the work place, or sexual harassment, this is a kind of insurance payment to preven being one of the lucky duckies caughing out their lungs or being told to bend over the desk and take it in the rear if they want to keep their job. Most workers (we hope) don’t suffer these experiences, but for those who do, (and it appears principally due to bad luck (karma, etc.)) who do, the results are devastating and not to be tolerated in a society that would like to claim it is “enlightened” and made up of people who feel empathy for human beings who don’t necessarily belong to their own tribe. (Really, libertarians should read Adam Smith and David Hume more, and Ayn Rand less, and perhaps the little man that Adam Smith says we all have in insider our heads in TMS would not be such a “dick” in their kopfs.)

141 Careless July 11, 2012 at 12:40 pm

Yes, it was an amusing exercise in agreeing with Alex while claiming to disagree with Tyler about something Tyler didn’t say.

142 Wonks Anonymous July 11, 2012 at 12:02 pm

“Now. what are you expecting to follow this paragraph? Surely, some empirical evidence that the model is wrong. None follows. Henry seems to think that economists have never thought about their assumptions or tested their models.”

Farrell doesn’t cite Dorman on wages, but he does reply to Yglesias argument about unemployment in a frictional labor market due to regulations by citing an empirical paper by Dean Baker & others, along with a reply from Heckman saying neither the standard RBC literature nor Baker et al’s critique can definitively say anything (although Heckman indicated some of his own work better supports the standard story).

143 Robert Reichardt July 11, 2012 at 12:53 pm

Another example is the teaching labor market. There are some schools that are hard to work at and schools that are easy. If the schools exist in the same district you see experienced teachers flowing to the “easy” schools. If the schools exist in different districts, then the district with the “easy” schools will be able to attract experienced teachers with lower salaries.

144 ohwilleke July 11, 2012 at 8:16 pm

Except that it doesn’t often work that way in real life.

Schools are considered “hard” because the students come from poor families. Poor families are concentrated in districts with weak tax bases and often aren’t willing to tax themselves enough to increase school budgets at schools in which their own children are often failing. Schools with weak tax bases can’t afford to pay their teachers as much and also have a lower cost of living since rents and housing prices in low income districts are lower. Schools with kids from affluent families are considered “easy,” also tend to have fatter tax bases, also are more willing to tax themselves to pay for schools in which their children are thriving, also have more support from private grants from local PTAs etc, hence the “easy” school districts can pay more and have to pay more so that the teachers that they hire can afford to live in or near that district.

For example, my own children’s central city elementary school in a wealthy neighborhood within the central city had a PTA that managed to raise enough private funds to put a union wage paid teacher’s aide in every classroom, despite receiving the same per student funding as every other elementary school in the district (or less since there are “at risk” kid premiums in the funding formula) where schools can’t afford this luxury. Tellingly, one off the main fundraisers for my children’s school was a house tour of the homes of the school’s affluent neighbors.

145 Dick King July 11, 2012 at 1:11 pm

Wage increases are taxed, but working condition improvement changes, even luxuries such as the racquetball court mentioned above are not taxed.

Furthermore, raises must be shared with a spouse but working condition improvements cannot be* .


* unless the spouse also likes racquetball and can be snuck in, in which case it’s an even bigger plus :-) .

146 someguy July 11, 2012 at 1:32 pm

You cannot actually make your employees wet their pants by refusing bathroom breaks. The Nabisco employees sued and won.

Forget econometric proof. Henry’s theoretical model is flawed.

If firms capture a surplus via crappy working conditions because of the unequal bargaining power of workers – you cannot reallocate via regulations curtailing crappy working conditions – because as long as bargaining powers remain unequal firms can just reallocate away those gains in the form of less monetary pay.

147 Go Kings, Go July 11, 2012 at 1:47 pm

Why does high risk of sexual harassment move opposite to expensive workers comp bills have the same effect on wages? In all three cases, employer cost increases (by purchase of insurance, whether public or private, and regulatory compliance.) Why are only workers comp bills passed through to employees?

For example, no one has more sex, age, race, disability, et al law suits than California, forcing California businesses to purchase insurance far more expensive compared to jurisdictions where it is difficult to win such cases? Shouldn’t California employees, especially in the protected groups, suffer lower wages than comparable Utah employees?

148 derek July 11, 2012 at 9:42 pm

As is said often, wages are sticky. What isn’t sticky is the opportunities to work elsewhere if your boss is an idiot. California is losing workers and businesses, partly due to the high regulatory costs. The tough rules and generous awards may benefit individuals, but on the whole takes opportunity away from most people.

Similar dynamic in unionized shops, private sector. Benefits and work rules, all geared for the protection of the workers, cause employment in these industries to decrease. Individuals who have seniority do just fine, in diminishing numbers. Market pressure makes the pool of individuals who have those benefits shrink. We even see two tier wage and benefit scales for new hires.

149 ohwilleke July 11, 2012 at 9:52 pm

Sexual harassment is taken so seriously by employers in part because it is largely an uninsurable risk that present a risk limited in dollar amount from any given incident, in practice, only by the generosity of the jury.

For reasons that have roots mostly in the moral hazards involved, insurance contracts exclude coverage for intentional conduct and for breaches of contract, and there is generally no vicariously liability of an employer for intentional conduct by an employee who is acting directly contrary to company policy. Since sexual harassment is defined in a matter the makes it an intentional wrong by the harassing individual, insurance usually doesn’t cover claims against the most culpable individuals. A company that adopts a compliance sexual harassment policy and doesn’t every communicate a tolerance for violations of its to anyone is almost bullet proof from liability, so firm exposure to the kind of liability for unintentional conduct that drives up insurance costs isn’t very meaningful. Insurance costs in California are high, but not meaningfully because of greater sexual harassment liability (which is, incidentally, a uniform federal law for the most part).

So, unlike fringe benefit laws like worker’s compensation, the employer costs associated with sexual harassment laws have very litle to do with insurance costs. This is not the mechanism by which sexual harassment laws (unlike other tort liability laws) impact firms economically.

There are compliance costs associated with sexual harassment laws (which make particular sense in big organizations where the worst case scenario from a big judgment is such a large dollar figure compared to small businesses with fewer assets and where informal control of employees by top management is so much weaker), but honestly, the out of pocket costs of generalized complicance with sexual harassment laws is trivial relative to a company’s payroll, and grows smaller over the years as public understanding of what fits the definition of sexual harassment is more widely assimilated into the culture (since most prevention measures focus on defintions of what is prohibited) and as legally compliant workplace policies (which cost money to draft and adopt the first time around) become commonplace and SOP for businesses with many employees.

The employer costs from sexual harassment laws comes almost entirely from defending lawsuits and threatened lawsuits where there are credible allegations of sexual harassment, and from situations when an employer is forced to choose between firing a harassing employee who is otherwise a productive one and facing exposure to legal liability for not doing so – a lose-lose situation for the firm. If the harassing individual is a key employee, maybe a genius (and those are the employees who manage to get away with harassment despite employer knowledge in reasonably well run businesses) or for some other reason has economic leverage vis-a-vis the firm (tenure, familiy ties, an ownership interest in the company or ties to a key source of financing), the compliance cost to the firm forced to fire this person or buy off the accuser, if they want to comply or settle, is high indeed.

150 The Original D July 11, 2012 at 1:58 pm

Wouldn’t an easy way to model this be to examine companies that make “best places to work” lists? Are their salaries higher or lower than industry averages?

151 Jason July 11, 2012 at 2:48 pm

Which model are we talking about?!!!

I assume we are talking about deadweight loss. However, arbitrary workplace rules also introduce deadweight loss.

A rule at a company for mandatory searches produces deadweight loss and increases the company’s cost in exactly the same way as a rule across all companies forbidding such searches. The question is whether the deadweight loss is overcome by the amount of stealing.

With perfect competition, that deadweight loss (search costs) and priced externality (workplace stealing) is priced in the market and firms that do too much searching eventually lose money and go under.

If you don’t have perfect competition, then there is no telling whether that deadweight loss of searches is effectively pricing the stealing. And there is no way for Yglesias’s example ($8/hour + searches vs 7$/hour + no searches) to form. That is what is wrong with the Economics 101 analysis. Mandatory drug tests are not priced into the labor market effectively and so are probably overdone or at least inconsistently applied.

You can’t just ignore perfect competition.

152 Joe Smith July 11, 2012 at 3:06 pm

All else being equal, a workplace that allows sexual harassment will be able to pay the harassers lower wages. Since harassers tend to be more senior employees with higher salaries than the victims, a company might save money on balance by allowing harassment. The company could go further and high designated “sex objects” as a perk for the senior managers (oh wait, that already happens implicitly).

153 Harold McClure July 11, 2012 at 3:07 pm

“Henry gets the economics wrong (the model does not require perfect competition)”. Not to pick on a minor point but there can be differences in how labor markets react in the presence of monopsony power v. perfect competition. See my 1994 Journal of Labor Research paper – “Minimum Wages and the Wessels Effect in a Monopsony Model”:

The Wessels model suggests that firms respond to increases in the minimum wage rate by decreasing the level of fringe benefits — an action which produces an inefficiency effect that lowers workers’ utility and the supply of labor. Standard models of monopsony, however, argue that wage floors prevent the exercise of market power and increase employment. I show that wage floors, even with fringe benefit curtailment, may increase employment by lowering the marginal expense of labor. Employee utility and employment will rise somewhat but not as much had the firm acted competitively in setting both wages and fringes.

154 Glen July 11, 2012 at 3:08 pm

This post is full of win. Best thing I’ve read this cross-blog debate thus far. I am strongly reminded of Alex’s posts back in 2004 about the “nonwaivable warranty of habitability” laws. Many of the same economic principles, as well as economically unsound responses, are in play.

155 James Vonder Haar July 11, 2012 at 3:11 pm

Granting your argument, how do we differentiate between acceptable interventions in labor rights and unacceptable ones? Most libertarians will grant that protecting against bodily harm, or theft, even against an employer, is a justified use of government power. I don’t think that changes even if you believe that such measures will reduce wages, as Tyler’s argument would indicate.

I think it also depends on what rights you think ought to be contracted away by default as part of the the employment contract. A significant portion of the liberal camp is perfectly fine with an employment contract including sex work, but the question is whether, in the absence of something stipulating it, sex work can be assumed as part of any employment contract. As a matter of custom, this is clearly not the contract that most employees believe that they are signing. The law reflects this through sexual harassment provisions that make sex work an opt-in part of employment contracts rather than opt-out. I don’t see this as a tremendous violation of liberty.

156 Misaki July 11, 2012 at 3:28 pm

>All else equal, an improvement in workplace conditions will reduce wages. Now that looks slightly different than “A burger and fries costs more than a burger” but it’s really not.

This may be true “at the margin”, but if you imagine rising standards of workplace conditions, possibly as the result of regulation, then the expectation people will have for workplace quality will rise and the supply of labor for a workplace with higher conditions will shift back to “trend”.

Regarding what that trend is, as Matt Yglesias pointed out the best way to improve workplace conditions is full employment. This can be achieved without higher government spending, inflation, or trade barriers as described here:

157 Steve Sailer July 11, 2012 at 4:06 pm

My mother became a bank teller out of high school in the late 1930s. She said the pay was very low, but it was a highly respectable job. Presumably, respectability of employment was worth sacrificing short term income for to maintain her value on the marriage market.

158 Vernunft July 11, 2012 at 5:56 pm

Matty made a smart post? That’d be a first. Dude seems not to grasp the most basic of concepts.

159 ohwilleke July 11, 2012 at 7:32 pm

“better working conditions (burger plus fries) cost more (lower wage) than worse working conditions (burger alone).”

The biggest flaw in this logic is that in the case of working conditions, unlike the case of consumption decisions, production of the good and consumption of it are not independent of each other. If better working conditions improve productivity then better working conditions may indeed lead to higher wages. Comparisons to fringe benefit and vacation regulation, which involve things that are consumed separately and distincting from doing the work itself, aren’t really valid. Working conditions are something different than merely being non-monetary compensation.

A classic example would be the Japanese practice of mandating extreme levels of cleanliness and order in usually chaotic environments like automobile repair shops – the clean and orderly working conditions mandate may not have had a productivity motivation, but the effect was to reduce workplace accidents, to reduce mistakes by those making the repairs, and to increase worker productivity despite the increased time spent on seemingly unnecessary cleaning and putting things away tasks when they could be actually doing the work. There is every reason to believe that the increased productivity and increased safety result would follow in conditions where a Japanese standard for order and cleanliness in automobile repair shops would work almost as well if imposed by government regulation as it does when imposed by management. Indeed, the story of the last century of the manufacturing industry is essentially one of just that happening via government regulation and resulting in increased productivity partially attributable to the working condition improvements.

Likewise, union shops are on average more productive than non-union shops, despite a reasonable inferrence that the union shops have better working conditions.

160 John Thacker July 11, 2012 at 7:58 pm

There is every reason to believe that the increased productivity and increased safety result would follow in conditions where a Japanese standard for order and cleanliness in automobile repair shops would work almost as well if imposed by government regulation as it does when imposed by management…
Likewise, union shops are on average more productive than non-union shops, despite a reasonable inferrence that the union shops have better working conditions.

So you’re discounting the possibility that union shops simply exclude the less productive workers and encourage firms to overinvest in capital relative to labor, increasing unemployment? You’re making the “if everyone went to Harvard, we’d all be geniuses” argument.

Alex’s argument of course also predicts that higher wage and compensation shops will have higher average productivity. They have to, to stay in business. But that doesn’t mean that less productive workers don’t deserve jobs, and it doesn’t mean that you could raise productivity by raising wages without costing jobs.

161 ohwilleke July 11, 2012 at 8:48 pm

I am saying that Alex’s argument leads to deceptive inferences because it rests on the counterfactual assumption that working conditions only function as a form of fringe benefit compensation for workers rather than impacting the productivity of the workplace directly, and because it also puts a great deal of weight on the desirability of a job compared to jobs that take comparable skill levels, when econometrically this factor is so trivial in influencing wage levels that it is hopelessly obscured by noise from all sorts of other second and thrd order factors that have similar influences on wages.

Alex is involved in the fundamentally misguided, and to some extent latently politically motivated, errand that economists seem to love of trying to tease out the economic impact of some trivial influence on wages, when that factor is dwarfed by factors that loom far larger in real life.

To use an analogy from physics – general relativity can allow you to make a slight adjustment to the expected orbit of the planet Mercury around the sun, which is except for the very slight adjustment perfectly explained by Kepler’s laws which can be derived entirely from first principles from Newton’s very simple F=GMm/r^2 formula for the law of gravity. Focusing on the correction, however, makes absolute no sense until you have pinned down all of the constants that are fed into the Newtonian equation (the mass of Mercury, the mass of the Sun, the mass and location of the other planets, the precisely location of the planet Mercury, the precisely value of the gravitational constant G) with great precision. What Alex is trying to do is like trying to draw conclusions from the tweaks in the law of gravity arising from general relativity in a situation where nobody has yet pinned down the exact form of its Newtonian approximation or the quantiative inputs that together set the baseline condition and are profoundly more important. As a result, its argument leads naturally to counterfactual expections.

Alex, like so many economists, starts with theory and attempts to reach a conclusion by logic alone from first principles in a system that is sufficiently messy that it makes much more sense to start with the statistical trends and work one’s way back to empirical relationships that implicitly show which factors (like skill, education, productivity, and local cost of living) are important, and which (like workplace conditions) have the opposite statistical relationship to wages that a simple minded Econ 101 analysis of the type that Alex employs would suggest. Econ 101 reasoning works well in systems that are well understood by the people using it (who unconsciously abstract away irrelevant factors) that involve simple tranasctions (e.g. commodity markets), and far less well in rich, multiply interrelated environments where there is deep underlying reasons in the way the world is organized for an “all things being equal” disclaimer being not just unlikely, but inherently counterfactual.

We need more descriptive economists in the world, are fewer theoretically oriented ones that our graduate schools are so insistent on churning out.

162 Steve Sailer July 12, 2012 at 2:11 am

I agree. My experience in life, much to my surprise as an old econ major, is that better working conditions go together with better pay and better productivity.

This was an argument that ran for decades in Victorian England as the Earl of Shaftesbury tried to get a law through Parliament to prevent five year old boys from working inside chimneys. The economists explained that people would freeze to death come winter if master chimney sweeps couldn’t use five year old boys. Finally, in 1875 after 35 years of trying, Shaftesbury got the law through. And everybody in England froze. No, actually chimneysweeps immediately figured out ways to sweep chimneys that didn’t involve 5 year old boys.

163 msgkings July 12, 2012 at 11:25 am

Nice post, ohwilleke

164 ohwilleke July 11, 2012 at 8:26 pm

“people exposed to a higher risk of death are paid more.”

Just not so. Who are the workers expose to the highest risk of death, by a long margin, compared to other workers in our economy? Mostly, farm workers. So obviously, farm workers must be the best paid low skill workers in the economy, by the long shot, right?

Not so much. In fact, they are some of the most poorly paid workers in the economy and have working conditions that are far worse in other respects as well. Rather than improving wages or working conditions, farms find employees that no one else will hire, pay them a pittance, and treat them like shit. In this case, the guys at Harvard writing the linked paper seeing the world through equation filtered glasses peering at government statistics miss the realities that are obvious to anyone remotely familiar with economic and social history, the history of the labor movement, or the way the gray market in labor (that conveniently doesn’t make its way into government statistics that are easily available to lazy econometric modelers) operates.

165 derek July 11, 2012 at 9:49 pm

You forget the ‘all else being equal’.

Compare linemen and farm workers. Both dangerous jobs. One gets paid very well, the other not. All else being equal doesn’t apply because there are skill requirements, supply of labor, etc.

166 ohwilleke July 11, 2012 at 10:10 pm

The are lots of jobs that are comparatively skilled to farm work (or less so), draw on the same supply of labor, are in the same geographic regions, etc. (e.g. being an orderly in a nursing home, a job favored by high school dropout ex-felons since they aren’t barred from it), that pay better and pose far less of a risk of death.

For that matter, take exactly the same skilled job, e.g. welding, at both an urban construction site and at a farm. The risk of death on the job at the urban construction site is vastly lower than on the farm doing the same work, and yet the urban construction site work is almost always going to pay more than the farm position, even if both jobs are done by the same person (many farmer workers spend the off season working construction in the city). What the employing firm can afford to pay is going to usually dwarf any risk premium.

“All other things being equal” is a sensible disclaimer when it concerns factual points that are severable from each other. But, when the risk of death involved, the relevant facts are almost never severable and the non-coincidental, non–independent facts that drive down wages in high risk jobs dwarf the effect of any residual wage premium.

Put another way, the econometics folks who devise the models that show wage premiums are using inappropriate statistical methods that rest upon implicit assumptions that are true where there are being applied in these cases. If you control for all of the non-independent factors that really matter, your highly filtered end result is junk.

167 derek July 11, 2012 at 10:51 pm

I agree with you. And yet we as employers face regulations where a very specific and narrow problem ends up being addressed by rules that are applied to all, even when there isn’t a problem to be solved in most cases.

168 Will McLean July 11, 2012 at 10:48 pm

Farm workers aren’t in the top four for death rates.

And the category that’s #5 is farmers and ranchers. That includes the owners driving tractors, which seems to be one of the main causes of death. Lettuce picking, not so much.

169 byomtov July 11, 2012 at 9:09 pm

All else equal, an improvement in workplace conditions will reduce wages. Now that looks slightly different than “A burger and fries costs more than a burger” but it’s really not. Think of the worker as buying better working conditions with a lower wage, thus better working conditions (burger plus fries) cost more (lower wage) than worse working conditions (burger alone).

Actually, it really is different.

McDonald’s has a menu with a variety of choices, and if those don’t appeal you can easily go elsewhere. That’s not true, in general, of potential employers. How many companies ask prospective employees whether they’d like to sign up for “being sexually harassed – daily, weekly, monthly, never,” with suitable wage variations?

170 derek July 11, 2012 at 10:11 pm

Indeed. You would think that the appropriate policy response to this situation is to make sure that nothing is done that would decrease the opportunities for people to find another employer. And would decrease the costs of moving from one employer to another.

171 Tom July 11, 2012 at 10:45 pm

All else equal, an improvement in workplace conditions will reduce wages. Now that looks slightly different than “A burger and fries costs more than a burger” but it’s really not. Think of the worker as buying better working conditions with a lower wage, thus better working conditions (burger plus fries) cost more (lower wage) than worse working conditions (burger alone).

Don’t better “workplace conditions” involve greater capital per worker and thus higher wages?

172 GiT July 11, 2012 at 11:28 pm

‘Dare one mention it,’ indeed. Reads like the Hersch paper is a mess.

From Farrell’s updated response:

“On the specific question of sexual harassment, an ungated version of the relevant article is available here for those who don’t have subscriptions to the AER. It’s indeed a remarkable piece. One of Alex’s commenters suggests that the author doesn’t do anything to convince anyone that the causal story is a good one; this is (very, very slightly) unfair. The author does mutter a bit about not having an instrumental variable, and waves vaguely in the direction of a strategy to identify causation. But what she does not do, at any point whatsoever, is to think about whether or not complaints about sexual harassment serve as a good proxy for actual sexual harassment in the workplace. As Alex’s commenter notes, there is a strong case that the causal story is precisely the reverse of what the author claims to have found- less well paid employees may be less likely to complain about sexual harassment. Not only that, but this prima facie provides a much better explanation of the findings than the author’s own argument. This bit from the paper is genuinely priceless.

“The log wage difference between a job with zero sexual harassment risk and a job with the mean sexual harassment risk is 0.0155, or about 25 cents per hour for women, and 0.0252, or about 50 cents per hour for men. The large compensation for sexual harassment risk for men is surprising. One possible explanation is that since men infrequently file sexual harassment claims, those claims that are filed are particularly egregious, and exposure to such risk warrants a larger compensating differential than received by women.”

So, on the one hand, we have a Tabarrok-world story in which power relations don’t exist, everyone is contracting fairly with everyone else, people are compensated for the higher risk of sexual harassment in particular industries with higher wages, and men demand and receive particularly high compensation for the risk of being sexually harassed in harassment risky industries. I don’t think that this claim passes the laugh test, myself. Clearly Alex disagrees, and I look forward to his fuller explication of this quite striking empirical finding. On the other, we have a story in which different levels of industry compensation are associated with different levels of bargaining power and reinforcing social norms, so that the likelihood of reporting sexual harassment varies substantially with compensation. This provides a far more plausible explanation of the awkward finding (it’s prima facie believable that working class guys are less likely to report sexual harassment, especially by male supervisors, than are men in better paid professions)”

And a commenter notes:

“That paper Tabarrok quotes is generally terrible. It has numerator-denominator bias in the calculation of the exposure variable, for starters; it doesn’t adjust for individual confounders, so it’s not able to explore relationships between workplace variables or individual variables and the supposed relationship (let alone account for variation between workplaces within industries); the author hasn’t allowed for the possibility that the data is missing not at random; and hourly wages are calculated on the assumption that “usual hours worked per week” are correctly reported in the CPS. Also, is the CPS a probability survey and if so did the author use probability-weighted sample estimates for denominators?

Plus of course there’s the completely obvious reporting problem that Henry identifies in his post, that throws out the entire conclusions anyway.”

173 Steve Sailer July 12, 2012 at 2:14 am

As Kingsley Amis said in Lucky Jim, generally speaking, there is no end to the way that nice things are nicer than nasty things.

174 R July 13, 2012 at 3:17 am

Women exposed to sexual harassment in the workplace have higher wages than women whose jobs don’t expose them to sexual harassment. But what about wage differential — how does that play in? Shouldn’t women exposed to high levels of sexual harassment be paid more, on average, than their male counterparts, assuming that their male counterparts aren’t exposed to the same levels of harassment? How does that factor into these models?

175 Welcome in the webshop of Signature Gravure. Master engravers since 1974 July 14, 2012 at 10:33 am

hey there and thank you for your information – I’ve definitely picked up anything new from right here. I did however expertise several technical points using this website, since I experienced to reload the site many times previous to I could get it to load properly. I had been wondering if your hosting is OK? Not that I am complaining, but sluggish loading instances times will sometimes affect your placement in google and could damage your high quality score if advertising and marketing with Adwords. Well I am adding this RSS to my e-mail and could look out for much more of your respective intriguing content. Ensure that you update this again soon.

Comments on this entry are closed.

Previous post:

Next post: