Myth: The Wage Gap is Caused by Men’s Higher Pay for Dangerous Jobs (wage gap series, part 10)

(This is one of a series of posts on the wage gap.)

Over on Amanda’s blog, reader “JenK” writes:

Men are more willing to take on dangerous jobs so can find better paying jobs than those who are not willing to risk their lives.

This is an argument I’ve seen before. When anti-feminists explain why the gender wage gap doesn’t exist or is justified, they frequently claim the wage gap reflects men being paid more for taking hazardous jobs or dangerous jobs. Often men’s rights activist (MRA) Warren Farrell is cited. The following arguments are typical:

• John Leo: Farrell argues that many men outearn women by a willingness to take risky and dangerous jobs as well as work that exposes them to stress and bad weather…
• Arrah Nielsen (from the IWF’s website): The real reason than men tend to out-earn women is the choices they make. Men are far more likely to take unpleasant and dangerous jobs, what Farrell calls the “death and exposure professions.” For example, firefighting, truck driving, mining and logging — to name just a few high-risk jobs — are all more than 95 percent male. Conversely, low risk jobs like secretarial work and childcare are more than 95 percent female.
• Glenn Sacks: Of the 25 most dangerous jobs in the United States (according to the U.S. Department of Labor), all of them are overwhelmingly or exclusively male. Over 90% of American workplace deaths and serious injuries occur to men. It is not unfair in the least that dangerous jobs pay more than safe jobs at the same skill level.

The anti-feminist argument here sounds logical and just. It’s true that men are much more likely to die or to be injured on the job than women. Surely no one would be willing to risk their life without getting paid a premium for it; and no reasonable person would argue that extra pay for extra danger is unjust. So how could feminists object to a “danger premium” that raises men’s wages?

The problem is, there is no premium for dangerous jobs. And since the “danger premium” doesn’t really exist, it can’t explain the wage gap.

This post will first look at some general evidence, from the Bureau of Labor Statistics, showing that high pay doesn’t equal high wages. Next, I’ll discuss the dubious right-wing assumptions implicit in the belief that dangerous jobs are paid for with higher wages. Finally, I’ll briefly discuss some of the peer-reviewed economic studies showing that high risk isn’t associated with high pay (and is even associated with lower pay, for non-union workers).

There is no premium for dangerous jobs.

Let’s look at some graphs (all graphs in this post were taken from the Bureau of Labor Statistics website). Here are some of the most dangerous industries to work in in the USA, based on on-the-job deaths:

Just looking at that graph should make people suspicious of the “high risk = high pay” myth. Yes, construction workers and miners earn decent pay, but agricultural workers? They face the highest risk of death, and get paid less than almost any other class of workers in the USA. From a BLS page entitled “lowest paying occupations in 2002“:

If danger jobs really paid a premium, we wouldn’t expect the most dangerous industry in America to be the second lowest-paid. Indeed, when the Bureau of Labor Statistics investigated job traits that are associated with wage premiums, they found that “Job attributes relating to … physically demanding or dangerous jobs… do not seem to affect wages.” Here’s a bar graph. As you can see, what pays most is specialized knowledge. The very tiniest bar, all the way over on the right, that’s actually slightly negative? That’s the “death and exposure” effect on wages Warren Farrell is talking about.

The right-wing economic assumptions behind the anti-feminist economic theory

Many anti-feminists are conservative or libertarian in viewpoint (the IWF, for example, exists chiefly to put a “good for women” face on whatever the Republican party’s current talking points are). However, some MRAs – including Warren Farrell and Glenn Sacks – think of themselves as liberal on many issues, despite their opposition to feminism. This makes their easy acceptance of right-wing economic assumptions implicit in the “high risk = high wages” theory somewhat surprising.

The key right-wing assumption – one frequently used to argue against policies such as the minimum wage and worker protection laws – is the belief that the free market produces the best possible outcome for workers. Obviously, workers would never accept jobs that risk life or injury without getting paid extra for it, right?

Well, no.

Believing that high risk is paid for by a wage premium means making a lot of assumptions; and if even one of those assumptions is off-base, then risk and wages might not be connected at all. From an article by economists Peter Dorman and Paul Hagstrom:

The theoretical case for wage compensation for risk is plausible but hardly certain. If workers have utility functions in which the expected likelihood and cost of occupational hazards enter as arguments, if they are fully informed of risks, if firms possess sufficient information on worker expectations and preferences (directly or through revealed preferences), if safety is costly to provide and not a public good, and if risk is fully transacted in anonymous, perfectly competitive labor markets, then workers will receive wage premia that exactly offset the disutility of assuming greater risk of injury or death. Of course, none of these assumptions applies in full and if one or more of them is sufficiently at variance with the real world, actual compensation may be less than utility-offsetting, nonexistent, or even negative – a combination of low pay and poor working conditions. [Source: Dorman and Hagstrom, “Wage Compensation for Dangerous Work Revisited,” Industrial and Labor Relations Review vol 52(1) Oct 1998]

What would make a labor market less than perfectly competitive? Many things. Feminists and liberals are likely to think of the effects of discrimination and persistent unemployment, which may leave some workers without the option of refusing to take a low-paying, high risk job. There are also industry premiums – workers do not move freely between industries, and some industries simply pay higher than others, in a pattern that cannot be reliably accounted for by skill requirements, education, risk, etc..

And of course, workers often lack the ability to accurately access risks. For instance, an agricultural worker may assume that she or he (most likely he) isn’t doing anything risky if his job doesn’t involve operating heavy-duty farm equipment; but he’s far more likely to be killed on the job if his duties involve driving. And the construction worker hanging from a girder thirty stories above the ground? He’s much less likely to be killed than the construction worker who stays on the ground driving a pick-up. (Leigh & Garcia, “Some problems with value-of-life estimates based on labor market data” Journal of Forensic Economics, Spring-Summer 2000 v13)

Because workers do not move freely from one industry to another, differences in how much different industries pay may prevent wages from being perfectly competitive. (As I’ll explain later this post, this is a particularly important factor when looking at wages and risk).

The point is, the assumption that the marketplace compensates workers for risk is, in the end, another example of blind ideological faith in the market to always produce the best outcome. We should be skeptical of such assumptions.

What academic studies have found

Several academic studies have found a significant connection between risk and higher wages. These studies generally don’t include agricultural workers – which is possibly a problem, since this cuts out the US workers who face the highest risks for the lowest pay. Furthermore, these studies usually don’t account for the differences in pay between industries – meaning that they can easily mistake the higher industry wages in an industry like construction or mining, with higher pay for risks.

How do we know that higher average pay in those industries aren’t premiums paid to workers in physically risky jobs? By comparing employees who face comparable levels of risk in different industries. A secretary working for a mining firm is not more likely to die on the job than a secretary working for an elementary school, for example. But when economists J. Paul Leigh and Jorge A. Garcia compared clerks across industries, they found that the so-called “danger premium” paid to construction and mining workers applied even to clerks facing no danger. The standard economic theory – stating that firms pay a premium to workers facing a higher risk of death or injury – cannot explain why a construction firm would choose to pay a low-skill clerk much more than an insurance firm would.

Dorman and Hagstrom’s analysis (pdf link) found that if industry wasn’t accounted for (and agricultural workers weren’t included), higher risk seems to be associated with higher wages. But once other factors were accounted for, there was almost no association between risk and pay. And what little association existed was negative – that is, workers who face a higher risk of death actually get paid lower wages than similar workers facing less risk.

This “negative premium” – workers getting paid less for facing risk – only happens to non-unionized workers. This result is not easily explained by conservative economic assumptions. It is, however, not unexpected to left-wingers, who would expect that worker bargaining power would have more to do with wages than risky work conditions.

Conclusions

First conclusion: The anti-feminist argument that the gender wage gap is (partly or fully) caused by justified higher pay for men who take on riskier work is not true. Evidence shows that taking on risky work isn’t associated with higher pay.

(Note that a related argument made by some MRAs – that sexist occupational segregation leads to men being more likely to be injured or killed on the job – holds true. That is sexist, and unfair. Men’s greater likelihood of workplace injury and death has nothing to do with the wage gap, but that doesn’t mean it’s not unjust.)

Second conclusion: The widely-shared conservative assumption that the market produces just and fair outcomes is not supported by looking at how the market compensates for risk. Workers who risk their lives often receive very low compensation, and for non-unionized workers they may be paid even less than similar workers in less risky jobs. Quoting Dorman and Hagstrom:

In plain terms, nonunion workers in dangerous jobs are, in many cases, simply unlucky; they have found their way in to situations of high risk and low pay and would presumably move to a better job if they could. …

From the perspective of public policy, dropping the assumption that risk coefficients fully reflect workers’ desired tradoffs strengthens the case for regulatory policies to promote safe working conditions… [and there is a basis for] assigning a higher priority to policies that target the conditions of the less-compensated.

The bottom line: Neither the anti-feminist, nor the conservative, assumptions about risk and pay hold water. The wage gap between men and women is not fair or justified; and the market is not fairly compensating those workers (mostly men) who face the highest risk of death or injury at their jobs.