The "Job Killers," Part 2

The American Association for the Advancement of Science was kind enough to send me a free issue of Science last month, in the hope that I'd subscribe. I didn't—too much “hard” science, in both senses of the word—but I found an interesting evaluation research study by economists David Levine, Michael Toffel and Matthew Johnson on the effect of government safety inspections on worker injuries. Unfortunately, the study is gated.

As noted in Part 1 of this post, corporate spokespersons and conservative politicians often label government-enforced health and safety regulations—along with environmental protections, health insurance, and taxes—as “job killers.” Opponents claim that these laws seldom actually prevent injuries, and their implementation is so expensive that they make companies unprofitable, leading to plant closures and layoffs, and hurting workers in the long run.

There is a large existing literature on the subject, but it is inconsistent and inconclusive. Studies usually look for changes in injury rates from before to after an inspection. But workplaces are not randomly chosen for inspection; they are usually chosen due to recent accidents or safety complaints. If some of these accidents are actually random events, there could be an apparent improvement that is merely regression to the mean. In addition, inspectors often find incomplete safety records and require better record keeping. This can lead to the false perception of an increase in accidents after the inspection.

Fortunately for research purposes, between 1996 and 2006, California's Occupational Safety and Health Administration, lacking the resources to inspect all workplaces, randomly selected 409 workplaces from high injury industries for inspection. The authors matched them with 409 control firms, each one from the same industry and region of California as an inspected plant. They counted the number of injuries and the cost of these injuries during the period from four years before the inspection to four years after. They also looked at several measures intended to determine whether inspections had the unintended effect of being job killers: firm survival, credit ratings, total sales, number of employees, and payroll.

They found that the inspections reduced injuries by an average of 9.4% in comparison to the control group—not a trivial difference, especially if you work at one of these plants. Workers' compensation costs were reduced by 26%. The difference was significant for both major injuries (costing more than $2000) and minor ones. The improvement was not just temporary. The reduction in injuries was greater in the third and fourth years after the inspection than in the first two years.

There was no evidence of job loss or any other negative effect of the inspections on the competitiveness of these firms. The plants selected for inspection were no more likely to go out of business, and they did not have lower credit ratings, sales, numbers of employees, or payrolls, when compared to the control plants. In fact, all five of these measures favored the treatment group over the control group, although none of the differences were statistically significant.

This is just one study, but it appears to be a good one. It's worth bearing in mind the next time some lobbyist or politician claims that health and safety laws, and their enforcement, are “job killers.”

The "Job Killers," Part 1

Students of persuasion from Aristotle to Adolf Hitler have advised propagandists that statements can acquire the illusion of truth through repetition. Let's suppose you hear a plausible statement, but you are unsure whether it is true, i.e., “Greenland has about 50,000 inhabitants” or “Zachary Taylor was the first president to die in office.” Research in social psychology has shown that the more frequently you are exposed to such statements, the more likely you are to judge them to be true. A recent statistical analysis of 51 studies found this “illusion of truth effect” to be highly reliable. The effect is largely unconscious. People judge familiar statements to be true even when they can't remember whether they've heard them before.

Here's a statement I'll bet you've heard quite often in the past few years: “Government regulations are job killers.” Peter Dreier and Christopher Martin, Professors of Political Science and Communication respectively, did a content analysis in which they looked at every instance of the phrase “job killer” in the Associated Press, New York Times, Wall Street Journal and Washington Post from 1984 to 2011. There were 381 articles that contained the phrase “job killer” and its variations. (If you think that's not very many, remember that stories from these four organizations are widely reprinted. The Associated Press alone serves 1700 newspapers and 500 radio and TV stations. The authors cite one article that received 12,800 citations.)

The first thing they noted is that use of this phrase has become more frequent in recent years. In fact, it increased dramatically after Barack Obama took office. From the first three years of the George W. Bush administration to the first three years of the Obama administration, appearances of “job killer” increased by 1156%; that is, there were about 11.5 times as many “job killer” stories.

Most of the accusations focused on federal (65%) and state (12%) government policies designed to regulate business. The top four policies claimed to be job killers were environmental regulations (18%), tax policies (17%), health care reform (10%), and wage policies (8%), usually proposals to increase the minimum wage.

Not surprisingly, the most common sources in stories claiming that something was a job killer were Republican politicians (42%) and business spokespersons (19%). But the third most frequent source was the authors themselves. 17% of the time, the newspaper used the phrase in articles or editorials without attributing it to any source, suggesting that the claim that government regulations are job killers is part of the conventional wisdom. The Wall Street Journal was more likely than the other organizations to use the phrase without attribution.

91.6% of the articles alleging that a government policy was or would be a job killer did not present any evidence to support that claim, either from the source or from the article's author. In the 8.4% of cases where evidence was presented, the most frequent types reported were anecdotes (3.4%), quantitative data (1.6%), a study (1%), and the opinion of an economist (.8%). The authors did not find a single case in which journalists attempted on their own to investigate the veracity of a job-killing claim.

In cases where a source claimed that a policy was a job killer, reporters provided a counterclaim from an opposition source only 7% of the time. I find this especially noteworthy. Even when claims have broad empirical support, such as the evidence for global warming, reporters usually engage in false balancing by citing the opinions of a small minority of global warming skeptics. However, claims of job-killing seem to be exempt from the norm of balance which requires journalists to present both sides of the issue. Are other conservative statements similarly exempt? Does labeling a statement “controversial” only apply to claims made by liberals?

Concerns about unemployment are higher during economic hard times, so you might expect articles about job killers to increase during a recession. However, there was no correlation. The only historical variable that predicted job-killing claims was the occupant of the White House. They were up during Democrat administrations and down when Republicans were in power.

Dreier and Martin cite journalism textbooks and the Society of Professional Journalists' Code of Ethics, which state that it is a reporter's job to verify what newsmakers say. This advice is apparently now archaic. To report that a politician's or businessman's statement is true or false risks being accused of bias, especially at this historical moment, when the truth appears to have a liberal bias.

Failure of reporters to investigate the claims of politicians has the obvious consequence of making it more difficult for the public to determine whether their statements are true. It also means that politicians pay no penalty for lying, since they are never called out for it. This would seem to encourage further falsehoods. Perlstein has argued American politics has become a fact-free zone where lying has become the new normal.

It's obvious that the job killer meme—along with the equally suspect claim that the wealthy are “job creators”—has become a powerful weapon in the hands of Elephant politicians and their allies in the corporate media. Few journalists seemed concerned when the House of Representatives passed an attempt to kill the Affordable Care Act (ACA) entitled “Repealing the Job-Killing Health Care Act.” In fact, an analysis by the nonpartisan Congressional Budget Office predicted that the ACA would not lead to significant layoffs, only voluntary retirements. Politifact rated House majority leader Eric Cantor's statement that Obamacare was a job killer as “false” for that reason.

To what extent do government policies protecting workers, consumers or the environment actually lead to job loss? It's hard to say, since there hasn't been a lot of good research. One policy that has been extensively studied is the minimum wage, where research shows that increasing the minimum wage does not increase unemployment. For another example, see Part 2 of this post.