How concerned should we be about the downward trend in real wages?
Average hourly earnings have failed to keep up with inflation over the last two years, a fact noted with concern recently by
Deinonychus Antirrhopus, Economic Policy Institute, Sirotablog, and
Brad DeLong, among others.
Dave Altig notes that this is in fact a continuation of a trend going back to 1974 that had been temporarily reversed in the late 1990′s, as seen in the graph at the right. However, Dave also points out that although wages have fallen behind inflation for over a generation now, other nonwage components of worker compensation, particularly health care benefits, have grown more quickly than inflation. The graph below shows that in fact total compensation shows a steady long-term upward trend relative to inflation that has if anything accelerated in recent years.
Is it a good thing from the point of view of workers to be receiving an ever-shrinking fraction of their total compensation in the form of wages? One’s first thought might be that cash beats an equivalent dollar value in perks, since you could always use the cash to buy the perks yourself. You’d be at least as well off if you bought the identical perks as provided by your employer, and better off if you made some substitutions that better reflect your own preferences. But such an argument abstracts from taxes– since the perks come free of income, payroll, and sales taxes, you might be better off with a big set of employer-provided perks, even if there are substantial efficiency losses relative to what you could do with an equivalent amount of tax-free cash. To try to reduce these inefficiencies, some have advocated taxing employer-provided health care expenditures as if they were the same as direct worker income.
Even if there were no taxes, another reason why workers might prefer to receive income in the form of health care benefits rather than salary is the issue of adverse selection. If the insurance company cannot charge different premiums so as to fully reflect the different risks, healthy people end up paying more than they expect to get out, and so might not opt for the plan. The result is that the pool of insured is sicker, the premiums necessarily even higher, and many other people who might otherwise want insurance might get priced out. By insuring a pool of workers as a group through the employer, such adverse selection might be minimized, with the result that the average worker is better off than he would be trying to buy insurance on his own. This argument is persuasive to many economists, though Arnold Kling has his doubts.
But I worry whether other factors might be involved in the rising share of non-wage compensation. I wonder whether the people whose decisions determine this share really understand that the basic question is whether a given dollar of value-added that is produced by the worker goes to the employee in the form of salary or perks. I realize that I’m venturing out of economists’ usual rational-decision-maker framework in making such a claim. But let me give you two examples.
The first is based on observation of the political process. For example, one big component of nonsalary compensation is employers’ contribution to Social Security. Why did Congress set up a system in which, supposedly, the worker makes half the contribution and the employer makes a matching contribution? Economic theory suggests that it should make no difference who allegedly pays the payroll tax. The firm will look at the cost it pays for labor (which includes its Social Security contribution) in deciding how many workers it is profitable to hire, while the worker looks at what she personally gets paid (which excludes her Social Security contribution) in deciding on where and how much to work. The final outcome should be the same regardless of which party delivers the taxes to the government. And yet, the fact that the Social Security payment system is set up in the way it is suggests that there are at least some people who see the “employer contribution” and “employee contribution” as not coming out of one and the same pile of money.
The second reason I wonder about the efficiency of the size of current nonwage compensation is based on observation of the faculty benefits committee at UCSD. It seems there is a universal assumption of the noneconomists on this committee that the more benefits we get, the better off we are, as if the money were coming out of nowhere and had no implications for the budget for salaries. If professors act this way, I’m inclined to entertain the possibility that those who negotiate contracts for labor unions may have a similar mindset.
Finally, before we conclude that the solid trend in total compensation means there’s nothing to worry about in the declining trend in real wages, there is the difficult question of whether the growing non-wage compensation is acting to reduce or to exacerbate the rising gap between the wages of the top 10% and bottom 10% of earners; see for example the concerns raised by Kash at Angry Bear.
Should we worry about the declining trend in real wages? At least a little, I should think.