In a recent posting we reported on the highly commendable new Department of Labor regulations regarding H-1B workers and their wages.
Toward the end of it, we wrote that DoL was estimating that the annualized costs of the new program to employers (with the department using the bland term "transfers" rather than additional costs) would average $23.253 billion a year over the next 10 years. That's a lot of money, and might push some employers into thinking about hiring U.S. workers instead of foreign ones.
A reader — one of the American workers displaced by the H-1B program — replied, saying that this estimate was much too high. He then suggested that if 100,000 workers each year each got a raise of $20,000 the new annual costs would be $2 billion, not $23.253 billion. That inspired us to take a closer look at those estimated transfer payments.
The first thing we found was that after a couple of years the new and appropriately higher wage rates would cover virtually all H-1B workers, and DHS estimates that there are 585,420 of them, not 100,000 of them. And, as previously noted, I think the DHS population estimate is too low.
If you divide $23.253 billion by 585,000, you get an annual increase in costs of close to $40,000 a year per worker. That is still a huge number, and a somewhat misleading one, for it also includes two other major cost factors: fringe benefits and overhead.
The Department, on page 63872 of the Federal Register on October 10 stated that it is using a fringe benefit rate of 42 percent, and an overhead rate of 17 percent in its calculations. My sense is that one uses 100 percent of salary, applies the 42 percent, for a total of 142 percent, and then calculates an additional 17 percent (which adds 24 percent) to get 166 percent of wages. At least that was the formula when I wrote a number of contract proposals to the government in past years.
Working backward from the $40,000 or so in extra costs, and using the formula above, we get about $24,096 in wages, $10,120 in fringe benefits, and $5,760 in overhead, for a total of $39,966 or about $40,000.
Thus the average annual wage increase is about $24,000 for the workers, not about $40,000. And close to the $20,000 used by my informant.
The second thing we found is that the formulae for both fringe benefits and overhead seriously overstated the real costs of these factors to employers. The use of these formulae is both understandable (it is easy to use), and highly unfortunate because it exaggerates the cost of both factors, probably by billions of dollars.
You see, when you bring on high-wage workers the costs of fringe benefits do not rise in easy harmony with the rise in wages. Let's take just three examples:
- Medical insurance: Most private insurance costs are based on factors other than income; they relate to family structure and the age of the worker; these factors do not change when the worker gets a raise;
- FICA or Social Security taxes max out at $137,000 a year and many H-1Bs will be making more than that under the new scheme; and
- In New Jersey, where I worked with social insurance, the wage ceiling for unemployment insurance and the state's own temporary disability insurance program both max out at $35,300, and all H-1Bs make more than that. Thus, there would be no increase in fringe benefit costs for these workers after they got raises.
Similarly with corporate overhead, it is foolish to apply a standard formula in situations where workers' salaries are increased, even a lot. Does the rent go up when an H-1B gets a $24,000 a year raise? Does the cost of electricity and running water rise? Does the cost of supervision increase?
So while DoL has brought forth a fine new set of regulations, raising the incomes of the H-1Bs and perhaps leading to some more jobs for Americans, its total dollar cost estimates have been overstated, probably by billions a year, and that is damaging to the cause of H-1B reform.