Blinder's Blunder

Economist Alan Blinder wrote in The Washington Post that “offshoring of service jobs from rich countries such as the United States to poor countries such as India may pose major problems for tens of millions of American workers over the coming decades. In fact, I think offshoring may be the biggest political issue in economics for a generation.”
The telling phrase is “political issue.”

“To be marketable in the political arena,” Blinder once explained, “an idea must be short and snappy enough to be emblazoned on a T-shirt. But … any economic idea expressed that tersely is almost certainly wrong.”

“Offshoring” is short and snappy enough to attract nativist and protectionist interests, left and right. Blinder’s speculations were recently praised by William Greider in The Nation, for example, and by Phyllis Schlafly in The Conservative Voice.

“Nowadays,” says Blinder, “a growing list of services can be zapped across international borders electronically. … We’re all familiar with call centers, but electronic service delivery has already extended to computer programming, a variety of engineering services, accounting, security analysis and a lot else.”

Blinder is speculating about the next “decade or two,” yet the phenomenon he describes is going on “nowadays.” Infosys, India’s biggest software outsourcing firm, began in 1981 and now employs more than 72,000 worldwide, including employees in 16 U.S. offices. Wipro employs 66,000, including 11 offices in the United States. Perot Systems employs 5,000 of its 22,000 workers in India. If such offshore outsourcing of business services actually risked the loss of “tens of millions” of U.S. jobs, we should have seen some sign of it by now. But we haven’t.

In a paper for Princeton’s Center for Economic Policy Studies, Blinder selects 22 major occupations in services (out of 291 lesser jobs, which include manufacturing) ranked into four groups according to their “offshorability.” In contrast to his press interviews and popular articles, the academic paper makes it clear that he is “guesstimating the number of jobs that might be offshorable, not the number that actually would be offshored. … Only a fraction of these offshorable jobs will actually be offshored.”

 His purely subjective list makes it very clear, however, which service jobs are supposedly most threatened. And that, in turn, makes it possible to test his theories against actual job change between 2002 and 2005. I had to estimate the 2002 figure for “business operations specialists, all other” because that classification did not exist until 2004. Those offshorable jobs increased 8.2 percent between 2004 and 2005, but I assumed only a small increase from 2002 to 2004.

The total Blinder list of the 22 most offshorable major jobs totaled 15,732,670 in May 2005. That was up 7.7 percent from 14,603,140 jobs in May 2002. Employment increased in 18 of the 22 job groups, and wages increased in all of them.

 The two most vulnerable jobs on Blinder’s list are the most plausible, and the ones most often mentioned — computer programmers and telemarketers. They are followed in the same top tier by computer systems analysts, bookkeeping clerks, consumer software engineers and systems analysts.

The second, less risky category includes accountants, but also several jobs that do not seem easily “zapped across international borders electronically.” The group includes welders, for example, plus production worker helpers, packaging and filling machine operators, machinists, inspectors and even “first-line supervisors/managers.”

In the third group, we are asked to believe that “sales managers” may soon be relocated to foreign countries. Customers may get an e-mail or phone call, but don’t expect lunch or a handshake.

Employment increased between 2002 and 2005 in 18 of the 22 job categories supposedly threatened by electronic delivery of services. Employment did decline 4.5 percent among telemarketers, a loss of nearly 19,000 jobs, but that is an unpleasant low-wage job, and the job drop is largely explained by the national “do not call” registry. To put that loss of about 6,000 jobs a year in context, the United States routinely loses about 30 million jobs each year, yet gains even more. The net gain is why the unemployment rate is just 4.5 percent.

 Employment among computer programmers fell by 68,000 jobs, nearly 15 percent. I am not unsympathetic, since my son John is a PHP programmer of online applications (in New York, not India). Most software professionals in India are just coders, a rote job requiring little creativity. Besides, employment in four other computer-related “offshorable” U.S. jobs, such as computer software engineers, increased by 210,570 between 2002 and 2005, far outstripping the loss of mostly low-end programming jobs.

Blinder needed a short and snappy idea designed to be “marketable in the political arena.” He is trying to sell us on his public policy ideas. As the academic paper notes, “The appropriate policy responses (if any) to this problem probably depend on how many jobs might be susceptible to offshoring.” He wants a large policy response, which requires a large number.

It is always prudent to be suspicious whenever wildly sensational claims are used to market any public policy.

Blinder wants more generous unemployment benefits (which discourage work) and more costly retraining programs (which almost always fail). He even expects expert central planners to redesign “our education system so that it turns out more people who are trained for the jobs that will remain in the United States and fewer for the jobs that will migrate overseas.” If your children want to be computer programmers or accountants, they will simply have to adapt to the national plan.

Employment is increasing in nearly all the occupations Blinder imagines to be in imminent danger of being handled through electronic communication with some distant country. Unless the facts change, Blinder’s imaginative guesstimates and policy advice need not be taken too seriously.