The combustible mix of anxiety and uncertainty that fuels the offshoring debate has left policymakers and the public thirsting for hard data about jobs moving overseas. Specifically, Americans are concerned about how much bigger offshoring is going get and how it is going to affect their jobs. Although policymakers are scrambling to develop short-term policy fixes for the immediate problems associated with offshoring, we will need much better data to understand the broader scope of its impact and propose long-term policy solutions.
A host of public and private actors have recently released gross offshoring job-loss numbers. Mainstream estimates for gross job loss due to offshoring over the past three years range from 300,000 to over 1 million. Projections of the potential gross job loss in the next decade run from 3 million to 15 million. However, these numbers, although relevant, do not account for job creation (as a net job figure would) and are a fraction of the total number of jobs lost in the United States every year. More importantly, even the most knowledgeable analysts admit that these projections are merely rough estimates. None comes close to telling the full story of the impact of offshoring on our economy.
The most high-profile projection is the widely cited estimate by Forrester Research that 3.3 million jobs will move offshore by 2015. Almost immediately after its release, this projection became the go-to figure in policy circles and the popular press. But as the author of the report, Forrester Vice President John McCarthy, explained to the Wall Street Journal recently, his team's estimates were no more than "educated guesses" and were never meant to receive the kind of attribution they did. The public's intense desire to understand the scope of the problem has bred a reliance on statistics that even McCarthy admits are based heavily on guesswork.
Likewise, the misrepresentation of recent data from the Bureau of Labor Statistics' (BLS) quarterly Mass Layoff Survey highlights the downside of the public's new thirst for statistics. Normally, the release of the BLS Mass Layoff Survey is met with little fanfare outside the research community. It is well understood that the survey has a very specific, limited objective, looking only at instances where 50 or more workers have been laid off, and only at firms with only 50 or more employees. In fact, the 240,000 mass layoffs they recently reported represent only 3 percent of the 7.6 million gross jobs lost in each quarter of last year.
Yet the inclusion of a question about whether mass layoffs were associated with moving jobs overseas – and the finding that only about 2.5 percent of them were – sparked considerable attention. The Financial Times said the report would "quiet the debate" about offshoring in the United States and the New York Times added that "according to the first government effort to actually measure [offshoring] . . . fears may be overblown."
In reality, the Department's statistics provide only one small piece of the larger offshoring puzzle. Yet when a survey that represents only 3 percent of job loss in our economy elicits such assured responses, it means it's well past time to do some serious work to get a comprehensive, reliable set of numbers.
Accurately measuring job losses and gains due to offshoring is admittedly difficult, and a number of factors complicate the data gathering. To begin with, companies often do not collect offshoring statistics. When they do, different departments (accounting and human resources, for example) collect different data.
Second, there are a host of tricky questions surrounding what counts as an "offshored" job, particularly in a global economy characterized by multinational corporations. What if a German-owned company has an American subsidiary which contracts with an Indian call center for customer service? Are those jobs considered offshored or were they never part of the United States job market to begin with since they are encompassed within a German-owned firm?
Finally, offshoring occurs not only on the job-destruction side of the equation (a U.S. firm closing up operations and moving to another country), but also on the job-creation side (a U.S. firm deciding to create its next new marginal job abroad rather than at home). Our statistical agencies have little capacity to capture the counterfactual job-creation side of the equation.
Still, all of these formidable challenges shouldn't stop us from making a far more serious effort at measuring the effects of offshoring. While it is humorous to think of Forrester researchers doing cocktail napkin calculations that suddenly become the industry standard, the implications of imperfect and incomplete offshoring data is no laughing matter.
To begin with, it is clear that more resources are necessary. Congress should be sure that collecting this data is a priority. Appropriated money should go to the Bureau of Labor Statistics, Bureau of Economic Analysis and the U.S. Census Bureau – the agencies best able to collect offshoring statistics – so that they can improve their methodology and work together to begin capturing these numbers. In addition, Congress should establish an inter-agency task force to define problems, propose solutions and put together specific cost estimates on collecting more comprehensive data. Last week, the Brookings Institution convened an all-day seminar with economists, academics and experts from the government's statistical agencies to start the dialogue on the data challenges surrounding offshoring.
After a few months of solid job growth and in the absence of a serious discussion on data collection, some have argued that the offshoring debate can be put to rest. Unfortunately, this position is currently bolstered by statistics that do not tell the whole story. We risk an incomplete and ineffective policy debate – and a missed opportunity to take the serious steps necessary to address both the short- and long-term affects of offshoring – if we do not begin collecting accurate data.