Made-up job gains from the Trans-Pacific Partnership

The Fact Checker frequently warns readers to be wary of claims by politicians that various policy initiatives will yield tens of thousands of jobs. Such claims are often based on studies that rely on a variety of assumptions, any of which can be called into question. So we were interested when we received a call from a reader who wondered how the administration calculated that a proposed international trade agreement, known as the Trans-Pacific Partnership, would support some 650,000 jobs.

The Obama administration is on a full-court press to complete negotiations on the trade pact, involving 12 Pacific Rim nations, and also win from Congress the authority for an up-or-down vote from lawmakers, which officials say is necessary to close the deal. The TPP is one of the few areas in which the White House can expect strong support from Republicans, though many Democrats remain skeptical or hostile.

The Fact Checker of course takes no position on whether the proposed trade deal is good or bad. But we were curious about how this number was calculated.

The Facts

Notice that Vilsack referred to the Peterson Institute for International Economics, which is a well-regarded centrist think tank that focuses on international economic policy. The Peterson Institute advocates for free-trade agreements but also for programs that aid people who may be hurt by globalization.

The Peterson Institute in 2012 published a book titled “The Trans-Pacific Partnership and Asia-Pacific Integration: A Quantitative Assessment,” by Peter A. Petri, Michael G. Plummer and Fan Zhai. The book does include an estimate that, by 2025, the United States would experience a gain of $77 billion in income from TPP, as well as a $124 billion increase in exports. (More on those numbers, which are expressed in 2007 dollars, below.) But nowhere in the book does it says 650,000 jobs would be created.

Asked about the statistic on 650,000 jobs, the White House referred us to the Office of the U.S. Trade Representative. USTR spokesman Matthew McAlvanah directed us to page 58 of the book. “They do not provide an estimate on jobs,” he acknowledged. “However they do provide a methodology that one could use.”

Essentially, the book suggests that an income gain of $121,000 would be “roughly equivalent to creating an extra job.” So the Obama administration took the figure of $77 billion and divided it by $121,000, which yields 636,000. Rounded up, that becomes 650,000.

There’s just one problem: Petri says this is the incorrect way to use his research, especially when officials such as Kerry combine the jobs figure in the same sentence as the income prediction: “The TPP could provide $77 billion a year in real income and support 650,000 new jobs in the US alone.”

That’s because the calculation on jobs can only be done if one assumes that wages have been frozen and there is no income gain, Petri said. So it’s completely misleading to suggest there would be both a gain in income and a gain in jobs.

Petri said that his book did not discuss job gains because mainstream economists do not believe that the number of jobs is significantly affected by trade policy.

“The reason we don’t project employment is that, like most trade economists, we don’t believe that trade agreements change the labor force in the long run. The consequential factors are demography, immigration, retirement benefits, etc.,” he said. “Rather, trade agreements affect how people are employed, and ideally substitute more productive jobs for less productive ones and thus raise real incomes.”

The same dynamic exists with the claim that the trade bill would increase exports by $124 billion by 2025. The Commerce Department estimates that about 5,500 jobs are supported by every $1 billion in exports, so in theory that also would yield about 650,000 jobs. But that calculation would ignore the fact that the Petri book found that imports would increase by virtually the same amount as exports, meaning the net number of new jobs is zero.

The mix of jobs would change, however. “Employment could be negatively affected by the adjustment implications of a trade agreement,” Petri said. “We estimate ‘job shifts’—employment moving from one sector to another—and in difficult labor markets such shifts can lead to transitional unemployment, retirement or wage cuts.”  But, he added, “in the case of the TPP such shifts will be small and slow, dwarfed by routine job separations and new hires in the economy. So adjustments and costs should be covered many times by gains. This makes possible strong transitional assistance for workers and communities that are adversely affected.”

Finally, let’s put these numbers in context. Petri’s book says that a gain of $77 billion in income amounts to just a 0.4 percent increase in the pre-trade-deal baseline for the United States’ $20-trillion gross domestic product. You read that right—0.4 percent.

Indeed, a gain of 650,000 jobs would also be just 0.4 percent of projected employment of 168 million people, Petri said. “The percentage change is small,” Petri acknowledged, which he said is what one would expect from a large and efficient economy such as the United States. (Vietnam, by contrast, would see a gain of nearly 14 percent in income according to Petri’s model.)

There is, of course, a long history of presidential administrations touting imaginary job gains from trade deals. “I believe that NAFTA will create 200,000 American jobs in the first two years of its effect,” then-President Bill Clinton said in 1993, when he signed supplemental agreements to the North American Free Trade Agreement. “I believe that NAFTA will create a million jobs in the first 5 years of its impact.”

Clinton was relying in part on analyses generated by the Peterson Institute. Two years later, after a financial meltdown in Mexico and collapse of the peso evaporated any job gains from NAFTA, the economist who generated the forecasts famously said he would stay away from job forecasting in the future.

In 2012, C. Fred Bergsten, the founder of the Peterson Institute, conceded the Institute had “big internal debates” over whether to calculate job numbers. “Congress always wants to know how many jobs they’re going to create,” Bergsten said. “As good economists, we all take the view I think that trade agreement does not unbalance, create, or destroy jobs, it alters the composition of the workforce.”

McAlvanah provided the following response:

“The Peterson Institute study provides a variety of different analytical pathways to estimate the job supporting potential of TPP.  Peterson provides estimates of both the increase to U.S. income and to U.S. exports as a result of TPP.  The methodology laid out in the Peterson study for calculating potential employment gains based on their projected increase in income suggests a jobs number just below 650,000.  A different approach is to use the sectoral value added estimates from the Peterson study.  This approach was used by the U.S. Chamber of Commerce to estimate approximately 700,000 jobs.  Another approach is to map the incremental increase in national economic output to a proportional increase in employment.  The Peterson study estimates an increase of 0.4 percent of GDP attributable to TPP.  Based on a 2025 working population of 168 million cited by Peterson, a 0.4 percent increase would be approximately 672,000 jobs.  It is also possible to estimate jobs supported by new exports by multiplying estimates of jobs per billion exports by the 124 billion dollars of increased exports projected in the Peterson study.”

We note his statement did not address the fact that the study itself did not offer these calculations — or that effects of one action, such as more exports, are canceled out by another action, such as imports.

The Pinocchio Test

Clearly, with the Peterson Institute refusing to play the game this time and cough up a jobs number, the administration decided to concoct its own. But, as we have shown, one cannot at the same time claim both a gain of $77 billion in income and a gain of 650,000 jobs; the same effects simply cannot happen at the same time.

Moreover, these are big numbers with virtually no context. It is pretty lame to use such huge numbers to tout what, in the context of the U.S. economy, amounts to minuscule changes in income —10 years from now.

Our advice remains: be wary whenever a politician claims a policy will yield bountiful jobs. In this case, the correct number is zero, not 650,000, according to the very study used to calculate this number. Administration officials earn Four Pinocchios for their fishy math.

Four Pinocchios