Election Promises, Trade Policy Ripples: Behind New PIIE Study

September 20, 2016 11:45 AM
Photo Credit: 
REUTERS/Srdjan Zivulovic

Methods matter in economic research, so it's worth taking the time to explain the work behind a preelection study from the Peterson Institute for International Economics (PIIE) on the potential economic effects of the two presidential candidates’ proposed trade policies. 

Economists at PIIE figured out a way to gauge local effects from national trade policies by taking a more granular look at a macroeconomic dataset generated by Moody's Analytics. The Moody's model is a large, traditional, multiequation econometrically estimated model that looks at an economy’s most basic components: gross domestic product (GDP), employment, private consumption, government spending, investment, exports, and imports. It also includes a variety of financial variables. Importantly, the model allows for the possibility of unemployment from a given macroeconomic shock. It views these employment effects as transitory but fairly long-lasting—the transition back to full employment can take over five years in their model. The weak performance of the US economy since the financial crisis indicates that the effects of such shocks could last up to a decade, absent adequate policy response.

Extending the Moody’s model, the study’s authors Marcus Noland, Sherman Robinson, and Tyler Moran use a separate, detailed, industry-level model to disentangle the macroeconomic results by both sector and geography. They start with a table of intermediate goods demanded by producers—like steel demanded by auto producers—that includes 389 “industries,” which they organize into a social accounting matrix (SAM). The US SAM dataset is used to trace the direct and indirect impact of changes in the macroeconomy on detailed industry production and employment at the national level. The national employment changes are further broken down into changes in employment at the state and county levels in a slightly aggregated set of 369 industries. This employment disaggregation process draws on private employment data from the Quarterly Census of Employment and Wages (QCEW) of the Bureau of Labor Statistics (BLS) for 2013, which provides county-level data on employment.

Changing one variable in the SAM model, such as a reduction in exports, will show the ripple effects on the rest of the economy in the form of changes in demand for intermediate goods, imports, and incomes. Reducing trade sharply will lead to severe, direct job losses in trade-related industries, which will then lead to indirect job losses in industries that produce intermediate goods for the affected industries. This in turn will lower US incomes, consumption, and investment, throwing the economy into a recession. As incomes and consumption drop, retail and wholesale purchases will fall as people buy fewer goods.

 A contribution of the PIIE study's approach is that it allows separate but linked models to connect macro shocks to their effects on specific industries operating in states and even counties. Many US workers’ jobs are linked indirectly to international trade and the performance of US exports, whether they know it or not.  

The authors model three plausible retaliatory scenarios under Trump’s proposals to slap punitive tariffs against imports from China and Mexico:

  • Full trade war: The United States imposes a 45 percent tariff on nonoil imports from China and a 35 percent tariff on nonoil imports from Mexico. (Under this scenario, China and Mexico respond reciprocally, imposing the same tariffs on US exports.)
  • Aborted trade war: US tariffs are imposed for only one year, because China and Mexico concede to US demands, the US Congress overturns the action, President Trump loses in the courts, or the public outcry is such that the administration is forced to stand down.
  • Asymmetric trade war: China and Mexico do not reciprocate with an across-the-board tariff but instead retaliate in other, more industry-specific ways.

The methodology described above was used to calculate the effects of the first two scenarios, a full-fledged trade war and an aborted trade war. Thousands of counties are affected throughout the country and hundreds of industries that are completely unrelated to trade are hit when China and Mexico retaliate in kind under a full trade war scenario.

"We're taking a snapshot of what happens when you hit the economy with a big hammer," says Noland, PIIE's executive vice president and director of studies, referring to the full trade war scenario. "In our case, we take that snapshot in 2019 at the trough of a policy-induced recession."

The asymmetric trade war scenario looks at China and Mexico reducing imports of US business services and ending imports of US aircraft and soybeans. The calculations for these simulations were done by reducing US business services exports by 40 percent (the share estimated to be purchased by Chinese state-owned enterprises) and terminating all aircraft and soybean exports to China. The ripple effects on the entire economy, while small, are still severe on counties that produce the affected commodities.

Readers wishing to find out the study’s estimates of the economic impact on specific industrial sectors and on employment in a specific state or country may do so by using this searchable database

Add new comment