Mitchell’s Musings 10-10-16: Makes Sense to Me
Daniel J.B. Mitchell
I came across an article in the Los Angeles Times business section recently with the print-version headline “Trade is seen as harmful.” The article noted that a Pew Research Center poll found that “eight out of ten adults regarded outsourcing of jobs overseas and the growth of imports of foreign-made goods as harmful to U.S. workers. By comparison only half of the people surveyed saw automation as hurtful – even though many economists believe that new technologies and the mechanization of work have led to as many job losses as imbalanced trade.” The article went on to relate the trade concerns to the current presidential election campaign.
Above are the actual poll results from the underlying Pew study. Although the LA Times article seems to imply that respondents have misdiagnosed the relative importance of trade vs. automation, there may be an explanation. Unfortunately, Pew does not provide the precise wording of the questions asked. So exactly how respondents interpreted words such as “outsourcing” and “automation” is not clear. Were they given definitions? The hurt vs. help contrast also raises some questions. Do the words mean destroy jobs vs. create jobs? Or do the terms suggest making jobs more difficult to do vs. assisting workers to do their jobs? Or do the terms suggest some kind of good jobs vs. bad jobs distinction? The LA Times article implicitly assumes the destruction/creation interpretation.
Despite the ambiguities, let’s limit the interpretation to destruction vs. creation when thinking about the trade issue and how survey respondents reacted. Other things equal, the large trade imbalance (deficit) that has characterized the U.S. for decades has to be a source of net destruction. There are caveats, of course. Important among them is the fact that workers whose jobs are lost may end up in the non-trade sector (such as retail). That is, imbalanced trade may shift the mix of jobs towards the non-trade sector without changing the total number of jobs.
What about automation as a concern? Note that the LA Times article seems to use automation and technology interchangeably. If that is also how respondents reacted, at least some of them may have been thinking about the way technology provides an assist to workers in doing jobs. Thanks to computer technology, for example, it is easier to access information needed on the job than it used to be. Medical records are now available readily without going through paper files. There are many such examples.
Finally, what about immigration? Respondents are roughly split on whether immigration hurts or helps American workers. The LA Times article notes that ten years ago, the result was much more anti-immigrant than it is now. Although the pace of illegal immigration is hard to measure precisely, another report from Pew suggests that during the past decade, net illegal immigration has halted. That is, the absolute of number of illegal immigrants residing within the U.S. has stayed about constant, as shown on the chart below. Presumably, the Great Recession and its aftermath had an impact in discouraging a net inflow; up until the Great Recession occurred, the number had been rising. That shift to a net of zero may explain the attitudinal change.
In short, the Pew survey results do not seem to be counter-intuitive. They seem to go with the election narrative. Reflected in the outcome, you have the followers of Bernie Sanders who think trade is a problem but not immigration vs. the followers of Donald Trump who think both are a problem. You have automation-technology seen as good and bad, perhaps because the question is posed ambiguously. It makes sense to me.
 The online version of the article had a different headline focused on immigration, not trade: “Americans are feeling better about immigrants' economic effect — but Republicans aren't, survey shows.” See http://www.latimes.com/business/la-fi-pew-study-jobs-20161006-snap-story.html.
 The full report is at http://assets.pewresearch.org/wp-content/uploads/sites/3/2016/10/ST_2016.10.06_Future-of-Work_FINAL4.pdf. A summary is at http://www.pewsocialtrends.org/2016/10/06/the-state-of-american-jobs/.
 The fact that decline of unions is seen as more hurtful than helpful in the survey suggests that respondents may have been thinking – at least in part – about job characteristics such as pay and benefits.
 Exports and imports may differ in their labor-using characteristics. At least in theory, however, U.S. imports should be more labor-intensive than exports which would intensify the net destruction effect. That is, even with balanced trade, there might be net displacement.
 http://www.pewresearch.org/fact-tank/2016/09/20/5-facts-about-illegal-immigration-in-the-u-s/. Even when the stock of illegal immigrants is constant, there may be gross flows into and out of the U.S. But the inflows and outflows must be balanced, i.e., a net of zero, for the total stock to stay the same.
 Immigration has had a U-shape in terms of skill with a concentration of unskilled immigrants and a lesser concentration of highly-skilled immigrants. Since this pattern is not a replica of the existing U.S. workforce, there may be both competition among substitutes (e.g., low-skilled natives vs. low-skilled immigrants) as well as complements, e.g., natives for whom demand for their labor is enhanced by the presence of immigrants. For example, native supervisors in southern poultry packing plants may benefit from the influx of low-skilled production workers. It seems unlikely, however, that this nuanced view of the impact of immigration across different groups within the workforce is driving the survey results.