Mike Konczal has a nice article about how important it is to distinguish “jobs” from “the economy”, and how Trump used that to his advantage. Noah Smith had a summary of the idea that a job is “more than a paycheck”. In my head, this is a complement to the arguments that Dani Rodrik has been making for years, as in his 1997 book Has Globalization Gone too Far?. In economist terms, the easiest way to say this is that having a job actually enters the utility function separately from the real wage it generates.

I have one idea to add to this, which is that even the “winners” of trade (or technological shocks) are unhappy about other people losing their jobs. If you like, the job status of others has an effect on the utility function of those with jobs. And it means they are willing to trade off some other source of utility (real wages) in order to give those people jobs.

As I’m not sure I’m explaining this terribly well, let me give you an example of a choice I posed to a number of “regular people” in my neighborhood, and during the after-game-beer-drinking phase of my hockey league. I’ve been doing this for a year or two, but recent events have made the responses more salient.

I’m not claiming that I had some sort of insight a few years ago. This all started because of a random conversation with Kenny, one of the guys in my hockey league, which led to him saying that he’d “absolutely” pay $100 more for a TV if it meant more American workers didn’t have to go on welfare. After that, I started posing the same kind of question to others in an ad hoc manner. I realize that this all has a very David Brooks-ish bullshit man-on-the-street vibe. Feel free to discount accordingly.

This is the essence of the choice I posed to people regarding a hypothetical trade deal (I typically didn’t use such precise numbers, but this gives you a concrete idea of what I was asking):

  1. Option A. With the trade deal, let’s say that the income of a winner - people who stay employed - is 1000. The price index with trade is 0.8. Let’s say there are four winners for every loser - the ones who lose jobs - but we organize a redistribution of the gains from trade to make sure the losers are compensated. We tax every winner 200, and use that to give every loser 800. The winners, after their tax of 200, are also making 800. With an price index of 0.8, every person - winners and losers - has a real wage of 800/0.8 = 1000.

  2. Option B. Don’t do the trade deal. Everyone is still employed, but their income is 700, which is lower than what a winner makes under the trade deal, and means that in aggregate, total income is lower than with the trade deal. Also, the price index is higher without trade, because we don’t get cheap imports. So let the price index be 1, and every person has a real wage of 700/1 = 700.

Which option would you choose? The pure economist answer is A, because obviously everyone is better off in terms of real consumption, with a real wage of 1000 versus 700. Noah’s post was about how job losers would probably prefer B. Most people would rather take the lower real wage of 700 with an actual job, versus a transfer with a real value of 1000. If you want to call this a form of money illusion, fine, but that doesn’t make it wrong. It is easy to say that their utility function puts more weight on income that is received from working versus income received as a hand out. It is not a stretch to get your head around this. I have spent years trying to instill this exact preference set into my kids.

What was more interesting in posing this question to various people I knew was the response of “winners”. Most of the time, I told people that they should imagine that they would be a “winner” under the deal. And for the most part, the people I was talking to were already winners with respect to trade, or were professionals relatively immune from the effects (e.g. doctors or lawyers). Regardless, even winners preferred B.

Why would potential winners prefer B? Anecdotally, they don’t like the idea of having to hand over some of their own wages to subsidize people who are not working. They know these people lost their jobs due to the trade deal or technological change, and a number of people I spoke with said they would be in favor of some kind of subsidy for the losers, but only for a very limited time. But it doesn’t seem fair to them that the job losers don’t have to do anything, even if it wasn’t the job losers fault in the first place. People care about the job status of others. This is, I would venture, some of the reason that individuals care about their own job status in the first place.

The hypothetical winners also tend to express a sort of endowment effect. Once you pay them 1000, it is much more tangible to them that they are giving up some of “their” money to subsidize job losers. On the other hand, under plan B they pay the subsidy implicitly through higher prices and lower nominal wages, and there is no endowment effect for them. And the people I asked this question are completely aware, like Kenny, of the fact that they are paying a subsidy through higher prices. The winners are for redistribution, but they care a lot about how that redistibution takes place. Again, this sounds like money illusion, but that doesn’t make it wrong.

I think the answers I got are in line with what Konczal, Smith, and Rodrik (and others) have said before about the value of jobs versus GDP. I’d suggest that even those with secure employment have preferences that would lead them to support jobs for others at the cost of some GDP.

Does this mean we should scrap every trade deal? No. The implication of my little study is that people are willing to trade off some of the gains from trade to keep more people employed. That doesn’t mean they are willing to trade off all of the gains from trade. If Option B were to scrap all trade deals and lower the real wage of everyone to 200, rather than 700, you would probably get a different answer from both potential winners and losers. And given the embedded nature of imports in supply chains, you could well make the case that the losses would be astronomical, and thus not worth it.

Other policies relating to job creation would face the same questions. Noah brought up the idea of incentives for long-term contracts, and noted the worry that you can take this too far. If you incentivize long-term contracts, do you gum up the labor market so much that new firms cannot afford to get stated or grow, and old ones cannot afford to shut down? Or do you end up getting the same issues we have now with trade; a class of “winners” that get long-term contracts and a class of “losers” that do not?

From the perspective of doing economics, I’m not sold that building the concept of this job preference into models would amount to much more than slapping some math around. I think our job is a little more humble. Be really clear about the economic costs and benefits of Options A and B on things like real wages and GDP that we have some hope of actually quantifying, and then let people decide if they are willing to pay those costs to save jobs or not.