As argued in numerous posts here, we have seen an apparently emerging disconnect between economic conditions and political outcomes in a variety of nations, with anti-immigrant or more generally nationalist or populist parties with authoritarian tendencies gaining strength in many nations despite apparently improving or even largely pretty good economic conditions. A list of those showing this includes Poland, the US, Germany, UK with the Brexit vote, Austria, Sweden, the Netherlands, and some others (I had Iran on the list, but it seems to a more complicated case). I labeled this phenomenon to be the "Poland problem," with the Law and Justice Party coming to power in 2015, despite Poland having been probably the best economically performing of all the European transition economies, as well as being the only European nation not to having gone into recession in 2009.
Now a study by Yann Algan, Sergei Guriev, Elias Papaionnou, and Eugenia Passari at Pro Market (and linked to today by Mark Thoma at Economistsview) finds that if one looks closely at local economic conditions in parts of European nations over recent years, specifically changes in unemployment rates, one finds that there seems to be a relationship between such increases and increased support for "populist" parties, a "specter hanging over Europe." I find the study reasonably persuasive. So this would mean that a possible explanation for the Poland problem is that unhappiness in suffering and poorer regions overwhelms broader good economic performance in nations, leading in some cases to actual takeovers by these parties.
I note that for at least some nations where I know the geographic details better, this fits with the odd situation where those regions voting most vigorously for strongly anti-immigrant parties are also th regions with fewest immigrants. This reminds me of the old wisecrack about Poland from 1968 when there was a major outbreak of anti-Semitism that Poland was showing how to have anti-Semitism while having few Jews.
We should note that while this local outcome finding reasserts the standard link between economic conditions and political outcomes, the Poland problem does not completely go away. The hard fact is that while indeed we see worsened economic conditions leading to voting for political change, we see these cases where in fact conditions in the nations as a whole are/were pretty good. But those good conditions in other parts of these nations fail to deliver satisfaction sufficient to save the incumbent parties. We have seen such local behavior previously without this sort of outcome happening at the national level.
I shall add two further notes on the case of the US, some of which I already made in my earlier post on the US having a Poland problem. One is that in fact looking at the popular vote, the better national conditions did offset the locally bad ones in the Rust Belt in that in fact Clinton won the popular vote fairly solidly over Trump. But the shifts in those Rust Belt states crucially shifted electoral votes so that Trump won the electoral vote and became president.
There is also a more complicating aspect to this. Without doubt parts of the states that shifted between 2012 and 2016 in the Rust Belt, IA, WI, MI, OH, and PA, fit the ProMarket story. : especially in PA and OH. We see locations such as Scranton-Wilkes Barre and Erie in PA having bad economic conditions and switching sharply from Dem to GOP between 2012 and 2016. In OH Youngstown also fits this as well as some manufacturing urban areas in MI. But important lrgions in these states did not fit the story. In particular reasonably well-off rural areas in parts of OH, MI, WI, and IA, with these areas the key for the switch in state outcomes in the latter two, especially southwestern Wisconsin and southern Iowa, the state with the largest percentage change in vote. These changes were not driven by bad economic conditions. Some of these, especially in SW WI, switched just prior to the election, with the infamous Comey email story likely playing a major role.
Of course today we have Trump facing a Poland problem: US economy clearly growing solidly, with well over 50% of voters saying so. But Trump's support is a good 20% lower. I have read that Trump and some of his associates quoting James Carville on "It's the economy, stupid," hoping that a big advertising campaign pushing the supposedly stimulative effect of the GOP tax cut. It may well be that they are not going to succeed, and Trump and GOP will fall victim to the US version of the Poland program as the GOP takes a large hit this fall in a Dem wave, despite the economy doing well.