The Skowronek Theory of Presidential Cycles

[Fmr. Gov. NickG]

Stephen Skowronek has a theory about presidential elections. He looks at the sequence of ‘political time’, the historical pattern of the American presidency that has repeated itself over the last 200 years. The sequence goes from “reconstructive” presidents who transform politics in their own image (Roosevelt, Reagan), followed by their handpicked successors (Truman, Bush ‘41) ; in turn they are usually succeeded by presidents Skowronek calls “pre-emptive”, who adopt the reigning orthodoxy of their parties (Eisenhower, Bill Clinton) followed by a faithful servant of that orthodoxy (Kennedy/Johnson, Bush ‘43) followed by another pre-emptive opposition leader (Nixon, Obama). The final stage is the “disjunctive” leader, who is outside their party’s orthodoxy, and that’s where we are now with Trump.  https://isps.yale.edu/news/blog/2016/12/skowronek-views-the-trump-win-through-political-time

It's an interesting idea, but I think Skowronek is off on his pairings.  I'd say
Hoover = Carter
FDR = Reagan/Bush '41
Truman = Clinton
Eisenhower = Bush '43
JFK/LBJ = Obama
Nixon/Ford = Trump/Pence

So, if you think Skoronek is right, Trump is due to lose in 2020.  If I'm right, Trump is due to win.  Anyway, political theories like this are fun fluff.  Any thoughts?

I teach this theory in my intro to American politics course every year (largely as a contrast to more quantitatively oriented political science).  I don’t understand how your pairings work.  How can Democrats be paired with Democrats some of the time and with Republicans other times?  The idea behind Skowronek’s theory is that at any point there is a dominant “regime” that gradually enervates, and the president can either be allied with the regime or opposed to it.  I’m not sure what the underlying idea to your pairings is.

[Fmr. Gov. NickG]

I don't believe it. Piecing together patterns from such limited data (especially if you're confined to the last 80 years) is difficult.

From a scientific perspective this theory (and most of these cyclical theories of Presidential elections) won't really be useful until it can predict the next 30 - 40 years of American politics.

To be fair, I believe Skowronek developed the theory early in the Reagan presidency, based on cases going back to Thomas Jefferson.  And it has predicted the subsequent 30-odd years pretty well.

[Fmr. Gov. NickG]

I tend to think of all these theories as bunk because they never go more than two generations back and break down as a result. Would Grover Cleveland really work as a disruptor to an era of GOP dominance that went back a quarter century, for example?

The thinking where politics seems to have started with FDR leads to a lot of myopia. It's this weird sort of Greatest Generation onward solipsism that history began with World War II/the Depression and there's nothing worth looking at further back, but that means you're dealing with a sample size of 13 presidents and 22 elections (assuming we're starting from 1932), and you cannot possibly make any serious trend argument with such a small sample size.

Skowronek’s theory tracks back way more than two generations.  His first article where he begins to develop the theory (published in the early ‘80s) is confined to the parellels between the 1824-1860 and 1928-1980 periods.  His book The Politics Presidents make, published in the 90’s, explains how almost every major president from Washington to Clinton fits into the pattern.