electoral downside of high unemployment

I don't like this meme DeLong and Krugman are pushing about high unemployment not mattering in elections.

They say political scientists like John Sides say it's a political fact. Well politics is even less of a science than economics. It could be true, but I believe that social democrat programs like Social Security and Medicare are popular.

So how does it make sense that the social democrat program of full employment with the economy running at capacity unpopular and electorally irrelevant. They say it matters if the unemployment rate goes up but not if it is high. My view is that they're making a partisan argument to keep the Republicans out of power no matter what. DeLong was arguing against Republicans who said Obama wouldn't get re-elected because of the economy. Krugman says Hillary will get elected despite the Democrats poor perfomance (even though it's mostly because of the blocking Republican minority.) However I see it as them saying the social democrat program of proper demand management (monetary, fiscal and currency policies) of full employment and an economy running at capacity is not popular. Even though Medicare and Social Security are popular.

Here's Krugman echoing DeLong and Sides.  First from his blog he writes he believes Hillary will win. Then in his column he echoes DeLong.

Not Like Ike

That will, one hopes, eventually change — but it’s going to take several big electoral defeats, and it’s not going to happen by 2016. If she becomes president, which does look fairly likely, Clinton will almost surely face the same environment Obama has faced all along — a completely obstructionist, hate-filled opposition. The only thing that might change this would be if her victory is really shocking — say, Democrats retake both houses of Congress and Clinton herself carries Texas. 

Defining Prosperity Down

You might think that a persistently poor economy — an economy in which millions of people who could and should be productively employed are jobless, and in many cases have been without work for a very long time — would eventually spark public outrage. But the political science evidence on economics and elections is unambiguous: what matters is the rate of change, not the level. 

Put it this way: If unemployment rises from 6 to 7 percent during an election year, the incumbent will probably lose. But if it stays flat at 8 percent through the incumbent’s whole term, he or she will probably be returned to power. And this means that there’s remarkably little political pressure to end our continuing, if low-grade, depression.

The evidence is unambiguous? I doubt that even though I'd much rather have Hillary than a Republican as President. I read DeLong and Krugman every other day and agree with their goals. I just think they may be wrong here. His link in support is:

Are Elections Too Much Like Musical Chairs? by Larry Bartels