Partisan Economic Patterns

Philippa Dunne and Doug Henwood:

The Liscio Report: Presidential economics: Do parties matter?: this is a good time to look at the partisan patterns in some major economic and financial indicators. The differences are significant, and worth thinking about for anyone with dollars at stake after January 20, 2009:

These partisan patterns are immensely strong. Yet, as Paul Krugman points out, it is very hard to point to any concrete policy steps taken by administrations that are strong and directed enough to produce them. Yet there they are.

Comments

[eightnine2718281828mu5e]

http://rodrik.typepad.com/dani_rodriks_weblog/2008/03/american-politi.html

---
It comes from Princeton political scientist Larry Bartels' new book, soon to be released. What it shows is the difference that the President's party affiliation makes to the distribution of income during the four years of the president's term.

(The distributional outcomes are shown with one year's lag.) When a Republican president is in power, people at the top of the income distribution experience much larger real income gains than those at the bottom--a difference of 1.5 percent per year going from the bottom to the top quintile in the income distribution.

The situation is reversed when a Democrat is in power: those who benefit the most are the lower income groups. If you are in the bottom quintile, the difference between having a Democratic or a Republican president in office is an income gain (or loss) of more than 2 percent per year!

Strikingly, compared to Republicans, Democratic presidents generate higher income gains for all income groups (although the difference is statistically significant only for lower income groups).
---

[eightnine2718281828mu5e]

Nice chart at the rodrik link, btw.

[David]

Inequality increased more under Clinton than it did under Bush Jr? Or am I reading the 5th chart incorrectly?

[derek]

Inequality is the rich reaping the benefits of increased GDP while leaving the poor little better off. What you're seeing is how little GDP growth there is under Bush 43 for the rich to reap.

And yeah, Clinton was pretty Republican for a Democrat. The right wing hysteria about him is more a matter of "working the referee", i.e. always insisting the other side are dirty commies no matter how much they try to please you.

[James Wimberley]

.."[I}t is very hard to point to any concrete policy steps taken b administrations that are strong and directed enough to produce [these effects]".
How about some implicit barriers theory, as with monopolies? The monopolist's rent-seeking is mainly constrained by the risk that a new competitor may enter the market. Under a Democratic administration, capitalists know that the government might do something if they go too far in rent-seeking. Under a Republican administration, they know this won't happen. Similarly workers know that a Dem administration might back them, and a GOP one won't.
The argument only helps on the distributional side. I don't see how it explains the growth differential.

[MB]

It seems more likely causality may run from economic conditions to who gets elected. High inflation leads to republicans and also higher real rates and recession. Recession and unemployment gets democrats elected and they govern through rebounds. Effects of policy would expect to have long lags so credit can't be measured by returns in that administration.

[eCurrency Arbitrage]

Its a known thing, when a Democrat is in administration its first benefit the lower income people and than only others. If we satisfy them its great democrat.

[save_the_rustbelt]

Roosevelt III, Roosevelt-Truman and Truman were influenced by WWII and then the reconstruction of the economy. Given the circumstances both deserve credit for managing through this period, especially Truman in the post-war period. Ike also benefited from Truman's relatively smooth transition.

[bakho]

Some of these are positively correlated, such as GDP and employment growth.

Under Clinton, the bottom quartile kept pace with the gains of the very top, but the very top did very well in the 1990s. The tech boom made a lot of new millionaires.

[Robert Waldmann]

there is a booboo. Roosevelt-Truman employment growth is shaded red making the employment growth graph even more extreme looking than reality.

I am struck by the similarity of GDP growth under Clinton, Carter and Reagan. Shows the importance of timing recessions at the beginning not the end.

There is a hint of an explanation in the real bond yields, CPI and fiscal stance graphs. Republicans have (slightly) declining inflation and huge increases in the deficit causing high real interest rates (partly changes in inflation are correlated with unexpected changes in inflation partly loose fiscal tight monetary implies high expected real interest rates). I'd also look at real exchange rates. I think the story is crowding out leads to poor growth of GNP and employment not next quarter but within a presidency.

[skeptonomist]

"These partisan patterns are immensely strong."

But the time patterns are even stronger:

http://www.skeptometrics.org/IncomeGrowth.htm

This directly applies to Bartels' analysis of income inequality, but other macroeconomic variables will mostly look the same, as they have the same dependence on time.

The patterns do not indicate that presidents can turn economic trends on and off as with a switch, as most people assume. There are differences between presidents and parties all right, but there are more profound differences between economic policies adopted by the governing authorities of whichever party, over time.

Focussing on the party of the president is bad science, will likely lead to dead ends, and could result in nasty surprises in rebuttal from conservatives who are smart enough to figure out how fallacious the argument is.

More profitable avenues of research on the time trends might include the prevalence of monetary versus fiscal policy; the ideological and partisan leanings of monetary authorities; changes in tax policy; changes in international trade; the economic position of the US relative to other countries; and many other things which have changed over the period in question.

[John Salvatier]

Doesn't that suggest (weakly) that the presidents are the result not the cause of those changes?

[Jason]

I make video games. I once did the difficulty-tuning on a football game. We wanted to be able to adjust the strength of the game to let newer players win their first few games, and to make more experienced players sweat. Close games are always more fun, with an important caveat:

If the player can see where you're making the adjustments, they resent it and it's not fun. If they can't see your hands, they just sense the game getting harder, and feel more pressured, and have more fun.

So, how to tweak the difficulty without doing anything so obvious that the player can see it?

You might adjust the probability of a fumble when the ball carrier takes a hit. But if that's the only knob you're turning, you're going to have to turn it a lot to get the result you're after. Everyone will see it. When they're ahead they'll be coughing up the ball left and right.

There are, however, hundreds of knobs. The game action moves in "frames" of 1/60 of a second. In every one of those frames there are several interactions governed by chance. Blockers struggle with rushers. Ball carriers try to slip out of tackles. Receivers try to evade their coverage. Computer-controlled players make countless decisions--do they notice gaps? How quickly do they move to close/exploit them? How likely are they to be fooled by fakes? How quickly do they notice their error when they are? How quickly and accurately do they correct the error?

And so we found that when we have a vast number of knobs to turn, a tiny adjustment to each one of them adds up to a huge change, but you can't point to any one thing and say "I see what you did there."

Make a fumble 1% more likely than it was. Make your blockers 1% more likely to let rushers slip by. Make your receivers 1% less likely to make that fingertip grab, and 1% more likely to cough up the ball if they're hit right after making the catch. Make the secondary 1% more likely to see through a fake, 1% faster to read where the pass is going, 1% more accurate when they jump to make the interception. All those 1%'s add up to You Can't Win.

I imagine the economy works about the same way. How many knobs does the executive branch control? I don't know--I'm just some computer programmer out in the small-business world, but I bet that someone like Professor DeLong can list a hundred knobs that an administration can turn, each a tiny bit, to add up to a huge result that seems to come from nowhere in particular and everywhere at once. I bet he could list a hundred before finishing his morning coffee.

[max]

I would like to see those numbers without the year shift (you eat the previous guy's leftovers) and without WWII. (Leave in Korea, Vietnam and Iraq. Fair's fair.) I don't doubt the numbers would level out significantly. (I also suspect, as always, that the deflator is being understated.)

That said, I am with Waldman at least partially: crowding out + lower taxes + deficit spending = big bond gains + slow growth + big increases in inequality. (Poor people pay taxes to pay off the interest on bonds owned by rich people.)

Or rather: deficit spending where the budget grows at the same rate as always is bad for the economy. If you tax it and spend it (or even tax more and spend it) that's ok.

max
['Reagan: Keynes for the rich.']

[anonymous]

I think Jason fundamentally nails it.

And I also think one of the big areas affected are that the rent-seekers know they're going to get a much less receptive audience if they simply burn, loot, and pillage.

They're also probably forced to avoid the more blatant "let's rip off orphans and widows" stuff, and also throw in something closer to fair wages in their activities.

Add it up, and it works because what Democrats believe is true--that if you treat people more decent & pay 'em more fair, things work better across the board. This is opposed to R models, which are based on letting laws be used to aid the powerful in their exploitation of the less-powerful.

We can't see the data well enough to get it. But it does appear to be true, that when you stop systematically screwing people over, you end up with a stronger economy that grows more, in unexpected ways.

[Barry]

IIRC, even Krugman once pointed out that a president has a large number of levers which can be moved somewhat, and that the cumulative impact could well be large.

[marych]

Is there a set of graphs that breaks this down into terms? Roosevelt III, Eisenhower, Reagan, and Clinton (and Bush43 to come) are the eight year spans, and the rest are four. Reagan's terms, e.g., were hugely different, weren't they?

[rdan]

http://angrybear.blogspot.com/2008/07/by-cactus-and-before-anything-else-lets.html can add a few charts as well.

[Fifi]

There were those incredibly spooky charts of income growth by decile or quintile and GDP growth by year in the presidency.

They showed a very consistent pattern of vastly better GDP growth, income growth for everybody and inequality reduction during Democratic presidencies than during Republican presidencies with one exception. During election year, GDP growth was always much better under a Republican incumbent (and then, went splat if elected) while it was much crappier for a Democrat, making it harder to be elected/re-elected.

I can't remember where I saw those and I can't find them anymore. Was it Krugman? It could even have been on this blog. Did I dream those up?

[The Objective Historian]

Hmmm ....

But since the Fed controls monetary policy and Congress controls tax and spend legislation, perhaps it would be better to compare years of party control of Congress (and the Fed which is more nuanced) and to compare FDR Democrats with LBJ Democrats. The real point made here is that we had great annual growth years in this nation before the "Great Society Program" and after and more specifically that the pre-Reagan Economy was worse than the post-Reagan economy. Moreover, how much credit can the Democrats really claim for the 90s when Gingrich and Greenspan ran the economy and the only signature economic policies were "Republican," i.e., welfare reform and NAFTA-Mexico?

TOH