Saturday, June 28, 2008

Playing with Political Numbers

I'd seen numbers implying that, in general, the economy does better (higher GDP growth) when a Democrat is in the White House. So, I decided to grab some GDP data (from http://www.bea.gov/national/index.htm#gdp) and see what I could find. Among other things, I wanted to see if there was a time-delay effect, or if the party controlling Congress mattered. I'm using the "chained (2000) dollars" numbers, as that seems to be the preferred one in the BEA press release.

Since the data provided is since 1930, let's take that as our first cut. Since 1930, the highest overall increase in GDP is...when the President was a Democrat last year, followed by years when the President is a Democrat. The lowest was when the Senate is controlled by Republicans that year, followed by years when the House is controlled by Republicans. For every position, the GDP is higher for Democrats than Republicans.

AVERAGE SAME-YEAR GDP:
PRESIDENT: Republican 1.855263, Democrat 5.075000
SENATE: Republican 1.209091, Democrat 3.883582
HOUSE: Republican 1.580000, Democrat 4.415094

AVERAGE NEXT-YEAR GDP:
PRESIDENT: Republican 2.064865, Democrat 5.142500
SENATE: Republican 1.800000, Democrat 3.974242
HOUSE: Republican 2.008000, Democrat 4.459615

AVERAGE 2-YEAR GDP:
PRESIDENT: Republican 3.058333, Democrat 4.460000
SENATE: Republican 1.850000, Democrat 4.090909
HOUSE: Republican 3.129167, Democrat 4.103846

AVERAGE 3-YEAR GDP:
PRESIDENT: Republican 3.560000, Democrat 4.422500
SENATE: Republican 3.422222, Democrat 4.101515
HOUSE: Republican 3.995652, Democrat 4.030769

AVERAGE 4-YEAR GDP:
PRESIDENT: Republican 3.829412, Democrat 4.315000
SENATE: Republican 3.937500, Democrat 4.110606
HOUSE: Republican 3.968182, Democrat 4.144231


Now, it's possible that time frame gives the Democrats too much of an advantage - the Republicans, after all, are penalized with the first three years of the Depression while the Democrats get all of World War II and the recovery afterwards. So let's try again with the lower boundary of 1953, the start of the Eisenhower administration. And, indeed, the numbers get closer. But the highest GDP growth average is for years when a Democrat is president, followed by years when the Democrats controlled the House four years prior. The lowest, oddly enough, is when the Republicans controlled the Senate 4 years ago, followed by years when the Republicans control the Senate. The numbers are much closer, with the Republicans having higher averages in 7 of the 15 comparisons listed.

AVERAGE SAME-YEAR GDP:
PRESIDENT: Republican 2.814286, Democrat 4.090000
SENATE: Republican 2.300000, Democrat 3.444681
HOUSE: Republican 3.200000, Democrat 3.322857

AVERAGE NEXT-YEAR GDP:
PRESIDENT: Republican 2.855882, Democrat 3.930000
SENATE: Republican 2.787500, Democrat 3.334783
HOUSE: Republican 3.350000, Democrat 3.197059

AVERAGE 2-YEAR GDP:
PRESIDENT: Republican 3.442424, Democrat 3.140000
SENATE: Republican 3.328571, Democrat 3.328261
HOUSE: Republican 3.705263, Democrat 3.117647

AVERAGE 3-YEAR GDP:
PRESIDENT: Republican 3.318750, Democrat 3.155000
SENATE: Republican 2.616667, Democrat 3.339130
HOUSE: Republican 3.172222, Democrat 3.300000

AVERAGE 4-YEAR GDP:
PRESIDENT: Republican 3.145161, Democrat 3.495000
SENATE: Republican 1.840000, Democrat 3.439130
HOUSE: Republican 2.535294, Democrat 3.655882


Okay, let's move forward a bit more. Let's try 1977, the start of the Carter administration; that'll skip over all of Vietnam and its immediate aftereffects. Now, the Republicans kick in a bit. The highest average GDP increase is when the Republicans controlled the House two years prior, followed by years when the President is a Democrat. The lowest average increase is when the Democrats controlled the House two years prior, followed by a tie between years when the Republicans control the Senate and when the Democrats controlled the House three years prior. Again, the Republicans have the advantage in 7 of the 15 comparisons.

AVERAGE SAME-YEAR GDP:
PRESIDENT: Republican 2.768421, Democrat 3.583333
SENATE: Republican 2.416667, Democrat 3.244000
HOUSE: Republican 3.338889, Democrat 2.730769

AVERAGE NEXT-YEAR GDP:
PRESIDENT: Republican 2.888889, Democrat 3.250000
SENATE: Republican 2.650000, Democrat 3.129167
HOUSE: Republican 3.366667, Democrat 2.533333

AVERAGE 2-YEAR GDP:
PRESIDENT: Republican 3.311765, Democrat 2.425000
SENATE: Republican 2.860000, Democrat 2.962500
HOUSE: Republican 3.611765, Democrat 2.000000

AVERAGE 3-YEAR GDP:
PRESIDENT: Republican 3.237500, Democrat 2.533333
SENATE: Republican 2.950000, Democrat 2.933333
HOUSE: Republican 3.325000, Democrat 2.416667

AVERAGE 4-YEAR GDP:
PRESIDENT: Republican 2.980000, Democrat 3.141667
SENATE: Republican 2.733333, Democrat 3.091667
HOUSE: Republican 2.806667, Democrat 3.358333


So, what does this mean? Well, it's hard to tell overall. The most consistent numbers seem to be that having Republicans control the Senate is bad - those are the lowest numbers in 10 of the 15 sets, and the Republican Senate average is lower than the Democratic Senate in 13 of the 15 sets...and those last two are very small differences (0.01667% and 0.00031%). That may just be because there are relatively few years with Republicans controlling the Senate, just 22 years out of the full 78-year range. Over the very long term, the Democratic numbers are higher, but recent trends seem to indicate that Republican control of the House is good economically.

I've made the script available at http://www.kazrak.com/gdpanalysis.py in case anyone else wants to play with it. (Yes, I know it's ugly.) Other analysis I want to play with: the effects of having multiple groups controlled by the same party at once (is there a more pronounced effect for, say, controlling both houses of Congress, or even both houses plus the Presidency, or is economic growth better when there's a party split?), or the effects of larger vs. smaller majorities in Congress.

Some notes on methodology: Party control information came from http://en.wikipedia.org/wiki/Party_Divisions_of_United_States_Congresses, with Presidents and Congresses sworn in during a year counting for that year. For the years marked as having changes (1931 House, 2001 Senate), both were counted as Democratic because they spent more time as Democratic than Republican.

No comments: