A Scientist’s View of Political Party Performance

by Carl R. Summers

Carl R. Summers MS’80 enjoys evaluating the performance of Republicans and Democrats to see if their claims are correct—or not.

One of a scientist’s duties is to help society differentiate myth from reality. And for a hardcore researcher, there is nothing quite as exciting as watching results appear with strong evidence that contradicts the conventional wisdom.

One of my favorite hobbies is to evaluate the performance of the two dominant U.S. political parties. Some may find this avocation a bit of a leap from my day job as a statistician/researcher with the Defense and Veterans Traumatic Brain Injury Center at Walter Reed Army Medical Center in Washington, D.C. But given my training in psychology, economics, business administration, and research and evaluation methodology, my choice of hobbies makes sense.

Just like the medieval philosophers who allegedly spent hours debating how many angels could dance on the head of a pin, many of our politicians commonly make assertions based on myth, rather than available data, to justify their claims.

To disseminate my observations to as broad an audience as possible, I wrote A Voter’s Guide to Political Party Performance, a seven-part series of commentaries geared for a general audience and published in the new online magazine OutsidersDC (www.outsidersdc.com). The first in the series, Political Angel Dancing, makes a simple point that unlike the Founding Fathers’ vision of running our democracy using reason, far too often we Americans base our voting decisions on myth or emotion. A more sensible way to make our political decisions is to compare the two parties’ actual performance.

I compared the performance of Republicans and Democrats by averaging the relevant data for the years each party occupied the White House, beginning in 1950 and continuing to the present, using the latest available data. I also included an analysis of the Senate and House taxing and spending records based on majority party.

Historically, the conventional wisdom holds that the Republican Party is the “party of business”-that is, the party that institutes policies that create superior business conditions. If that supposition is true, then the annual real (adjusted for inflation) increase in output of goods and services, or Gross Domestic Product (real GDP), should be greater during Republican rather than Democratic administrations. I tested this by comparing the average annual increase in real GDP between Republican and Democratic years. The results showed that Republican administrations had an average real increase of 2.8 percent, while the Democratic average was 4.4 percent. Thus, in spite of claims to the contrary, businesses perform an average of 57 percent better under Democratic administrations than under Republican rule. While we may not understand the reasons for the large difference in performance, it’s clear that the Republican claims of improved business output during their stewardship are at odds with this empirical evidence.

What about unemployment? If the Democrats are the party of the working person, then the unemployment rate should be lower during Democratic rather than Republican administrations. In fact, the average unemployment rate was significantly lower statistically for Democrats (5.1 percent) than for Republicans (6.1 percent). Thus, the Democratic claim is confirmed.

As for inflation-always a timely topic-it has historically been almost equally distributed between the parties. During the post-WWII years, Republicans averaged 3.9 percent inflation and Democrats 3.8 percent—too close to call statistically and indicating that inflation appears to be unrelated to the political party in power.

One of my favorite topics to take on is the Republican “trickle down” economic theory. Although various versions of the theory have been expounded over the years, the general idea is that by lowering taxes on businesses and high income citizens, they will, in turn, invest in plants and equipment, largely via the stock market. This, in theory, raises the value of businesses as reflected by the broad stock indexes. The value gained then “trickles down” to less affluent Americans, boosting the overall economy for all.

The key assumption here is the marginal increase in stock value. Most Republican presidents since World War II claimed to have implemented some version of the trickle down theory during their tenure, so it would stand to reason that the annual gain in stock market indices would increase faster, on average, during Republican rather than Democratic administrations.

To test this hypothesis, I deviated from official government statistics and used a respected business data source, the Dow Industrial Composite Index, and computed the annual changes. The results were intriguing: The average change in the Dow during Republican administrations was 7.4 percent, and 10.4 percent during Democratic rule. While it may seem (ironically) that the Democrats had a clear advantage, the conclusion is not that simple. This is because the stock market varies so much over time that what appears to be a substantial difference is not large enough to meet the statistician’s standard of statistical significance, so this case is too close to call. This means that the Democrats can’t legitimately claim to have the secret to stock market success. However, the Republicans’ key claim that the financial markets respond positively to their policies is likewise not supported. The variability in the Dow is significantly greater under Republicans than under Democrats. I describe the Dow during Democratic years as “merry-go-round” markets-with relatively small, somewhat consistent ups and downs-and the Republican stock markets as “roller coaster” markets, marked by large, unpredictable peaks and valleys.

Next, I investigated the Republican claim that Democrats “tax and spend.” If so, then average taxes should be higher during the years that a Democrat occupies the White House. Likewise, because Congress must approve the budget, spending would be higher when Democrats are in the majority in the House or Senate.

To control for the effects of inflation, changes in tax policy, variations in the economy, and population growth, I selected “federal receipts as a percent of real GDP” as the appropriate measure of overall taxation. The results showed that the Republicans taxed an average of 18.9 percent of real GDP, while the Democrats taxed at 19.1 percent. With only one fifth of a percentage point difference, these results were statistically too close to call.

The situation in the U.S. Senate, however, is an eye opener. Since 1950, when the Republicans were in control, the federal government collected 18.5 percent of GDP, while a Senate controlled by Democrats brought in 17.7 percent of GDP. The difference of .8 percent of GDP is large enough to be of statistical significance.

Likewise in the House, Republican majorities taxed 18.7 percent of GDP to the Democrats’ 17.7 percent, even more statistically significant. This means that after all of those years of Republicans blaming high taxes on their opponents, the reality is that taxes are almost exactly as high regardless of the President’s party affiliation but higher when the Republicans control the House and/or Senate.

As for spending, I used federal outlays (expenditures) as a percent of GDP to compare Republican to Democratic presidential years, finding that Republican presidents outspent their Democratic counterparts by a statistically significant 5.2 percent: Republican presidents spent 20.2 percent of GDP, and Democratic presidents spent 19.2 percent. [The spending difference is calculated as: (20.2-19.2)/19.2 x 100 = 5.2 percent.]

Similar results were found in the U.S. Senate, where the average annual federal expenditure was 20.5 percent of GDP under Republican majorities and 19.4 percent of GDP under the Democrats, which indicates that Senate Republicans have been more proficient at outspending their Democratic counterparts than were Republican presidents.

The House results were statistically too close to call, with House Republicans spending 10.6 percent of GDP to the Democrats’ 10.9 percent. To use a classic understatement, the Republicans are not in the strongest position to credibly label their opponents as “big spenders.”

So, who should the American people vote for in the upcoming presidential election based on economic considerations? Well, I firmly believe that in a free society, scientists and scholars should provide the public with correct information and principles and the public should govern themselves.

Return to Summer 2008 table of contents | Back to top