Market update from James Marple, Economist TD Bank.
PRESIDENTS, THE ECONOMY AND FINANCIAL MARKETS
October 27, 2008
Even amidst all the economic and financial turmoil taking place in the United States, it’s hard not to notice that the U.S. is also in the final leg of a presidential election campaign. The election taking place November 4th 2008 is an historic one for the United States. Beyond the unique qualities of the candidates, it will be the first election since 1952 without a sitting president or vice-president on the ballot and takes place in an environment of extreme financial market uncertainty and anemic economic growth.
The importance of the U.S. president in determining economic policy and the regularity of presidential elections has led many to speculate on the possible correlation between the timing of the presidential election and the performance of the economy and financial markets. This paper explores the link, finding little evidence of fiscal policy – either through taxation policy or government spending – cycling around a presidential election and finding that the increasing importance of an independent Federal Reserve has made the United States less vulnerable to politically opportunistic monetary policy.
Stock market gains have historically performed best in the third year of a president’s term but this outperformance is not matched by outperformance in either economic growth or government spending. Strong economic growth and low inflation are, however, important predictors of an incumbent party retaining the Presidency. The stock market is also a good indicator of financial market confidence in the incumbent-presidential party. Stock markets perform better in years when incumbent parties are re-elected than when they are defeated.
Political-Business Cycles, Fiscal Policy and Fed Independence
Since 1960 the U.S. has experienced seven business cycles. According to the NBER (the official arbiter of U.S. business cycles) the average length of a business cycle over this time period is seventy-five months or a little over six years.1 Presidential election cycles are even more regular, occurring every 48 months. The basis for the relationship between elections and economic performance is the popular notion that politicians up for re-election will attempt to stimulate the economy in order to increase their chance at re-election, even at the cost of higher inflation and slower growth later on. The example that perhaps best exemplifies this notion is the action of President Richard Nixon in the lead up to the 1972 election. In the year before the election, Nixon hiked Social Security benefits by close to 20%, while also aggressively lobbying the Federal Reserve to loosen monetary policy.2
Considering the political environment at the time, it’s not surprising that the first economic literature on politically-led business cycles took shape in this period. A 1975 paper by William Nordhaus of Yale University, entitled “The Political Business Cycle,” proposed that opportunistic monetary policy would result in unemployment and inflation cycling around election periods. According to this theory, the unemployment rate would fall in the final two years of the presidency and rise in the first two. He tested his hypothesis by looking at movements in the unemployment rate in elections from 1948 to 1972. The data backed him up – in five of six pre-election periods, unemployment rates fell and in five of six post-election periods unemployment rates rose.
The 1970s marked an important turning point for both the application of monetary policy and theoretical thinking about its effectiveness. The ability of administrations to influence monetary policy played a roll in the run-away inflation experienced in the late 1970s and rendered ineffective the Federal Reserve’s ability to ensure long-term price stability. Incidentally, Nordhaus’ hypothesis also broke down in this period, with unemployment rates rising in the lead up to both the 1976 and 1980s election.
The stagflation experience of the 1970s also led to an increased emphasis on Federal Reserve independence and on its role as an inflation fighter. The appointment of Paul Volcker, as Chairman of the Federal Reserve, marked a shift in the attitude of monetary policy makers towards price stability, even at the cost of higher unemployment in the short run. This was evident in the actions of the Federal Reserve under Volcker. Although appointed by President Carter, Volcker raised interest rates precipitously through the end of Carter’s term in order to break the back of inflation. The recession that followed has been noted as one of the reasons for Carter’s drubbing by Ronald Reagan in the 1980 election.
Following the 1970’s experience, the independence of monetary policy has been evident in several more instances surrounding election periods. Interest rates remained high through the 1980s and were raised in the lead up to both the 1984 and 1988 elections (but in contrast to 1980 the higher rates did not hurt Reagan’s and Bush Sr.’s election bids).
While interest rates declined before the 1992 election, the recession that took place near the end of 1990 had resulted in an economic contraction, a rising unemployment rate and a falling inflation rate in the lead up to the election. The poor performance of the economy led to George H.W. Bush losing the election (and the coining of the term “it’s the economy, stupid” by Clinton adviser James Carville). Bush at the time had been calling for the Federal Reserve to lower interest rates and placed part of the blame for his defeat on the Fed’s reluctance to do so.
It is no coincidence that the increased importance of Federal Reserve independence has resulted in a break down in observable patterns of unemployment, inflation and interest rates around elections, as well as a move towards greater stability in economic growth, and lower levels of inflation and interest rates. It also implies that forecasting economic variables on the basis of past relationships between political and economic variables is not likely to yield useful predictions.
What about fiscal policy?
The 1972 Nixon case is more than anything an example of fiscal policy being conducted in a way as to benefit the incumbent president. However, looking at the data over a longer period of time, there is little evidence of a consistent cyclical pattern in either tax policy or government spending around election periods. In fact, federal government spending has on average grown faster in the first two years of a presidential-term than in the second two years. Even as a share of economic growth, the first two years of a president’s have grown faster in the first half of the president’s term. Further, a regression of taxes on the election cycle shows no relationship between the two variables. Despite popular opinion that presidents cut taxes and spend to win re-election, the data fail to confirm this notion. There are several reasons why this is likely the case. For one, fiscal largess is not always the best way to win votes. In situations where voters have become increasingly concerned about the state of the nation’s finances, tighter fiscal policy might just as easily be the best way to win electoral favor. Certainly in this election, both candidates have signaled to voters the need to reign in spending rather than raise it. Secondly, the political system in the United States is characterized by multiple checks and balances, which limit the ability of an incumbent president to singularly dictate the course of fiscal policy. Congressional elections are held mid-way through a president’s term and it is a relatively rare occasion for the party of the president to also control the Senate and the House of Representatives. Indeed, since 1944 a Republican has been in the White House 56% of the time, but had a majority in the Senate only 38% of the time and the House of Representatives only 23%. Republicans have only held both the Presidency and controlled Congress for 8 out of the last 64 years (13% of the time), while the Democrats have had this “triple-majority” in only 20 of the last 64 (31%).
Economic performance as a political predictor
Even if fiscal and monetary policy do not cycle around election periods, this does not mean that economic performance is unimportant in determining who is elected president.
Election prediction models have become quite popular for estimating the impact of economic variables on who is elected president. One of the first and perhaps most famous economic model for predicting the outcome of the presidential election is that of another Yale University faculty member, Ray Fair who first developed the model in 1978. The Fair model uses both political and economic variables to predict the percentage of the popular vote expected to be received by the incumbent party in a Presidential election.3 The economic variables used in the analysis are inflation and growth in real GDP. In addition to economic variables, political variables are used as predictors. For instance, one variable takes into account whether the candidate is an incumbent or a member of the incumbent’s party (both are positive for re-election chances), while another captures how long the party has held the presidency (negative as voters tire of one party in power for too long). These variables are then used to predict the percentage of the popular vote received by the incumbent’s party.
The results predicted in the Fair models can also be observed simply by examining average growth rates over the course of presidential term. Between 1960 and 2004, real GDP grew by an average of 5.1% in years when the incumbent party is re-elected, a full 1.9 percentage points higher than the 3.2% average growth in years when they are defeated. Inflation on the other hand is significantly lower in years when incumbents are re-elected, averaging 2.9% in re-election years and 5.2% in defeats. In short, a buoyant economy and low inflation are positive for an incumbent parties running for president, while a struggling economy is bad news for re-election.
Presidents and stock markets
If it’s the economy that influences the outcome of elections, and incumbents are more likely to be re-elected when the economy is doing well, then financial markets too should be expected to perform better when incumbents win re-election. Looking at the data, this is exactly what we observe.
Since 1960, the incumbent party has won the election six times and lost the election six times. The election of 2000 is an important caveat since the Democrats lost the election but won the popular vote. Excluding 2000, the tally is six wins for the incumbent and five losses. Unsurprisingly, stock markets in the year leading up to an election victory outperform in years leading up to defeats. In the six elections won by incumbents the average annual change in the Dow-Jones Industrial Average is 10.8% compared to 6.4% when incumbents lost.
Looking at the average performance of a stock market during a full election year, a few patterns emerge around the election (see chart next page). Patterns that are also dependent on whether or not an incumbent wins. As shown in the graph, a pre-election sell-off in stock markets seems to occur fairly dramatically in years when an incumbent party loses the race. This is likely related to the increase in uncertainty surrounding the outcome of the election, especially as it becomes more evident that the incumbent is likely to lose. There is a post-election stock market rally in both cases when an incumbent wins and when an incumbent loses but this is more pronounced in the former case.
Bottom Line
The influence of the economy on politics has been on display throughout the current presidential election campaign. The importance of fiscal and monetary policy in determining economic outcomes has led to comments on the cycling of policy around political schedules. The movement of economic variables around election periods appears to have weakened with the increased importance of an independent Federal Reserve (as have economic fluctuations in general). Nonetheless, a strong economy and low inflation is still important for a political party hoping to retake the presidency.
Stock market performance is also correlated with political outcomes. Stronger confidence and higher returns appear with some consistency in years when incumbent parties are re-elected. There also appear to be certain “within-year” patterns in financial markets that also depend on the timing of the election (especially elections where it is becoming more apparent that there will be a change in governing party). Post-election rallies are not uncommon in election years but appear to be greater when incumbent parties are re-elected. Nonetheless, with current economic weakness expected to persist well into next year, investors may have to wait a bit longer for their post-election celebration. However, a case could be made that it is not the election that ultimately is driving these trends, but instead the underlying economic and financial environment.
James Marple, Economist
416-982-2557
Endnotes
(1) National Bureau of Economic Research . “Business Cycle Expansions and Contractions.” . Business cycles were the same length over this period, whether defined as peak-to-peak or trough-to-trough.
(2) Rogoff, Kenneth. “Bush Throws a Party.” Foreign Policy. March/April 2004.
