Share |

Defining the Baseline Down

e21 | 09/12/2012

Click here for a printer-friendly version of this article.

When defending the Administration’s economic record, the President and his supporters try to deflect blame for the slow and halting nature of the recovery by pointing to the condition of the economy at the time he was inaugurated. The basic argument is that the economy was losing 800,000 net jobs per month in January 2009 and is now adding jobs, so the Administration’s policies must therefore be a success. Using the depths of a financial crisis as the baseline from which the recovery should be measured is problematic to say the least

The shock of the financial crisis could be likened to an earthquake. Just as earthquakes eventually end and the ground stops shaking, economic crises eventually abate and the economy returns to its prior path. Except, in this case, the crisis ended but the U.S. economy never returned to normal; instead, it started on a new, unfamiliar, slow-growth trajectory that falls well short of being called a true recovery.

Yet, likening the financial crisis to an earthquake is not simply a useful metaphor. Contemporary econometric models look at exogenous shocks like a financial crisis in a similar manner to an earthquake. Both seismograms and time series econometrics rely on the estimation of impulse response functions, which measure how dynamic systems respond to external shocks. Chris Sims was awarded the 2011 Nobel Prize in economics for his work in developing such models. The seismogram is the record of the ground motion as a function of time. The idea is to measure the magnitude of the change in ground motion after an earthquake and how quickly it returns to normal levels. Time series econometric models measure the magnitude of the change in economic aggregates like GDP, employment, and asset prices in response to an extreme event and how quickly those series return to previous trends.

Figure 1: Daily Returns on the S&P 500

Figure 1 plots the daily returns on the S&P 500 from 2007 to September 2012. The area circled in red is the period immediately following the Lehman Brothers’ bankruptcy filing. Although the crisis was primarily focused on the credit markets, overall financial market illiquidity caused investors to liquidate stocks since they could not sell other obligations or raise external finance. The initial period of extreme volatility was followed by an “aftershock” in March 2009, but by May 2009, the series returned to is normal trend.

Compare the stock market data to Figure 2, which plots the seismograph from the Jan Mayen Earthquake off of the coast of Iceland. The obvious similarity to the stock market graph makes it easy to understand why similar methods would be used to measure both series. The only substantive difference is the duration of the shock, with the earth movement returning to normal within about 16 minutes while it took about 8 months for normality to return to the stock market after the shock of September 2008.

Figure 2: Seismogram

There are two ways to measure long-run growth in the economy in contemporary econometrics. The first, posits that the economy is “trend stationary” or follows a long-run trend no matter how far it may fall off of the trend at any point in time. No previous recession – not even the Great Depression – has knocked the U.S. economy off of this trend: the deeper the recession, the more robust the recovery. In the 1933-1937 period following the Great Contraction of 1929-1933, the economy grew by 39% on a cumulative basis. After the recession of 1938 and World War II, the economy again grew rapidly and returned to the previous path. This recovery is the only one in history to have failed to return the economy to its previous trend.

Figure 3 plots the 1929-2007 trend for the economy relative to actual GDP since 1980. Rather than accelerating to return to the trend growth rate, the economy has actually grown at a slower average pace after the recession ended than it did before the recession began. This is anomalous because one would have anticipated an especially brisk recovery to match the size of the contraction. For the first time in U.S. economic history, the depth of the contraction and the strength of the recovery have been asymmetric, causing the economy to fall well below the level consistent with its long-run trend.

Figure 3: Actual GDP versus Trend

The alternative assumption is that the economy is “difference stationary.” According to this view, the economy returns to its long-run average growth rate rather than the previous trend. If the economy had previously grown at an average of 3% prior to the shock, it would then be expected to grow by an average of 3% in the recovery. This method does not assume any “catch-up” growth where the economy grows faster than average to return to the previous trend. The economy’s trajectory is assumed to be permanently altered by the recession.

However one models the economy, the shocks end and growth returns. Of all the assumptions made in econometrics, this may be the one that’s on the most solid ground. Were this not the case, the perpetual recession would eventually cause all factories and offices to sit empty, capital equipment to go unused, and the entire labor force to be unemployed. No one believes this would be the natural endpoint for any recession – no matter how deep – so using this trajectory as the baseline to measure the President’s economic record makes little to no sense.

Incidentally, both CBO and OMB assume the economy is trend stationary, which is why they assume growth will accelerate to 4% in 2014 and stay that high until the economy returns to trend. As described above, this may seem unrealistic since businesses today are behaving in ways contrary to how one would expect during a recovery. Specifically, businesses are holding $2.5 trillion in cash balances, according to the Fed, an all-time record (see deposits and Treasury holdings in B. 102 and B.103). However, this figure could seriously understate cash holdings because it does not account for cash holdings in overseas subsidiaries. Including these balances could bring total cash holdings of nonfinancial businesses to over $6 trillion. In a normal recovery, businesses would be reinvesting cash flow into capital equipment and new productive capacity. While some claim that businesses are not reinvesting cash flow because of a supposed lack of demand, the jobs created by business investment would generate that demand, at both the companies selling the capital equipment as well as the businesses who install it. Instead, regulatory uncertainty and expectations of future tax increases encourage businesses to hoard cash rather than invest it.

Perhaps the most troubling aspect of this narrative is how it has emboldened the Federal Reserve to try to solve problems it is powerless to address. The weak August unemployment report has led many to assume that the Federal Reserve will re-launch quantitative easing at its policy meeting that concludes today. Were this policy simply ineffective, there would be little to complain about. But, as the Bank for International Settlements explained in its 2012 annual report, extreme monetary accommodation could be a net negative for the economy when it keeps zombie borrowers alive, reduces banks’ incentives to make new loans (due to the narrow spread between deposits and longer-term interest rates), and encourages speculative fund flows to higher risk asset classes or markets.

The economic crisis of 2008 is now blamed, at least in part, for the financial excesses that naturally arise when interest rates are kept too low for too long. Instead of learning from this mistake, the Federal Reserve seems intent on repeating it.


e21 PROJECTS & PARTNERSHIPS