If a pause on this day of reflection is needed, the debate over the economic recovery offers a diversion. Market players have convinced themselves that the decline in consumer confidence over the past couple of months is setting the stage for a drop in retail spending and a general malaise in the consumer sector. It is certainly true that over the long run, declines in confidence often reflect a dispirited consumer and often precede a downward trend in spending. But this has not been the case this past year.

Market players have been disappointed all year long with the anemic pace of the economic recovery, despite prior predictions from economists that a mild recession would be followed by no more than a mild recovery. Normally, a steep recession involves sharp declines in consumer durable spending. Pent-up demand for cars and furniture, and usually housing, will drive the early stages of recovery and boost GDP growth significantly. But this time, there was no drop in housing nor declines in motor vehicle sales or furniture. Without pent-up demand, consumer spending in a recovery is bound to be mild.

Confidence and motor vehicle sales: no relation
Consumer confidence can be affected by any number of factors. For instance, consumer confidence was shaken to its core a year ago today when terrorists attacked U.S. soil. But consumer confidence had also been dazed when the stock market bubble burst and a bear market awakened. Corporate scandals among more than a few major companies have cast shadows over the validity of financial statements in general.

When consumer confidence is affected by special factors, such as the terrorist attacks or a stock market plunge, it can recovery quickly. Sharp declines might have only minor effects on spending. The Conference Board's consumer confidence index asks key questions regarding labor market conditions. Consequently, this index tends to rise and fall with the labor market, more so than the University of Michigan's consumer sentiment index which is geared toward questions on personal finances. One would certainly expect that long-term negative trends in the labor market - a rising unemployment rate for instance - would lead to falling consumer confidence. In turn, if consumers feel that their jobs are on the line, then they are less likely to purchase large ticket items such as motor vehicles. In fact, confidence was near its peak for this cycle at the same time the unemployment rate bottomed out. But consumer confidence began to turn around immediately after the attacks even as the unemployment rate stabilized. The corporate scandals took a toll on confidence, as the index again began to decline in the spring, even as the unemployment rate remained in a tight range.

Based on conventional wisdom among equity and bond market participants, the declines in consumer confidence would have suggested a sharp drop-off in purchases of large ticket items such as cars and SUVs. In fact, in the 10 months after the terrorist attacks, motor vehicle sales averaged 13.9 million units, a tick higher than the 13.8 million-unit rate recorded in the 10 months prior. Counting the August data, the average selling pace since October was a 14.1-million-unit pace. Granted, automakers devised (and still are devising) dramatic incentives to drive customers back to the showrooms, but nevertheless, consumers were willing to buy. If confidence had really been an issue, even the zero-percent financing incentives wouldn't have been enough to entice buyers.

Fed Chairman Alan Greenspan pointed out that consumer confidence and retail sales don't move in tandem month by month. In fact, look at the chart above which shows confidence and motor vehicle sales. The month of May appears to have relatively low sales - but that also coincides with a relatively "high" level of confidence. The weakest level of confidence, in October following the attacks, also corresponds to the highest selling pace ever recorded by auto and light truck dealers.

Consumer spending since September 11
Motor vehicle sales are only part of retail spending. The chart below shows how retail sales fared in the 10 months before and after September. In the 10 months before the attack, when the U.S. had already fallen into recession, monthly gains in retail sales averaged 0.3 percent. After September, when confidence was more sluggish but the economy was starting to recover, retail sales averaged gains of 0.7 percent per month. Monthly retail sales gains have more than doubled. It is true that rising auto sales have helped to boost the total. But then who are we to judge what consumers choose to purchase? If other (non-auto) retailers want to boost sales, then they might also consider offering big incentives. The bottom line? Retail sales have improved during this recovery.

Retail spending was mainly spurred by lower interest rates. Aside from zero-percent financing on new cars and trucks, many credit card companies and department stores also offered zero-interest or low interest financing for short-term periods. Aggressive refinancing, triggered by the lowest mortgage rates in nearly 40 years, also helped spending. Many homeowners took cash out of their home equity to pay down credit cards or make additional purchases. Real (inflation-adjusted) disposable income grew 0.3 percent per month on average before and after September. Nevertheless, though income growth has been stable during the past eighteen months (the early phase of the recovery), it hasn't improved significantly.

Housing activity up in recovery
Going into 2001, economists were looking for housing starts to moderate after several years of robust activity. Starts did moderate slightly at the end of the year, but once again strengthened in the first part of 2002. The monthly pace of housing starts averaged a 1.6 million-unit rate in the 10 months before September but increased to a 1.65 million-unit rate in the 10 months after September. While the Fed did not reduce its federal funds rate target in 2002, yields on 10-year Treasury notes declined in 2002 from 2001 levels and this helped reduce mortgage rates to 40-year lows.

BOTTOM LINE
Looking at consumer indicators before and after September allows us to come to a couple of conclusions. First and foremost, many factors can affect consumer confidence: from acts of aggression to corporate scandals to economic factors such as a worsening labor market. Typically, economic factors such as fear of job loss usually cause confidence and retail spending to move in tandem. However, consumer confidence indexes are not necessarily the best measure of sentiment. Given the right incentives, consumers will be motivated to buy goods and services.

Economic growth was sluggish in the early part of 2001, but until the Bureau of Economic Analysis revised GDP figures a couple of months ago, most analysts and market players were not convinced that the U.S. economy had entered into recession. In fact, the National Bureau of Economic Research (NBER), the official arbiters of economic business cycles, did set March 2001 as the business cycle peak. The months before September 11 were a period of economic recession, yet some sectors of the economy were performing relatively well - such as motor vehicles and housing. Despite the fact that pent-up demand never got a chance to develop in these sectors, housing construction and motor vehicle sales still increased in the post-September period, which can also be called the early phase of the recovery. Market players have been disappointed with the modest recovery, but in light of the relatively healthy pace of housing and motor vehicle sales during the recession, the recovery isn't all that bad.

Evelina M. Tainer, Chief Economist, Econoday