FRB Dallas Home » Research & Data »Update » U.S. »2012 »Outlook Remains Positive but Cloudy

Research Publications
National Economic Update

Outlook Remains Positive but Cloudy

April 25, 2012 · Update in PDF PDF

The employment report, one of the most closely watched economic data releases, was unexpectedly weak in March. Most other indicators have been more positive. Consumption growth held up in the first quarter of 2012 despite higher energy prices that typically have a dampening effect. In the long term, matching previous rates of growth may be difficult due to demographic trends.

Disappointing Job Growth in March

After unexpectedly strong employment reports in December through February, it appeared that the labor market might have been finally gaining some traction. Nonfarm payrolls increased by only 120,000 in March, roughly half the pace of the previous three months and more consistent with the slower-growth period of last summer. This was well below the average forecast of about 200,000 jobs (Chart 1). The household survey, from which the unemployment rate is derived, reported employment falling by 31,000, although this measure is noisier and considered to be a less reliable signal of labor market conditions. Despite the reported drop in employment, the unemployment rate actually fell from 8.3 percent to 8.2 percent in March. This decrease was due to a greater decline in the labor force, and thus, it was not necessarily a signal of improving conditions.

Mild Winter Clouds Interpretation of Indicators

Many analysts attributed weak job growth in March to the unseasonably warm weather in January and February. Employment statistics are typically seasonally adjusted, meaning that they take into account predictable seasonal variation. Not seasonally adjusted employment falls in December and January as holiday-related employment ends and winter halts weather-sensitive projects. This is expected and should not be interpreted as market slack, so seasonal adjustment increases the winter month figures relative to other months. Seasonal workers are hired on again in the spring, and seasonal adjustment decreases reported employment, giving a clearer read of labor market conditions (Chart 2).

However, in a mild winter, firms can start weather-sensitive production earlier in the year, which causes a boost in the seasonally adjusted figures. This does not reflect underlying economic strength, but a statistical illusion. Once the weather normalizes, the boost would fade and seasonally adjusted employment would appear to slow, “paying back” the previous gains because the hiring that would normally occur in the spring has already taken place.

Is there any evidence that this was the case this winter? This winter was indeed unseasonably mild. According to a population-weighted index of U.S. temperature, this winter ranks as the warmest since the series began in 1980 (Chart 3). Low temperatures are not the only hindrance to hiring. Disruptive weather, such as blizzards and hurricanes, plays a large role. Ideally, one would look at how many people would be employed if it were not for disruptive weather, but this is not directly measured. So, the number of people who are employed but cannot work due to disruptive weather is a good proxy. This measure was well below normal this winter, reflecting that it was calm as well as warm (Chart 3). This would indicate that some of the gains in earlier employment reports can be ascribed to pull-forward effects.

However, these indicators do not suggest that the weakness in March was the resulting payback. The seasonally adjusted temperature in March was even warmer than in the previous months, and conditions remained relatively calm. Remnants of winter presumably prevent some hiring in a typical March, so firms should have been able to continue to pull-forward hiring in this atypical year. If this was the case, most of whatever payback is due from the mild winter is yet to come, and the weakness in the March employment report should not be completely written off as seasonal effects.

Automobile Sales Hold Up Despite Higher Gasoline Prices

Gasoline prices have climbed back near the highs reached in May 2011 and July 2008. In both of those cases, economic activity appeared to suffer as a result, with automobile production being one of the hardest-hit sectors. In May 2011, sales of light-weight vehicles had increased by only 0.7 percent over the previous year, and in July 2008, sales had fallen by 18.1 percent over the previous year. Of course, some of this was due to supply-chain disruptions in 2011 and the recession in 2008. In March 2012, sales were up by 10 percent over the previous year, a stark contrast to the previous two oil-price spikes (Chart 4).

This suggests that consumers are not drastically decreasing spending in response to the higher oil prices. Other indicators, such as retail sales and personal consumption expenditures, also show few signs of a first-quarter retrenchment. This resilience supports the hypothesis that oil-price increases dampen economic growth only to the extent that prices rise above the maximum of the previous few years, as consumers have already adjusted their spending patterns to the higher prices.[1]

Housing Indicators Sending Mixed Messages

Housing construction indicators have shown some softness in recent months, but expectations are for a gradual improvement. Housing starts fell to 654,000 in March from 694,000 in February (Chart 5). Starts were down 5.8 percent over the previous 12 months. Housing permits, alternatively, increased to 747,000 in March from 715,000 in February and are up 4.5 percent over the previous 12 months. Although residential construction may not have contributed much to output growth in the first quarter, permits, along with the apparent halt or slowing pace of house price declines, suggest that housing construction may begin to grow more steadily.

Demographics Affect Long-Term Growth

Since the end of World War II, U.S. economic output growth has fluctuated around a trend of about 3 percent a year. Analysts often use this trend to measure the amount of economic output loss due to recession. Trend growth is more than just a statistical technique, however; it is determined by inputs such as the amount of labor available to be dedicated to economic activity. In the postwar era, women in the labor market grew steadily faster than the population, which boosted economic growth. According to the trend in the female labor force participation rate, the boost likely ended sometime in the 1990s (Chart 6). Since then, the number of women who have a job or are looking for a job has grown at about the same rate as the population. This would have the effect of lowering the trend in output growth to around 2 percent. Thus, the gap between actual output and what it would have been in absence of the recession is likely smaller than with the assumption of a 3 percent trend. Nevertheless, the unemployment rates for both women and men remain high, and growth has been inadequate to quickly restore labor market conditions.

—Tyler Atkinson


  1. “Oil Prices and the U.S. Economy,” by James Hamilton, Econbrowser (blog), March 6, 2012,
About the Author

Atkinson is a senior research analyst in the Research Department of the Federal Reserve Bank of Dallas.


Federal Reserve Bank of Dallas Seal
Federal Reserve Bank of Dallas

2200 N. Pearl St., Dallas, Texas 75201 | 214.922.6000 or 800.333.4460
Disclaimer / Privacy Policy