Springtime for Jobs in February?

The month's employment figures are likely to be stronger than expected as many related economic indicators are pointing up

By Rick MacDonald

How strong will February's employment report be? The market will be braced for an upside surprise in nonfarm payroll growth when the monthly report is released on Mar. 4.

Wall Street's forecasts for February job growth have been on a steady upward march over the last couple of weeks, with the median estimate from Action Economics' weekly survey of economists jumping to 219,000 on Feb. 25, from 200,000 the previous week. And a "whisper" number of 300,000 circulating in the bond market fueled a sell-off in Treasuries -- and a jump in bond yields -- on Feb. 28.

In Action Economics' view, the odds are so one-sided in favor of a stronger-than-expected figure that the bond market may be set up for a relief rally if February payrolls come in below those lofty projections. Nonetheless, our own forecast calls for a rise of 240,000 on the month, well above the median estimate.

What makes us so confident? We think the litany of evidence supporting projections of a sharp bounce in February payrolls following the restrained 147,000 gain in January is fairly strong. Here's a review of the various data reports and other factors that support our outlook:

Initial Claims: The average reading for weekly first-time jobless claims in February will be close to 310,000. This is well below the average readings of 331,000 in January, 332,000 in December, and 337,000 in November. Similarly, the initial claims reading of 303,000 during the week that the Bureau of Labor Statistics (BLS) compiled data for the monthly report is well below the similar readings of 318,000 in January, 331,000 in December, and 335,000 in November.

Indeed, broader trends in payrolls and initial claims over the last two years have tracked fairly well. The recent drop in initial claims over the last four weeks raises the chance of an upside payroll gain in February.

Consumer Sentiment: The "current conditions" series from both the Conference Board's consumer confidence index and the University of Michigan's consumer sentiment survey have historically matched up well with the broader swings in the labor market. Both revealed jumps to multiyear highs in February.

The current component from the Michigan survey was unchanged in February, from January's 110.9, which left the series at the highest level since November, 2000. The present situations component of the Conference Board's index rose to 116.4 in February, from 112.1 in the prior month, which marked the highest level since September, 2001.

Help Wanted: The January Help-Wanted index from the Conference Board jumped to 41 in February, from 38 in January, which matches the highest level since September, 2002. The increase left a year-over-year growth rate of 8%, representing the biggest gain since the late 1990s. Moreover, the index has surged nearly 14% over January and February, which marks the strongest two-month growth rate in history.

Monster Employment: Employment-services outfit Monster Worldwide's (MNST ) January employment index (which will be updated later this week for February) surged to a record level, showing the second largest month-over-month gain since its inception in October, 2003. This strength ran in sharp contrast to the disappointing January payroll gain and may flag a need for reconciliation in the February employment report.

Factory Surveys: The employment components from the February factory sentiment surveys suggest a factory-payroll rebound following January's surprising 25,000 drop in manufacturing jobs. While it's clear actual payroll growth has undershot employment sentiment for much of this cycle, the decline in January appears to be an aberration that leaves room for a February rebound.

Weather: A return to more normal weather during the February survey week, compared to the abysmal conditions during the similar period in January, could provide an additional upside pop to payrolls. We have noted the risk that the January report was restrained by bad weather in the first half of the month. January data since then -- and anecdotal comments from auto dealers and retailers -- suggest that winter harshness indeed had a big impact.

The bad weather could explain other statistical oddities. Construction employment revealed an anomalous decline of 9,000 in January payrolls, compared to an average monthly payroll gain of 22,000 over the last 12 months. Aggregate construction hours worked dropped a hefty 2.4%. This weakness was in contrast to continued robust readings of housing sentiment measures, such as the National Association of Home Builders survey, as well other residential data that captured the full month (rather than just the first two weeks, when the weather was the worst).


  Similarly, the wage and salary (W&S) data from the January personal income report revealed a much stronger rate of growth than implied by the January payroll data. One can calculate an implied aggregate wages figure from the employment report by multiplying payrolls, the workweek, and hourly earnings. Historically, this provides a close estimate for personal income, which is derived from unpublished data from the BLS that gauges activity throughout the month rather than in the survey week. In January, W&S growth was a hefty 0.6% larger than implied by the employment data. This is one of the largest discrepancies of the last decade.

Inclement weather around the survey week was a clear factor for two of these January W&S discrepancies, seen in 1996 and 1999. These distortions were followed by sizable payroll rebounds the following month, with payroll growth in February, 1996, recovering 435,000 from the 18,000 drop in January. Similarly, payrolls surged 396,000 in February, 1999, from a 113,000 gain in January of that year.

Overall, the February data should confirm that job growth remains healthy, boding well for the economic outlook. It would also suggest the excessive level of Fed policy accommodation in place for the last few years can continue to be removed. As such, the data should support a Fed tightening trajectory that most likely will continue at a "measured" pace well into 2005 and suggests upside risk to current market estimates.

MacDonald is global director of investment research and analysis for Action Economics