Since early this spring, faint but distinct hints of slowing demand have begun trickling out of a variety of economic indicators. Beginning in March, retailers have consistently reported results that have fallen short of "plan." In addition, auto dealers had begun to report that the red-hot pace of vehicle sales late last year and early this year had moderated a bit. Last week the U.S. Commerce Dept. reported that May retail sales dropped 0.3%. Analysts had expected May sales to rise 0.1%. This marks the third straight month sales have fallen short of expectations. Meanwhile, automotive dealers sold fewer motor vehicles in May than in any month since last October. While the sales volume remains on par with the record-setting pace of 1999, sharp declines in two of the last three months suggest the steadily rising uptrend of the last 18 months may have ended. Importantly, auto sales remain especially sensitive to the promotional pricing and financing terms offered by dealers and manufacturers. DaimlerChrysler, which saw the biggest sales drop among the Big Three U.S. car makers, announced it would enlarge its cash rebates but would reduce its finance-rate subsidies. Rebates may help keep sales firm, but higher financing costs will probably forestall an acceleration of sales growth. Recent reports also indicate other interest-rate sensitive sectors have begun to moderate. Housing starts fell 3.9% in May. As with vehicle demand, sales volume remains high, but growth has clearly stopped accelerating. The orders index compiled by the National Assn. of Purchasing Management indicates demand for manufactured goods may not have improved much in May from April's decline. The NAPM new orders index fell sharply, hitting an 18-month low. Moreover, the employment report indicates this softer demand provoked a swift reaction. Manufacturers cut 17,000 workers from their payrolls in May, reduced the average work week by 48 minutes, and cut overtime by 24 minutes. The "diffusion index" of payroll employment in the manufacturing sector fell to 42.4, indicating that job cuts were broad-based. Reflecting a sharp curtailment of production, the index of aggregate hours worked in manufacturing fell 2% in May. Signs of a softer labor market were not confined to the manufacturing sector. Indeed, the Labor Dept.'s report on employment, hours, and earnings transformed the stream of data pointing toward slower growth into a flash flood. While the government hired 357,000 workers to assist in the decennial census, private-sector industries pared 116,000 workers from their payrolls in May. That was the biggest one-month decline since November 1991. Cutbacks spanned a wide array of industries, including construction (minus 29,000), retailing (minus 67,000) and "help supply" services (minus 36,000). As with the manufacturing sector, the average work week of all industries fell and the aggregate index of hours worked declined 0.6%. Data from the separate household survey showed an even more dramatic turnabout. As measured in that survey, employment fell a stunning 991,000 workers, the biggest monthly decline in the 52-year history of these data. Consequently, the jobless rate bounced up 0.2% from April's 30-year low of 3.9%. The pervasive softening of labor market conditions reflected across the entire breadth of this comprehensive report defies explanation. Rarely does an economic report as sweeping and comprehensive as the employment report so profoundly challenge prevailing assumptions about the economy. Other recent labor market indicators held no hint of such a dramatic change. Indeed, perceptions that jobs are plentiful helped push the Conference Board's index of consumer confidence to its second highest level ever in May. Similarly, the NAPM's employment index for May touched its second highest level of the current expansion. These observations cast doubt on the meaning and significance of the apparent sudden softening of the jobs market. To some extent, aberrations in normal weather patterns and other seasonal anomalies may have distorted the May numbers. For instance, unusually mild winter weather may have kept more construction workers on the job during the Winter so fewer needed to be rehired in the spring. When adjusted for "normal" seasonal variation, declines in construction employment in the spring offset the increases of the winter. The Bureau of Labor Statistics acknowledged it has difficulty making adjusting data from the household survey for seasonal patterns that occur in May. For instance, because the jobs survey occurred unusually early in the month, many college students may not have been able to join the workforce as early as usual. But the BLS knew of no comparable biases in the broader establishment survey. As perplexing and anomalous as the jobs data may be, they cannot be ignored. Qualitatively, if not quantitatively, they corroborate other signs of moderating economic activity. This evidence predates the Fed's most aggressive policy tightening move, suggesting that previous actions, after the always-unpredictable lags, had begun to exert their influence. But with the impact of the latest move yet to be felt, no responsible policy-maker would risk the consequences of further tightening. David H. Resler is chief economist and managing director of Nomura Securities International Inc., New York.