US Non-Farm Payrolls Preview: Worried but the signs are steady

• ADP private payrolls plunge to 27,000 in May after April's 275,000
• Initial jobless claims 4-week moving average 216,750
• Manufacturing employment PMI 53.7, services 58.1 in May

The Bureau of Labor Statistics (BLS) a division of the US Labor Department will issue its Employment Situation Report for May on Friday June 7th at 8:30 am EDT, 12:30 pm GMT.

Forecast

Non-farm payrolls are predicted to add 185,000 in May following April’s 263,000 increase. Manufacturing will gain 5,000 positions after the prior month’s 4,000.  Private payrolls should rise 175,000 after 263,000 new posts in April. Government payrolls rose 27,000 in April. The U-3 unemployment rate is projected to remain at 3.6%. Average hourly earnings on the month will increase 0.3%, in May they added 0.2% in April. Annual earnings are forecast to be unchanged at 3.2% and the workweek to rise 0.1 hour to 34.5 hours.

The Employment Situation Report

Normally called non-farm payrolls, payrolls or just NFP the Labor Department’s monthly assessment of the of jobs economy is the best known, most followed and actively traded US statistic.

The report consists of two surveys. The establishment survey queries non-farm businesses and produces the payrolls numbers, wages, weekly hours, labor force participation rate and other gauges. The household survey polls a representative sample of the working age civilian population and classifies each person as employed, unemployed or not in the labor force and computes the unemployment rates.

The best known measure of unemployment and the one that is usually meant in common referral, the U-3 rate was 3.6% in April. To be counted as unemployed under this measure a non-working individual had to have looked for work in the month prior to the survey.  The U-6 or underemployment rate includes someone as unemployed if they had looked for work in the prior year. It was 7.3% in April.  A person who is not employed or unemployed under the U-3 definition is considered as not in the workforce.

The non-farm payroll figure incorporates an estimate by BLS economists of the number of new jobs created at start-up firms that have not officially interacted with the government for tax or other reasons. These estimated jobs are revised at a later date against company and government information. The NFP report is one of the most up to date of the government’s many analytical efforts as its data is about one month old.

Coincident labor market statistics

Payrolls give the broadest picture of the US labor market but three other statistics are particularly useful in judging the health and direction of job creation: ADP payrolls; initial jobless claims; and the employment purchasing managers’’ indexes from the Institute for Supply Management.

NFP and ADP

The two payroll numbers are a part and the whole of the US labor market.

There are two important differences between the reports. The ADP information covers the 411,000 US firms that employ its payroll processing and accounting services. The BLS information reports on the entire economy including government employment at local, state and federal levels.

The second difference is that ADP includes only actual employees whose paychecks are produced by the company. The NFP figures from the BLS include a monthly estimate for the number of jobs created by newly formed companies as noted above.   

The ADP payrolls have seen several sharp declines over the past four and a half year though none were as steep or as far as the May plunge to 27,000 from 275,000. The 12-month moving averge has been climbing since September 2017 and even last month’s figure brought the average down to 202,200 still an excellent result and among the top scores in the history of the series.

Correlation

The correlation between the ADP and NFP employment tracks is trend based and directional. The correlation between the two on the month to month totals is weak.

Reuters

Of the six steepest declines in ADP over the last three and a half years, December 2015 to January 2016-271,000 to 130,000, April to May 2016-170,000 to 84,000, September to October 2016-181,000 to 91,000, November to December 2016-236,000 to 105,000, August to September 2017-174,000 to 111,000, and October to November 2018-238,000 to 136,000, four saw equally large declines in the NFP, one had a modest drop and in one instance NFP rose. Working the other way there were three large falls in NFP that were unmatched in the ADP results.

Of nine instances of large monthly variation in one of the individual series four were well matched in the other, one weakly matched and four were reversed. The May result for ADP increases the downside risk for NFP but not by very much

Employment PMI

The purchasing managers’ indexes for the manufacturing and service sectors track together over the past nine months. Both topped in September 2018, services at 60.4 and manufacturing at 58.2 and then descended through the year end and the first quarter. Manufacturing pulled a double bottom in February and April at 52.3 and 53.4. Services went to 53.7 in April. Both have rebounded unexpectedly in May, manufacturing to 53.7 and services to 58.1.

Reuters

The PMI surveys are considered sentiment indicators. Respondents are asked their characterizations of current and future business.

The US-China trade dispute has taken a toll on the optimism of business executives even as they have maintained their hiring. The deeper fall in manufacturing is logical as the sector is far more involved with overseas and China production. Services businesses tend to be local and less entangled with trade policy.

The strong return in May of the services employment index and the smaller bounce on the manufacturing side does not indicate a change in sentiment commensurate with the plunge in ADP hiring.

Initial jobless claims

Initial jobless claims have been at or close to five decade lows for the better part of two years. Claims are a reliable indicator of trouble in the labor market. They began to rise a year before the financial crisis and shot higher in June 2008 months ahead of the crash in the fourth quarter.

The 4-week moving average of 216,750 in the week of May 24^th brings the comparison back to December 1969.

Reuters

There is no indication from claims that the labor market is waning or that employers have reasons to pull back on hiring. Consider that claims are an absolute number and the current level includes the more than 60% increase in the US population since 1969.

Conclusion

Initial claims bear no sign of building problems for US employment. The decline in business hiring sentiment that has taken place over the past eight months has had little effect on actual job creation.

Business executives are clearly worried. Aside from the fact that fretting about the future is part of their job and that there are legitimate concerns from the US/China trade dispute to unsolved Brexit, political wrangling in the EU and a slowdown in global growth, none of these have prevent those same executives from continuing to seek and employ new workers.

The precipitous drop in additions to ADP payrolls tilts the NFP risk to the downside but it is most likely a single event.

Any one of the above problems may push the US and the world economy into a more serious slowdown or recession but as of the first week of June they are clouds on the horizon not a storm.