Reclaiming Common Sense

  The Tale of Two Data Sets May Come to and End this May

Expecting Strong Worker Growth and Strong Jobs Growth.

This year has been the Tale of two data sets, the Current Population Survey (CPS) jobs and unemployment data and the Current Employment Statistics (CES) Wages and Worker Data. The disagreement between the CPS and CES data was detailed in the May Jobs Report Forecast Article. There has been some disagreement between the ADP Payroll Report and the "Jobs Report." The February Payroll Report was strong while the February Private sector data  was weak. 

We have seen weakness in the May SA CES data when we have a weak economy. It is important to remember that the May 2016 SA CES private sector worker data went negative during the June and July Job Reports (Summary Table B for each report.) This was covered during 2016 by this author. The annual revisions that happen with the release of the January  Jobs Report pushed the May 2016 positive - it now stands at 8,000 workers added during May 2016. This Wednesday the ADP data was weak. What happened with the May Jobs Report?

This May we saw 778,000 workers added to the economy. Last year we saw over 1 million non-seasonally adjusted workers join the workforce. There are a couple of ways at looking at this. This is more than were added during recessionary years during the month of May. This is better than we saw during May 2010 and May 2011, as well as May 2016, during expansionary periods. The month to month NSA CES data recorded expansion in all sectors except IT, Education and Health Services (EHS,) and Government. This was a possibility. All sectors grew annually from May to May except IT. This, too, was a possibility. 

The Seasonally Adjusted data only grew by 90,000 Private Sector Workers. This is better than the recessionary years and better than 2016. Somehow, the seasonally adjusted data for 2010and 2011 were higher than this year. This is not the first time that we have seen strong NSA CES worker data reported weaker than weaker NSA CES data. The month to month and May to May story was similar with the seasonally adjusted data as the non-seasonally adjusted data. The largest month to month growth was found in Leisure and hospitality (LAH,) Professional Business Services (PBS,) Mining and Logging, EHS, and Construction. This makes sense based on recent JOLTS job openings, quits, separations, and hiring data. May to May we only saw a decline in IT. The Largest gains were in Mining, Construction, LAH, EHS, and PBS.

The Private sector only grew at 1.72, unadjusted, and 1.79% seasonally adjusted. The unadjusted annual CES Growth rate was been plus or minus 2.00% since September of last year. This was a disappointment. The month to month growth is expected in they 2011, 0.80% to 0.85% range, possibly 0.87%.. This number came in at 0.61% This was also a disappointment. It was not as bad as May 2016 when the economy grew at 0.57%

The seasonal factor impacts the final SA CES private sector worker number. The seasonal factor spiked this month, which boosted the weaker NSA CES worker data. If the seasonal factor last year was used then we would have only record 3,000 net workers added. 

What happened with the revisions?  The data for March and April were revised downward from their last values. This is similar to what happened duringMay 2016

when the April data was revised down by 37,000, allowing the May data to originally be reported at 38,000 seasonally adjusted non-farm payroll.

The Current Population Survey Data indicated the potential for a  strong spike in full-time jobs and a drop in part-time jobs, non-seasonally adjusted. While the full-time jobs number did spike and the part-time jobs number did fall, the spike in full-time jobs was not as good as it was during May 2017 and May 2018. We are going through a general replacement cycle of FT jobs for PT jobs.

We saw seasonally adjusted part-time growth and seasonally adjusted full-time jobs contract. It was possible that we could see strong Non-Seasonally Adjusted CPS growth and report weak growth, or even a contraction, non-seasonally adjusted.

Continuing Claims data fell Mid-April to Mid- May - Did U-3 Unemployment Fall? No. Both the NSA and the SA U-3 unemployment number rose. NSA U-3 unemployment rose from 5.387 million to 5.503 million. This is the second lowest May level since 1979. Only May 2000 was lower than this May. The SA U-3 rose from 5.824 million to 5.888 million.

The official unemployment level is the number of people out of work who are looking for work. The continuing claims data from the weekly unemployment report counts those who had been working, lost their one and only job, were eligible for benefits, and who are claiming benefits. The continuing claims data mid-May dropped from Mid-April. The continuing claims data is different from the U-3 unemployment data just as the CES data for workers is different from the CPS data for jobs and unemployment. 

What does all this data mean to the worker and wages situation?  We were anticipating seeing strong month to month and May to May growth in Construction, Mining/Logging, and Professional Business services, non-seasonally adjusted. These are some of the highest paying sectors. We were anticipating strong May to May and April to May seasonally adjusted worker growth in Education and Health Services and the  Leisure and Hospitality sectors, two of the lowest paying sectors. We could see a spike in Non-seasonally adjusted workers and wages masked by weaker seasonally adjusted worker and wage increases and a heavier bias on the seasonally adjusted EHS and LAH data. This will be covered in more detail in the wage and worker article.

What does the data men to men and women? This column addresses the changes in FT and PT jobs for men and women. Men work more FT jobs than women. Women work more PT jobs than men work. This means that as we see FT jobs spike and PT jobs fall that Men may be finding work while women are shedding jobs. It also means that the CES data regarding construction, manufacturing, and Mining/Logging may funnel more jobs to men than towards women.

What does the data mean regarding participation? The non-seasonally adjusted CPS jobs and unemployment jobs point toward a spike in jobs greater than the drop in unemployment, so NSA participation and U-3 unemployment rates should fall. The SA CPS data points toward a net gain or net loss of seasonally adjusted combined jobs and a net gain or loss of unemployed workers. The seasonally adjusted data is a crap shoot. It was expected that we could  see the seasonally adjusted participation rate reported as stagnating while it is actually growing. The addition of unemployed workers and full-time jobs were greater than the part-time job losses. The NSA Participation rate grew from 62.66% to 62.80% while the NSA U-3 unemployment rate ticked up from 3.32% to 3.38%

The Seasonally Adjusted Participation rate and unemployment rate were virturally unchanged. The SA U-3 rose from 3.58% to 3.62% while the participation rate edged lower from 62.82% to 62.80%

The data has been all over the place. We have had the best full-time job creation during the first 27 months of President trump's term than w. hat was measured under former Presidents Reagan, Clinton, George W Bush and Obama combined. President Trump and former President Clinton were the only two Presidents to reduce unemployment and increase full-time jobs during their first 27 months in office. This is month 28. If we see a spike in FT jobs, as expected, we will be climbing the jobs mountain through July and possibly August.

If wage increase and workers increase, as expected, net wages will grow by a multiple of the worker growth and the wage growth. If net wages increase dramatically then spending should increase and retail sales should increase. If retail sales increase then more retail workers will be hired. It takes a full-time permanent job to obtain a mortgage. If we have record full-time May workers this month then we can see new and existing home sales improve over the next few months, and over the next couple of years. This could also spur more construction and more construction jobs.  Wages did rise. Workers rose. This should doubly boost the total earnings picture. This will be covered in another article.

As employment increases, full-time jobs increase, and unemployment falls, then Gross Domestic Product should improve quarter to quarter and from 2018Q2 to 2019Q2. If GDP is reported higher than first quarter, and if it equals or exceeds 2018Q2, the headlines will be "Trump's Teflon Economy." Success will beget success. Jobs will continue to grow in real, non-seasonally adjusted, terms. The seasonally adjusted data may continue to grow for years, depending upon the seasonal factors used.

A Tale of Multiple Data Sets. This article started with the discussion of the "Tale of Two data sets." This month it is the tale of the seasonal factors and multiple data sets. The headline data is "always" seasonally adjusted. Would people believe that 1 million workers joined the workforce during May, or that 1.5 million full-time jobs were created during May? Remember, The CES worker and wage data is different from the CPS job and unemployment data. Remember that the JOLTS job opening data trails the Jobs report data by a month, and that it resembles the CES worker data as a fraternal twin resembles its twin. The JOLTS data is more like a cousin to the CPS data, part of the family tree, different parents. The CPS unemployment data is even further removed on the family tree from the CPS unemployment data.

We are still climbing the jobs mountain. We have 6.8 million more full-time jobs than we had during July 2007 at the pre-recession peak of full-time jobs. We also have 3.558 million more part-time jobs. We have 1.045 million more full-time jobs than we had last May. We also have 97,000 more part-time jobs than we had last May. "All recessions are job recessions." That has been said in this column in many articles. As long as the same month data is better than the prior year we are expanding.

It's the Economy.