There are many different data sets from which to view and analyze the employment picture in the US. The Non-Farm Payroll report (NFP) is one that generally attracts the most attention. It is also sometimes called the Employment Situation Report or the 'Establishment Survey'. Another series of data is presented by the Household Survey. Confused yet? Don't be.
The Establishment Survey (presented in NFP) polls 400,000 companies on how many employees they have and also covers government entities (i.e establishment). That data is then 'smoothed' based on seasonal hiring and firing patterns to provide an easily digestible series or trend for viewing employment.
The Household Survey questions 60,000 households each month on whether they have jobs and whether someone is looking for one. This data is then factored for the working population as a whole and used to calculate the unemployment rate. In the most recent NFP on Friday Sep, 6th, some commented about the weakening of the 'jobs' numbers. The employment report on the number of new hiring did indeed come in below expectations, but there are a lot of caveats that come with end of summer numbers.
The Sep NFP, which is actually data for August is heavily 'smoothed' for numerous factors such as the end of summer seasonal employment, vacations, and the return to school.
In contrast, the Household Survey Employment yielded a measure of 590,000 new jobs, far above the 130,000 reported in NFP.
Which one is correct? Both, and neither. Measuring the actual numbers in a highly fluid and mobile economy is daunting and estimating and trending is sometimes the best we can do. But, we can look at other data series to provide confirmation or non-confirmation as to whether we're seeing an outlier or a change in trend in jobs.
To really understand the 'jobs' picture, a main driver of economic growth, one can look at other data like unemployment (jobless) claims, wages, tax receipts and retail sales to give a deeper view of the employment mosaic. Weaker hiring could be temporary, or a mismatch geographically, a mismatch in labor skills, or a sign the economy is slowing. What we tend to see in an economy that is slowing is not just an slowdown in hiring, but an increase in unemployment, and that usually starts with an increase jobless claims.
As one can see in the chart below, increasing claims is usually an 'early warning sign' of a slowdown. Jobless Claims today are not just at historic lows, they've shown no sign of increasing.
Wage changes have historically been a good sign of a trend change in the economy and as we see from the most recent numbers average hourly earnings on a month over month and year over year basis are the best in a decade or so. Retail Sales have been strong and they can confirm the wage increases. In the chart below, Retail Sales are usually driven by wage increases and tend to rollover even prior to the rollover in income. While the two are not perfectly correlated, the cause and affect relationship between wages and spending is logical and a slowdown in either Retail Sales or wage growth would be reason for concern, but we've seen a slowdown in neither.
So, which employment numbers to believe? Actually, all of them because they should be viewed together to help us understand the jobs mosaic they create. We have a very low unemployment rate but have seen some seasonally adjusted numbers that show slower job growth. Since 2010 we've averaged around 200,000 new jobs per month, since 1950 there is no other period like that (or prior to I suspect). Yet, we're seeing no increase in new unemployment claims (firings) and job openings (reported in JOLTs) are still high (higher than the number of available workers). This has helped the wage growth that is leading to increased retail spending. No where in those numbers is the weakness that leads to recession.
We may have reached a point where the slowdown in new hires is due to a 'mismatch' in skills or geography that can hopefully be resolved by drawing in new or previously disheartened workers (who had stopped looking for work). This is not a bad problem to have, in fact with some luck and/or creativity, the long term unemployed may take on new opportunities that will also increase tax revenues, lower government spending and even strengthen programs like Social Security.