Lisa Katz| Crain’s Detroit Blog

The Workforce Intelligence Network (WIN) recently released its third quarter job demand data report and found that, while all three months of the third quarter are unprecedented and high, July is pulling the quarter over the top to a near doubling from the previous quarter.

We believe the July data is an outlier, but August and September also show very strong job demand, and the data for the fourth quarter will provide additional insight. What we can say is that October’s data also is higher than usual, so this is a trend we are watching. So far, such positive momentum, coupled with improved consumer confidence, increased labor market participation and employment, growth in GDP and other factors is pointing toward economic recovery.

With a mix of so many economic indicators to consider, why does WIN care so much about job demand, particularly in the form of online postings?

For quite some time, the WIN team has hypothesized that there is a fairly consistent relationship between online postings and actual hiring. In June 2013, there were 48,901 online job postings in Michigan. Does this mean employers had nearly 50,000 open jobs available? Not quite, but WIN researchers have been grappling with the big question: How many job postings do we need to see before someone gets a job?


The WIN team looked closely at net employment-change data and its relationship with postings for the state of Michigan as a whole. Among other things, the team began to look at online postings (the blue line in the graph below) and the six-month moving average in net employment change (dotted orange line). What emerged is a fairly obvious pattern between the two: As one increases, the other follows suit, and as one trends downward, the other does the same. This confirms that there definitely is a relationship between online postings and actual employment. Good! But how much of a relationship?

The WIN team went a step further, looking at the ratio of online postings to new employment. The analysis involved a simple math problem: postings divided by net employment change. Like the graph, employment data was lagged, meaning the team looked at employment changes one, two and three months following postings. This is because the job market is sticky and a job posting does not result in immediate hiring.

The ratio of online postings to lagged net employment change indicate that, for the initial time period analyzed (2007-13), there were approximately two online postings per net job increase in Michigan. The average ratio is near 2 whether the data was lagged at one, two or three months.

This finding is important because it suggests that while there are extreme fluctuations in the data on a monthly basis, there is a clear and stable relationship between online posting and hiring in the long run.

The majority of available data shows Michigan moving out of the recession. Most years of analysis have a ratio of postings to employment change that is close to 1.5, but 2010’s higher ratio of about four postings per hire is pulling up the average across the data set to 2.

Looking only at 2011, 2012 and to-date 2013, the ratio is closer to 1.75 postings per net job increase rather than 2. Also, the three-month lag shows an even lower ratio of 1.64 postings per net job increase. This data is evidence of a very sticky market (hiring occurs several months after postings go up) and, if the ratios continue to stabilize in post-recession Michigan, they can be used to predict long-run employment changes.

Here’s is what we can take away from WIN’s analysis:

  • Real-time job posting data has a stable long-run relationship with changes in total employment. With further research, these postings can be used as an “early warning system” for upcoming changes in the labor market.
  • More research needs to happen to focus on certain industries and occupations so more specific projections can be made for education programs and future workforce needs.
  • The data available is for Michigan’s recession recovery and Michigan is continuing to recover. As the economy stabilizes, we should start to see the ratio of postings to new jobs even out.

It looks like more postings do in fact lead to employment in the long run, but we will have to wait and see what happens as Michigan continues to move out of its employment drought. Based on what we saw in third quarter data, we very well may see a big jump in regional employment data in the next few months.

Happy job hunting!

Note: The third quarter data cited in this blog post was prepared by the WIN team’s Rebecca Cohen, Tricia Walding-Smith, Sarah Sebaly and Colby Spencer-Cesaro. The research and analysis prepared for the employment-to-postings analysis was completed by Spencer-Cesaro.

Share On