“Labor mismatch, also known as structural imbalance, can be defined as a poor match between the char- acteristics of unemployed workers and those required for vacant jobs” write
Maria E. Canon, Mingyu Chen, and Elise A. Marifian in
Labor Mismatch in the Great Recession: A Review of Indexes Using Recent U.S. Data *
In the wake of the jobless recovery from the Great Recession, economists have sought to explain the coexistence of a high unemployment rate and increasing job openings as a mismatch phenomenon. This article reviews five studies that have contributed to the development of mismatch indexes and computes the corresponding indexes over the period May 2005–May 2012 using job vacancy data from the Conference Board Help Wanted OnLine® (HWOL) Data Series. For most of the indexes, mismatch increased during the Great Recession, although the indexes exhibit a range of behaviors. According to an index developed in Jackman and Roper (1987), mismatch can account for at most 2.72 percentage points of the 5.30-percentage-point increase in the unemployment rate from the beginning of the recession to the unemployment rate peak.
The counterfactual unemployment rates
In addition to computing the indexes for industrial and occupational mismatch, the authors use ISSTV1 to compute the counterfactual unemployment rate (the unemployment rate that prevails when there is no mismatch). Following Şahin et al. (2012), they calculate the counterfactual unemployment rate, which, when subtracted from the actual unemployment rate, indicates how much of the recent rise in the unemployment rate can be attributed to mismatch.
Figures 4 and 5 show the actual unemployment rate plotted with our two calculations of the counterfactual unemployment rates, which use (i) industries disaggregated into 12 and 19 cate- gories and (ii) occupations disaggregated into 10 and 22 categories. The spread between the actual and the counterfactual unemployment rates is always greater than zero, even when the unemployment rate is low, indicating that mismatch unemployment exists throughout the sample. The counterfactuals tend to follow the path of the actual unemployment rate, with fluctuations appearing to lag the actual rate slightly. Figure 4 shows that the lesser-disaggregated indexes appear superimposed throughout the majority of the sample; alternatively, with the greater level of disaggregation, the industrial counterfactual unemployment rate generally falls at higher val- ues than those for the occupational rate (see Figure 5).
* Chosen excerpts by Job Market Monitor