The Workforce Employment Dynamics data presented on this site have been developed by linking Unemployment Insurance (UI) quarterly wage records to records from other Department of Labor programs, from the Department of Motor Vehicles (DMV) and from the Quarterly Census of Employment and Wages (QCEW), a statistical program conducted in association with the U.S. Bureau of Labor Statistics. Using these data sources and linking procedures has made the preparation of Workforce Employment Dynamics data possible. However, there are certain aspects of the data that are a consequence of these sources and procedures that should be known.

As a result of the data sources, information on certain subsets of Connecticut’s working population is unavailable. While unemployment insurance covers 93% or more of the working population, there is no information in this data on jobholders that are not covered by unemployment insurance. The largest groups excluded are the self-employed and federal civilian and military employees. Further, the WED data is limited to wage earners that live and work in Connecticut. Specifically, there is no information on individuals who either live in Connecticut and work out of state or reside out of state and work in Connecticut.

The matching procedure provides demographic data on age, gender, industry of employment, and region of residence. However, there are some instances when the demographic information is not available from the sources used. The data presented here only provide information for those individuals for whom demographic information is available. This is between 75% and 80% of the population of UI covered workers mentioned above.

Since the resulting data set is a substantially large subset of Connecticut’s working population, the Office of Research has used the data to estimate the distribution of all jobs held within the state by industry, age, gender, and the region where the wage earner lives. Specifically, in each quarter, the records derived from the linkage process provide an estimate of the proportion of jobs held in each demographic category, in each industry sector and subsector, and in each region of the State. These proportions are applied to the total count of jobholders in the State, providing estimates of the counts of all jobs held by demographic category, industry sector and subsector, statewide and by region.

Final notes: It is possible for a single person to show up multiple times within the wage record database during a quarter as each wage record represents an employer-employee relationship, i.e. a job held. In other words, if a person holds multiple jobs, then he will show up multiple times in this database. Also, more than one jobholder may occupy a job during a quarter due to employee turnover. As a result, the counts of jobs held that are shown in the WED data, which are developed from the UI wage records, will exceed the monthly counts of jobs that are reported in the Labor Situation and other products, which are obtained from other establishment surveys and reports of jobs filled during the week including the 12^th of the month.

By Age		By Industry		By Region
There are three broad age categories: young workers, adults, and older workers. Young workers are broken into four age categories, adults four, and older workers two categories.		The North American Industrial Classification System (NAICS) defines the industry data. Information is available at the industry sector and subsector level.		The geographical area is the place where the individual holding a particular job resides. The five Workforce Investment Areas (WIAs) define the regions of residence.
Measures Available:
The Disparity Ratio – better known by social researchers as the 50/10 Ratio. It is calculated by dividing the 50th percentile (median) by the 10th percentile, providing a measure of the distance between them. The closer the ratio is to 1, the more similar are the values at the middle and low end of the range; the higher the ratio, the greater the difference. For example, with respect to wages, if the disparity ratio equals 2, then the wage at the 50th percentile is twice as large as the wage at the 10th. With regard to the wage distribution, researchers have used this as a measure of income inequality between middle-income and low-income workers. Since the WED data do not delineate between part-time or full-time work, and do not identify occupations, the reason for large disparity ratios are more difficult to interpret. Nevertheless, it does provide a measure of the spread between the wages at the middle and lower portion of the distribution. Press this symbol  for additional information. Sample – A sample is a subgroup that comes from the entire population. Mean – The mean is a measure of the average of the sample (e.g. the mean wage is the average wage of all jobs in the sample). Median – The median represents the number where 50% of the sample lies below and 50% of the sample lies above the number (e.g. the median wage is that wage where 50% of the sample has a smaller wage than the median and 50% of the sample has a wage larger than the median). Percentile – The xth percentile is the number where x percent has a value lower and 100 minus x percent has a value larger than the number (e.g. the wage at the 10th percentile is that wage where 10% of the sample has a smaller wage and 90% of the sample has a larger wage).