Unemployment data: How it works

MONTGOMERY - With all of the hoopla about how the state’s employment and unemployment figures are better than ever, the Journal Record decided to dig deeper into what these numbers really mean for residents.
According to Alabama Department of Labor Communications Director Tara Hutchison, the employment figures that  calculate the amount of people that are in the work force, both employed and unemployed, are tabulated by the United States Census Bureau.
“The Census Bureau calls 1,500 households in Alabama each month,” said Hutchison.
These 1,500 households are considered a “sample” of the state’s population. The bureau then uses the answers from the sample to estimate the unemployment figures for the rest of the state.
Hutchison acknowledged that such a small sample size for the state may not provide the best look at unemployment figures.
“From the sample size, the bureau extrapolates the employment figures each month. I know 1,500 doesn’t sound like a lot of people, but that’s actually more than they used to do,” said Hutchison.
The employment figures for each state are averaged together to form the national employment data.
Similarly, the 67 counties in the state average together to equal the state’s unemployment figures.
Hutchison said that the Census Bureau can poll some of the same households multiple months in a row, or change the sample population.
For county employment rates, Hutchison said the employment data was estimated based on population size for a given county.
What has appeared to be specific numbers of employed and unemployed local people could actually not reflect the real local numbers.
With only 1,500 households polled in the state, it’s possible that none or too few from the Marion County area are selected to form estimates that reflect the real employment rates.
With 67 counties, the bureau could poll 22 residents from each county. However, Hutchison said that those 1,500 residents are divided based on population size. This means that counties with larger populations (and large cities) would have more representation in the sample size than counties with smaller populations.
This does not necessarily mean that these figures are not reliable, Hutchison pointed out. It simply means that they are not exact but, due to statistical procedures, attempt to reflect the real rates as much as possible.
In addition to employment figures, the department of labor also releases the increases or decreases of wage and salary employment throughout the state each month.
“These figures are more reliable because the sample size is so much larger--20,000 Alabama businesses,” said Hutchison.
Each month, the DOL sends information requests to 20,000 businesses, which report how many jobs they support each month.
While these surveys are also divided up based on county population size, the very large number of surveys makes it much larger that Marion County is better represented in the sample.
Hutchison pointed out that the wage and salary employment numbers will not add up to the estimates provided by the census bureau.
“They will never match up. One person might have three jobs and we wouldn’t be able to see that from the data alone,” said Hutchison.


See complete story in the Journal Record.
Subscribe now!