New research series on labor force status flows from the Current Population Survey

Each month, the Bureau of Labor Statistics (BLS) reports the number of people employed, unemployed, or not in the labor force as estimated from the Current Population Survey (CPS). The net changes in the number of people employed and unemployed from month to month are important gauges of the health of the U.S. job market. Underlying these relatively small net changes, however, is a great deal more churning. Millions of individuals move between employment and unemployment each month, and millions of others enter or leave the labor market. In addition, people move into and out of the survey universe of the civilian noninstitutional population age 16 and over; each month, for example, some people turn age 16. Information on these labor force status flows (commonly referred to as gross flows) can provide additional insight into current labor market developments and help analysts pinpoint the nature of the net changes in employment and unemployment.

The CPS can provide labor force flow data because households are interviewed for several months in a row. In any month, about three-quarters of the households in the survey also would have been interviewed the previous month. This month-to-month overlap allows the BLS to track individuals who change labor force status from one month to the next.

In a given month, a person is in one of three labor force states: employed (E), unemployed (U), or not in the labor force (N). The following month, the person could either have the same status or change to one of the other two states. Thus, one can express the complete set of labor force flows with the following 3 x 3 matrix.

Labor force status flows

	Status in current month
Status in prior month	Employed	Unemployed	Not in labor force
Employed	EE	EU	EN
Unemployed	UE	UU	UN
Not in labor force	NE	NU	NN

The notation of the matrix is such that the first letter of each flow denotes the labor force status of an individual in the previous month, and the second letter of each flow denotes the state of an individual in the current month. The diagonal elements (i.e., EE, UU, NN) represent individuals who did not change their labor force status over the month. (Not shown in this matrix are flows into and out of the civilian noninstitutional population age 16 and over.)

The flows into employment, listed in the first column, represent all individuals who remained employed, but not necessarily with the same employer, over the month (EE), the number of unemployed persons who became employed (UE), and the number of persons previously not in the labor force who became employed (NE). The flows into unemployment, listed in the second column, represent the number of employed who become unemployed (EU), the number of unemployed who remained unemployed from the previous month (UU), and the number of individuals not in the labor force who became unemployed (NU). Finally, the flows out of the labor force, listed in the third column, represent the number of previously employed individuals who leave the labor force (EN), the number of previously unemployed individuals who leave the labor force (UN), and the number of individuals who remained out of the labor force (NN).

Labor force flow data have been available from the CPS for decades; however, discrepancies existed between the labor force changes derived from the flows and the net changes derived from the monthly stock estimates. For this reason, the flow data have not been published on a continuous basis since 1952. To make the data more useful, BLS researchers developed methods to reconcile the flows and stock data and to seasonally adjust the flows. For more information on the development of the improved labor force status flows, see Harley J. Frazis, Edwin L. Robison, Thomas D. Evans and Martha A. Duff, Estimating gross flows consistent with stocks in the CPS, Monthly Labor Review, September 2005, pp. 3-9.

These research labor force flows will allow data users to analyze the movements that underlie the net over-the-month changes in employment, unemployment, or not in the labor force. For example, examining the flows would show whether a drop in unemployment over the month was due to an increase in the number of unemployed people becoming employed (the UE flow) or to an increase in the number of unemployed people leaving the labor force entirely (the UN flow). The historical series on labor force flows also allow for longer-term analysis which should be useful for business cycle comparisons. During a labor market slowdown, for example, the number of people remaining unemployed from one month to next (the UU flow) will increase.

The labor force flows can provide information from different perspectives. The examples above focused on examining the data in terms of the current month, that is, from which labor force status individuals came, often referred to as inflows. However, assessments also can focus on what happens to persons in a particular labor force category, such as the flow of individuals out of unemployment. This latter type of assessment (using total unemployment as the base) sheds light concerning the likelihood of individuals leaving their current state of unemployment and finding employment, remaining unemployed, or leaving the labor force. For more information on the usefulness of labor force status flows in analyzing changes in labor force measures, see Randy Ilg, Analyzing CPS data using gross flows, Monthly Labor Review, September 2005, pp. 10-18.

These new research labor force flow estimates will now be made available through the BLS website.¹ Three products will be updated each month. The first table (cps_flows_current.pdf) shows all the labor force status flows that are developed under the new methods. This includes the 9 core flows that track the movement of people between employment, unemployment, and out of the labor force. For example, the table shows the number of people who moved from employment to unemployment and the number who moved from unemployment to employment. In addition to the core flows, the table also shows the flows of people into and out of the scope of the survey, that is, the movement of people into and out of the civilian noninstitutional population age 16 and over. This table allows data users to see how the sum of all the flows equal the levels of employment, unemployment, and not in the labor force for both the current and previous month. The second table (cps_flows_recent.pdf) shows the flows into the three major labor force categories (employment, unemployment, and not in the labor force) for recent months, allowing users to see how the flows have changed over those months. The third product is a data file (cps_flows_history.pdf, cps_flows_history.txt, cps_flows_history_format.txt ) containing the time series of the major labor force status flows from February 1994 through the most recent month available.

Effective May 2, 2008, labor force status flows series derived from the Current Population Survey were extended to include the February 1990 to January 1994 period. In addition, the seasonally adjusted series from February 1994 to February 2008 have been revised to incorporate minor adjustments in the seasonal adjustment methodology.

Due to the major redesign in the Current Population Survey in January 1994, labor force flows for that month could not be estimated in the usual manner. A change in the way that households were identified in the CPS microdata made it impossible to link households from December 1993 to January 1994; thus all values for labor force flows (including individuals turning 16) were unavailable. In order to generate unadjusted estimates for January 1994, statistical models were fit to each of the existing historical series. The resulting estimated values were then used in the seasonal adjustment process to produce seasonally adjusted series back to February 1990.

For more information about the new research series on labor force flows and the new data products, contact the Division of Labor Force Statistics at 202-691-6378 or by email. 1. These estimates will differ slightly from those previously published in research articles due to updated population controls and revised seasonal adjustment factors.

 

Last Modified Date: May 2, 2008