|
Effects of Hurricane Katrina on Local Area Unemployment Statistics
The Local Area Unemployment Statistics (LAUS) estimates for the month
of September 2005 are among the first subnational data to reflect the
impact of Hurricane Katrina, which struck the Gulf Coast on August 29,
with catastrophic effects in parts of Louisiana, Mississippi, and Alabama.
For the September 2005 LAUS estimates, BLS and its state partners have made
a number of modifications to standard estimating procedures to better reflect
the employment and unemployment situation for the month. These include:
- Overriding the built-in feature of the methodology that smoothes
over large shifts in key inputs—unemployment insurance (UI) claims
and nonfarm wage and salary employment—to the models for estimating
State labor force measures.
- Developing non-model-based LAUS estimates for the New Orleans metropolitan area.
- Not publishing labor force estimates for very small areas—for example counties
within labor market areas or cities within counties—where the quality of input
data was severely compromised by the hurricane.
Although program operations have been disrupted to some extent by the
aftereffects of Hurricane Katrina along the Gulf Coast, BLS is adhering to
its previously announced publication dates of October 21 for September Region
and State LAUS estimates and November 2 for metropolitan areas.
Below is a summary of the LAUS program. This is followed by a set of
questions and answers that provide more detailed information on LAUS
concepts, impacts on data development, and estimation procedure adjustments
taken by BLS in this unusual situation.
LAUS CONCEPTS AND DEFINITIONS
The LAUS program produces the monthly unemployment rates as well
as estimates of total employment and unemployment for approximately
7,200 areas through a Federal-State cooperative effort. LAUS areas
include Census Regions and Divisions, States, and labor market areas
that encompass the entire country. LAUS estimates are based on statistical
models of varying degrees of complexity, with the strongest models for Census
Divisions, States, and selected Metropolitan areas. Key inputs to LAUS
estimation include data from the Current Population Survey (CPS), the Current
Employment Statistics (CES) program, and Unemployment Insurance (UI) systems,
all of which have been affected by Hurricane Katrina.
The concepts and definitions underlying LAUS data come from the CPS,
the household survey that yields the official measures of the labor force
for the nation. Monthly CPS estimates are not used directly in the LAUS
program, but for LAUS estimation of States and eight substate areas
(including the New Orleans metropolitan area), the monthly CPS estimates
are inputs to models. State models are controlled in "real time" to CPS-based
Division models that sum to national monthly labor force estimates from the
CPS. For a discussion of the CPS issues associated with Katrina and the
actions BLS is taking in that program, please see
Effects of Hurricane Katrina on the Current Population Survey.
The State and metropolitan area LAUS estimates also use data from
the CES program, the payroll survey that yields estimates of nonfarm
wage and salary employment. For a discussion of the CES issues associated
with Katrina and the actions BLS is taking in that program, please see
Effects of Hurricane Katrina on the Current Employment Statistics Survey.
Unemployment insurance (UI) data are an important input to LAUS estimation,
especially at the substate level. Detailed information on the UI
statistics and LAUS estimation issues is contained in the questions and
answers below.
-
What LAUS areas are affected by Hurricane Katrina?
The States of Louisiana, Mississippi, and Alabama were directly
affected by the hurricane. Louisiana is part of the West South Central
Division, along with Texas, Oklahoma, and Arkansas. Mississippi and
Alabama are part of the East South Central Division, along with
Tennessee and Kentucky. In Louisiana, 64 parishes were determined by the
Federal Emergency Management Agency (FEMA) to be eligible for either public or public and individual assistance.
In Mississippi, 82 counties had such designation, while in Alabama, 12
counties were so identified. Within the directly affected States, BLS
publishes the following metropolitan areas and micropolitan areas in the
most heavily impacted areas (those determined by FEMA to be eligible for
individual and public assistance):
Alabama:
-
Metropolitan Areas: Mobile and Tuscaloosa
-
Micropolitan Areas: Daphne-Fairhope
Louisiana:
-
Metropolitan Areas: Baton Rouge, Houma-Bayou Cane-Thibodaux, Lafayette,
Lake Charles, and New Orleans-Metairie-Kenner
-
Micropolitan Areas: Abbeville, Bogalusa, Crowley, Hammond, Jennings,
Morgan City, New Iberia, and Pierre Part
Mississippi:
-
Metropolitan Areas: Gulfport-Biloxi, Hattiesburg, Jackson, and Pascagoula
-
Micropolitan Areas: Brookhaven, Columbus, Laurel, McComb, Meridian,
Natchez (MS-LA, MS county only), Picayune, Starkville, Vicksburg, and Yazoo City
For a complete listing of affected counties and parishes, please see
Affected area definitions.
In addition to these areas, the impact of the hurricane may be seen
in LAUS estimates for areas that are housing evacuees and people who
have resettled, in particular through the UI claims counts.
-
How are inputs to LAUS estimation impacted by Hurricane Katrina?
Decisions made for the CPS regarding September data collection and
estimation have impacted State and area CPS estimates in the affected
States. No data collection occurred in Orleans and Jefferson parishes
in Louisiana and nonresponse was high in other affected areas because
housing units in the sample were destroyed or because people had not yet
returned to their homes. Details on CPS data collection and estimation
can be obtained from Effects of Hurricane Katrina on the Current Population Survey.
The CES (payroll) data are also an input to the LAUS State models
and to metropolitan area estimates, and decisions on September data
collection and estimation impacted the affected areas.
(CES data are not used as inputs to the Division models.)
While the CES input is typically not a major contributor to the LAUS
State model employment estimate, CES metropolitan area estimates are
important in those areas. Details on CES data collection and estimation
can be obtained from Effects of Hurricane Katrina on the Current Employment Statistics Survey
The LAUS program uses the count of State UI continued claimants
without earnings from employment as an input to State and labor
market area estimates of unemployment. At the State level, the UI
claims count typically is not a major contributor to the LAUS State
unemployment estimate. At the substate level, UI data are important
to the unemployment estimate, since the data are current and pertain
to place of residence of the individual. For labor market areas, LAUS
uses the UI data based on the State and county/parish of residence that
is provided through the UI claims process. In September, States
processed claims for State UI filed by Katrina-affected individuals
in accordance with procedures established in cooperation with the
Department of Labor.
-
Do LAUS estimates reflect the geographic moves of the
population affected by Hurricane Katrina?
No. Official estimates of the population shifts associated
with the evacuation and relocation of Louisiana, Mississippi,
and Alabama residents are not available at this time. Therefore,
explicit adjustment of the LAUS State estimates to reduce population
and, in turn, labor force estimates in these three States and increase
estimates in States such as Texas, Arkansas, and Georgia where
evacuees relocated is not possible. To the extent that UI claims
counts are based on current place of residence, LAUS substate
unemployment estimates may reflect some of the movement, but these
estimates will still be controlled to the State totals.
-
Does the receipt of Disaster Unemployment Assistance (DUA)
classify a person as unemployed under the LAUS concept?
No. The qualifications to receive Disaster Unemployment
Assistance (DUA) differ from qualification from regular UI. For example,
individuals can qualify for DUA even if they are still employed
under the CPS concept. The definition of employed persons under
the Current Population Survey (CPS) includes all those who were
not working but who had jobs or businesses from which they were
temporarily absent because of bad weather, whether they were paid
for the time off or were seeking other jobs.
In addition, individuals who are receiving DUA benefits must
not have been deemed eligible for regular State UI benefits.
These individuals are outside the scope of regular UI and are
either already accounted for in the LAUS methodology, such as
unemployed individuals who have exhausted UI benefits, or they
do not meet the CPS criteria to be classified as unemployed.
LAUS ESTIMATION PROCEDURES
-
Has BLS modified its estimation procedures for affected Divisions
and States for the September 2005 estimates?
Yes. While the Census Bureau undertook extraordinary efforts
to collect data in September, the CPS estimates for the affected
Divisions and States did not show the same disaster impact as other
important economic series the nonfarm wage and salary estimate and
the count of UI claimants developed specifically for BLS/LAUS estimation.
The UI claims in Louisiana and Mississippi rose dramatically in September,
and payroll employment dropped significantly in both States. In
contrast, CPS estimates for these states showed little change from
August to September; this probably resulted from the high nonresponse
in areas affected by the storm as well as the lack of adjustment to
state population controls. BLS determined that the CES and UI movements
more accurately depicted the labor force experience in these two States,
and, consequently, in the East South Central and West South Central
Divisions. Standard LAUS estimating procedures were modified accordingly.
-
How were procedures modified in developing the labor force
estimates for Louisiana and the West South Central Division and
Mississippi and the East South Central Division?
September State models for Louisiana and Mississippi were modified
to allow the CES and UI inputs to exert more influence on the final
estimates of employment and unemployment than they typically do since,
in this situation, they were capturing more of the economic effects of
Katrina. Special variables were also used that transferred the shifts
in these two variables directly to the model estimates. In this way,
the model estimates provide a better measure of the labor force values
for these areas. The modifications for Louisiana were incorporated in
the West South Central Division estimates, along with the use of a
special variable in the Division models. The Mississippi modifications
were reflected in the estimates for the East South Central Division.
However, a special division variable was not called for in this case.
-
Will the Division estimates equal the national CPS totals of
employment and unemployment?
No. Because the estimates for the West and East South Central
Divisions were adjusted to more fully reflect the impact of Katrina
on employment and unemployment in Louisiana and Mississippi, they will
be used in developing labor force estimates that exhaust the US.
Therefore, the labor force estimates developed by summing the Division/State
LAUS estimates will not equal the official CPS national estimate.
-
How will estimates be prepared for the New Orleans metropolitan
area?
Prior to September, the New Orleans metropolitan area labor force
estimates were developed by modeling CPS data. Because Jefferson and
Orleans Parishes were under mandatory evacuation, CPS data collection
was precluded. Thus, we will not make a model-based estimate for the
New Orleans metropolitan area. We will develop an estimate based on
the standard estimating methodology that is utilized in most of the other
labor market areas in the country. Lacking Jefferson and Orleans Parishes,
we will not disaggregate the New Orleans metropolitan area into component
parishes, or develop city estimates within the metropolitan area.
-
Will other small area estimates be impacted by the hurricane?
It is possible that the procedures used to disaggregate labor market
area estimates of employment and unemployment will be distorted because
of the effect of Hurricane Katrina on local population estimates and UI
claims filing, two key components of the procedures. If this occurs,
then, while LAUS estimates will be made for the labor market area, BLS
may determine that LAUS estimates for the component counties and cities
cannot be developed. This decision will be made in consultation with
the affected State.
-
Will BLS institute other special estimating procedures for
affected areas?
Yes. Commuting data from the 2000 Census is used in the LAUS program
to adjust place-of-work employment estimates to place-of-residence.
Commuting relationships in many areas of the directly affected States
have been severely altered. In such areas, BLS intends to adjust the
commuting relationships to reflect the current situation. These
adjustments will be determined in cooperation with the affected State.
Adjustment of existing procedures for distributing unemployed new
and reentrants to the labor market also may be warranted. These
procedures rely on population distributions which may be significantly
altered. These adjustments also will be determined in cooperation with
the affected State.
-
When will these special estimation procedures be discontinued?
BLS will continue to evaluate the data used in LAUS estimation and
determine whether special procedures are required on a month-to-month basis.
Last Modified Date: October 20, 2005
|