 |
Frequently Asked Questions
Thank you for your interest in the ADP National Employment Report®.
Below you'll find frequently asked questions about the ADP report. Please
contact us with any further questions about the ADP report, or visit www.ADP.com for more information about ADP.
Background:
Q. What is the ADP National Employment Report?
Q. Why was the ADP National Employment Report
developed?
Q. What is the history of the ADP National Employment
Report?
Q. What enhancements have been made since the ADP Report's
inception?
Q. Why did ADP enhance the ADP National Employment Report
at this time?
Q. What is unique about the ADP National Employment Report
compared with other estimates?
Q. When is the ADP National Employment Report
released?
Using the ADP National Employment Report:
Q. What information is provided in the ADP National Employment
Report?
Q. How do I interpret the information in the ADP National Employment
Report?
Q. Why is another measurement needed - and is the ADP National
Employment Report in any way different?
Q. How can I use the ADP National Employment Report to
better understand the U.S. employment situation?
Q. Are monthly changes in employment forecasted by the ADP
National Employment Report indicative of changes in the ADP client base?
Privacy:
Q. Is any client or personally identifying
data used to develop the ADP National Employment Report?
Technical:
Q. How are the estimates of private nonfarm
employment shown in the ADP National Employment Report constructed?
Q. How large is the sample of payrolls on which
the ADP National Employment Report is based?
Q. How does this compare to the size of the
monthly sample of payrolls on which BLS estimates of employment are based?
Q. What procedure is used to seasonally adjust
the estimates of employment shown in the ADP National Employment Report?
Q. Are estimates of employment shown in the
ADP National Employment Report adjusted for 5 week intervals between BLS
survey periods?
Q. Do estimates of employment shown in the
ADP National Employment Report include a "net birth/death" adjustment?
Q. Are there monthly revisions to estimates
of employment shown in the ADP National Employment Report?
Q. Are there annual revisions to the estimates
of employment shown in the ADP National Employment Report?
Q. What supporting data are available to analysts?
Background:
Q. What is the ADP National Employment
Report?
A. The ADP National Employment Report
is a monthly estimate of total nonfarm private employment in the United States derived
from aggregated and anonymous ADP payroll data.
Q. Why was the ADP National Employment
Report developed?
A. The ADP National Employment Report
was developed to help meet the need for additional, timely and accurate estimates
of movements in the national labor market among economists, financial professionals,
government policymakers and academics. Because ADP pays 1-in-6 private sector employees
in the United States every pay period across a broad range of industries, firm sizes,
and geographies, it has a unique and significant perspective on the U.S. labor market.
Q. What is the history of the ADP National Employment
Report?
A.The ADP National Employment Report
was first launched on May 3, 2006 as a forecast of April 2006 total nonfarm private
employment, and was based on a subset of aggregated and anonymous payroll data that
used approximately 225,000 of ADP's more than 500,000 U.S. business clients, representing
roughly 14 million employees working in all 19 of the major North American Industrial
Classification System (NAICS) private industrial sectors.
An enhanced version of the ADP National Employment Report
was launched on March 7, 2007 to reflect the February 2007 employment situation,
based on a subset of aggregated and anonymous payroll data utilizing approximately
270,000
of ADP's more than 500,000 U.S. business clients, representing roughly
3.8 million
employees working in all 19 of the NAICS private industrial sectors. This enhanced
version of the ADP National Employment Report also provides
private employment forecasts by select industry sectors and across various company
sizes.
The latest enhancement to the ADP National Employment Report
will be launched January 7, 2009, to reflect the December 2008 employment situation,
derived from a similarly sized sample.
Q. What enhancements have been made since the
ADP Report's inception?
A. The enhanced version of the ADP National
Employment Report is based on a subset of aggregated and anonymous
payroll data that utilizes approximately
270,000
of ADP's more than 500,000 U.S. business clients, representing roughly
3.8 million
employees working in all 19 of the major North American Industrial Classification
System (NAICS) private industrial sectors. Sophisticated statistical methods, similar
to those used by the U.S. Bureau of Labor Statistics, are then used to forecast
total nonfarm private employment, as well as to provide insight into nonfarm private
employment in the goods-producing and service-providing industry sectors, the manufacturing
and finance industries, and across small, medium and large company sizes. To gain
a fuller understanding of the methodology used to develop this enhanced version
of the ADP National Employment Report and its underlying
statistical properties, please visit the "Methodology"
section of our Web site.
Q. Why did ADP enhance the ADP National Employment
Report at this time?
A. The ADP National Employment Report,
first introduced in May 2006, is jointly produced by ADP and Macroeconomic Advisers,
LLC (“MA”). Since its introduction, ADP and MA have been committed to providing
the most timely and useful estimate of U.S. nonfarm private employment possible. As
a result of this ongoing commitment, the ADP Report methodology
will be revised on January 7, 2009 for the December 2008 Report.
Q. What is unique about the ADP National Employment
Report compared with other estimates?
A. The ADP National Employment Report
is unique in that it utilizes ADP’s exceptionally large, aggregated and anonymous
sample of active payrolls from approximately
270,000
of ADP's more than 500,000 U.S. business clients as compared to other forecasts
and estimates that use surveys, job postings, and smaller payroll samples.
The ADP National Employment Report is a credible monthly
resource derived from actual payroll data, and provides insight into short-term
movements in employment across various company sizes.
Q. When is the ADP National Employment
Report released?
A. The ADP National Employment Report
is released monthly based on a pre-announced schedule.
The release dates usually fall on Wednesday, two days before the release of the
Bureau of Labor Statistics' Employment Situation Report.
Back to Top
Using the ADP National Employment Report:
Q. What information is provided in the ADP National
Employment Report? A. The
ADP National Employment Report provides the following information
for the current month and the five months prior:
1. Forecast of total nonfarm private employment in the United States.
2. Forecast of nonfarm private employment in the goods-producing industry sector
in the United States.
3. Forecast of nonfarm private employment in the service-providing industry sector
in the United States.
4. Forecast of nonfarm private employment in the manufacturing industry in the United
States.
5. Forecast of nonfarm private employment across small (1-49 payrolls), medium (50-499
payrolls) and large (>499 payrolls) company sizes in the United States.
6. Select characteristics of the ADP payroll sample used to develop the
ADP National Employment Report, including size of sample and number
of workers represented by this sample.
Q. How do I interpret the information in the
ADP National Employment Report?
A. The key forecast of total nonfarm private employment
is represented as both a monthly employment level and a monthly change in employment.
The employment level represents the total number of nonfarm private jobs in the
U.S. economy at a common point in time each month. The change in monthly employment
represents the increase or decrease in employment between these common points in
time in two consecutive months. You can observe trends in employment by simply comparing
each month's employment level change. Further detail on the construction and interpretation
of the ADP National Employment Report are available in
the Methodology section of our website.
Q. Why is another measurement needed - and is the
ADP National Employment Report in any way different?
A. The ADP National Employment Report
contains unique information about employment that is not publicly available anywhere
else. Given the importance of employment changes as an indicator of interest rates
and the overall condition of the nation’s economy, credible estimates of employment
are highly valued by economists, financial professionals, government policymakers
and academics. The ADP National Employment Report is unique
for several reasons:
It is released early in each month, two days ahead of the U.S. Bureau of Labor Statistics
Employment Situation Report, providing a timely view of monthly employment.
It is the only forecast of employment that is both timely and draws upon a large
and diverse data set derived from real data about active U.S. employees. Other estimates
rely on surveys of employment, job postings, or unemployment claims as a proxy for
employment, or use smaller samples.
It is developed using methodologies similar to those used by the U.S. Bureau of
Labor Statistics (BLS).
ADP is committed to the continued enhancement of the ADP National Employment
Report, which benefits from an ongoing and active program of collaboration
with economic thought leaders from the private and government sectors. The cumulative
intellectual contribution of these individuals enables meaningful and ongoing improvement.
Q. How can I use the ADP National Employment
Report to better understand the U.S. employment situation?
A. The ADP National Employment Report
provides a unique perspective on total nonfarm private employment, select private
industry sector employment, and private nonfarm employment by company size that
is not available from any other source. Economic practitioners and financial market
professionals recognize the value of this unique perspective and use the
ADP National Employment Report in combination with other publicly
available estimates of employment or their own proprietary models of employment.
Q. Are monthly changes in employment forecasted by the
ADP National Employment Report indicative of changes in
the ADP client base?
A. No. The methodology used to develop the
ADP National Employment Report not only uses a subset of ADP's
total payroll database, it also relies on a number of econometric techniques that
filter out anomalous data and volatility, adjust for seasonality, and align disparate
payroll period dates. In doing so, ADP's aggregate and anonymous payroll data is
developed into a unique sample of monthly employment level, and change in employment,
that is not reflective of ADP's entire client base.
Back to Top
Privacy:
Q. Is any client or personally identifying data used
to develop the ADP National Employment Report?
A. No. ADP's raw payroll data is stripped of (has
redacted from it) all client and personally identifying content before it is used
to produce the ADP National Employment Report. In no event
does ADP supply client or personally identifying information to any third parties
in connection with the ADP National Employment Report.
Back to Top
Availability of the ADP National Employment Report:
Q. Where can I go to get the ADP National Employment
Report and all additional detail?
A. We invite anyone to visit the ADP
National Employment Report Web site where they can freely access the entire Report, view commentary
and analysis of ADP National Employment Report by noted
economic analysts, and utilize a number of analytical tools. The ADP National
Employment Report is also made available through a wide variety
of outlets, including the Bloomberg service, many economic calendars such as EconoDay,
and virtually every major news outlet.
Q. Are there any other ways to access the ADP National
Employment Report and its detail?
A. Yes. Anyone can sign up for monthly e-mail delivery
of the ADP National Employment Report by visiting the
"Subscribe to the Report" section of our Web site. The e-mail
delivery of the ADP National Employment Report, and access
to all the supplemental information that supports it, are free to all via this Web site.
Back to Top
Technical:
Q. How are the estimates of private nonfarm employment
shown in the ADP National Employment Report constructed?
A. Late in each month, anonymous aggregate data
on payrolls are collected by ADP from its customers, checked for anomalies, and,
by secure means, made available to Macroeconomic Advisers for further processing.
That processing normally begins on the Monday before the monthly publication by
the BLS of the Employment Situation. Macroeconomic Advisers, in conjunction
with ADP, assigns NAICS industry codes and pay periods to as many of ADP’s payroll
records as possible. After this initial processing of the underlying data, estimates
of employment published in the ADP National Employment Report
are computed as follows.
- If employment on an ADP payroll record is for a pay period that includes the 12th
of the month, that record is used directly in subsequent calculations.
- If employment on an ADP payroll record is for a pay period after the 12th of the
month, an estimate of employment on the 12th of the month is computed as a linear
interpolation of the current and previous level of employment shown on that record,
and then used in subsequent calculations.
- Historically, if employment on an ADP record is for a pay period before the 12th
of the month, an estimate of employment on the 12th of the month is computed as
a linear interpolation of the current and following level of employment shown on
that record, and then used in subsequent calculations. This “forward interpolation”
cannot be performed in the current month, and so is the basis for revisions the
following month.
- Using the ADP data, monthly “matched-sample” growth rates of employment are computed
in 90 “cells”: 9 size classes within 10 non-farm private industries.
- The monthly growth rates of employment in each cell are seasonally adjusted using
the Census Bureau’s X-12 ARIMA program. Both the seasonal factors and the centered
moving averages on which the seasonal factors are based are estimated excluding
the effect of estimated outliers in each cell. The sample period for estimation
of the seasonal factors and the centered moving average trend is advanced each month.
- For each ADP record, a growth residual is calculated by subtracting from the growth
rate of employment on that record both the seasonal factor and the centered moving
average trend estimated for that cell.
- These growth residuals are grouped into 9 classes by size of payroll and, in each
size class, each month, are ranked in descending absolute value. For each size class,
each month, the payroll records with the largest 0.01 percent of growth residuals
(in absolute value) are discarded as cross-sectional outliers, and the matched sample
growth rates are recomputed in each of the 90 cells cell from the remaining records.
- The new matched sample growth rates are seasonally adjusted using X-12 ARIMA, with
estimated outliers in each cell replaced with the predicted value from the underlying
ARIMA model; the occurrence of estimated outliers is rare. The sample period for
the estimation of the seasonal factors (and the centered moving average) is advanced
each month.
- An additional adjustment for 5 week intervals between BLS survey dates is made by
regressing the matched sample growth rates in each cell on a dummy variable that
has a value of one in “long” months and zero otherwise. If the variable is statistically
significant, the regression is used to eliminate the effect of the long month in
that cell. Significant “long-month” effects were found in approximately 25% of the
90 cells.
- Using the seasonally adjusted data, matched-sample growth rates by industry are
computed by taking a weighted average of the matched sample growth rates by size
within each industry. The weights are based on monthly interpolations of the March
estimates of employment by industry and size from the Quarterly Census of Employment
and Wages (QCEW).
- The BLS data (as currently reported) for growth of employment by industry is regressed
on: (a) the matched-sample growth rates by industry based on the ADP data; (b) a
weighted average of the historical average growth rates of employment in each cell
based on QCEW data; (c) a weighted average of the historical average growth rates
of employment in each cell based on the ADP data; (d) lagged values of BLS estimates
of growth of employment by industry; (e) initial unemployment claims filed during
the week immediately before the week that includes the 12th of the month.
- The regressions are estimated concurrently. The coefficient on term (b) above is
restricted to unity. The coefficient on (c) is restricted to the negative of the
coefficient on term (a). This method allows different trends of employment by size
of payroll within industries, while assuming that the other industry-wide relationships
implied by the regressions hold for all size classes within an industry. Inclusion
of lagged employment in the regressions controls for shifting differences between
the BLS sample and the ADP customer base, while inclusion of initial unemployment
claims controls for differences in the definitions of employment used by BLS and
ADP.
- A level of employment is established in each cell by cumulating the predicted value
of the matched sample growth rate in each cell forward and backwards from the most
recently benchmarked March estimate of employment in that cell. Such referencing
effectively weights the growth rates of the ADP data in each cell by the observed
distribution of employment by industry and size classification.
- These levels are then summed to the aggregates by select industry and size of payroll
that are shown in the summary table of the monthly report.
Q. How large is the sample of payrolls on which the
ADP National Employment Report is based?
A. The size of the ADP sample varies moderately
from month to month, but from January to June 2008 the monthly sample averaged approximately
270,000
payrolls covering roughly
3.8 million
employees.
Q.How does this compare to the size of the monthly sample
of payrolls on which BLS estimates of employment are based?
A. BLS regularly surveys approximately 400,000 establishments
employing roughly 50 million workers. However, historically the BLS initial estimate
of employment has been based on roughly 66% of this sample (or 268,000 establishments),
the second estimate (one month later) has been based on roughly 82% of this sample
(or 328,000), and the third estimate (two months later) has been based on roughly
89% of this sample (or 356,000). Hence the ADP National Employment Report
is based on more payrolls than the BLS estimates are based on establishments. However,
the average size of the establishments surveyed by the BLS is larger than the average
size of the payrolls processed by ADP.
Q. What procedure is used to seasonally adjust the estimates
of employment shown in the ADP National Employment Report?
A. The matched sample growth rates are seasonally
adjusted using X-12 ARIMA, with estimated outliers in each cell replaced with the
predicted value from the underlying ARIMA model; the occurrence of estimated outliers
is rare. The sample period for the estimation of the seasonal factors (and the centered
moving average) is advanced each month.
Q. Are estimates of employment shown in the ADP
National Employment Report adjusted for 5 week intervals between BLS
survey periods?
A. Yes. An adjustment for 5 week intervals between
BLS survey dates is made by regressing the matched sample growth rates in each cell
on a dummy variable that has a value of one in “long” months and zero otherwise.
If the variable is statistically significant, the regression is used to eliminate
the effect of the long month in that cell. Significant “long-month” effects were
found in approximately 25% of the 90.
Q. Do estimates of employment shown in the ADP
National Employment Report include a "net birth/death" adjustment?
A. No. However, much of the BLS net birth/death
adjustment is seasonal in nature, so the seasonal adjustment procedures used in
the preparation of ADP National Employment Report implicitly
capture much of this phenomenon.
Q. Are there monthly revisions to estimates of employment
shown in the ADP National Employment Report?
A. Yes. Initial estimates shown in ADP
National Employment Report are revised one month later, reflecting
the incorporation of a relatively small number of payroll records that were not
available for use upon the initial estimate. In addition, these monthly revisions
incorporate revisions made by the BLS to its estimates of nonfarm private employment
and unemployment claims for the prior months.
Q. Are there annual revisions to the estimates of employment
shown in the ADP National Employment Report?
A. Yes. Once a year, immediately after the annual
level benchmarking of the BLS estimates of employment, the histories of all series
published in the ADP National Employment Report are revised.
These annual revisions reflect: (1) updated industry regressions based on the newly
benchmarked BLS data, including the extension of the regression sample through the
most recent benchmarked BLS data; (2) the advancement by twelve months of the reference
period used to initialize the levels of employment shown in ADP National
Employment Report; (3) the re-statement of historical seasonal
factors.
Q. How should I use estimates of the monthly growth of employment
shown in the ADP National Employment Report in conjunction
with consensus or other forecasts of growth in employment?
A. Estimates of employment in the ADP
National Employment Report are well correlated with the corresponding
BLS estimates, are available earlier, and contain information not incorporated in
other sources. Hence they can be used to improve upon forecasts of the BLS estimates
that are based on either "the consensus" or other empirical models used by analysts
to anticipate monthly changes in employment. For example, historically the estimates
of the monthly change in employment shown in the ADP National Employment Report
have been on the "right" side of the Bloomberg consensus 70% of the time, suggesting
the wisdom of adjusting one's forecast of employment in the direction suggested
by the ADP National Employment Report. In addition, a weighted
average of the monthly growth of employment shown in ADP National Employment
Report and the corresponding consensus forecast has consistently
out-performed the consensus forecast in anticipating the initial, second, third,
and current BLS estimate of the change in employment in a given month.
Q. What supporting data are available to analysts?
A. On www.adpemploymentreport.com analysts can find posted a file with the
compete histories of all series shown regularly the ADP National Employment
Report, including the current BLS estimates of the comparable
series.
Back to Top
|
|