Big data offers potentially enormous benefits for improving economic measurement, but it also presents challenges (e.g., lack of representativeness and instability), implying that their value is not always clear. We propose a framework for quantifying the usefulness of these data sources for specific applications, relative to existing official sources. We specifically weigh the potential benefits of additional granularity and timeliness, while examining the accuracy associated with any new or improved estimates, relative to comparable accuracy produced in existing official statistics. We apply the methodology to employment estimates using data from a payroll processor, considering both the improvement of existing state-level estimates, but also the production of new, more timely, county-level estimates. We find that incorporating payroll data can improve existing state-level estimates by 11% based on out-of-sample mean absolute error, although the improvement is considerably higher for smaller state-industry cells. We also produce new county-level estimates that could provide more timely granular estimates than previously available. We develop a novel test to determine if these new county-level estimates have errors consistent with official series. Given the level of granularity, we cannot reject the hypothesis that the new county estimates have an accuracy in line with official measures, implying an expansion of the existing frontier. We demonstrate the practical importance of these experimental estimates by investigating a hypothetical application during the COVID-19 pandemic, a period in which more timely and granular information could have assisted in implementing effective policies. Relative to existing estimates, we find that the alternative payroll data series could help identify areas of the country where employment was lagging. Moreover, we also demonstrate the value of a more timely series.
-
Business Applications as a Leading Economic Indicator?
May 2021
Working Paper Number:
CES-21-09R
How are applications to start new businesses related to aggregate economic activity? This paper explores the properties of three monthly business application series from the U.S. Census Bureau's Business Formation Statistics as economic indicators: all business applications, business applications that are relatively likely to turn into new employer businesses ('likely employers'), and the residual series -- business applications that have a relatively low rate of becoming employers ('likely non-employers'). Growth in applications for likely employers significantly leads total nonfarm employment growth and has a strong positive correlation with it. Furthermore, growth in applications for likely employers leads growth in most of the monthly Principal Federal Economic Indicators (PFEIs). Motivated by our findings, we estimate a dynamic factor model (DFM) to forecast nonfarm employment growth over a 12-month period using the PFEIs and the likely employers series. The latter improves the model's forecast, especially in the years following the turning points of the Great Recession and the COVID-19 pandemic. Overall, applications for likely employers are a strong leading indicator of monthly PFEIs and aggregate economic activity, whereas applications for likely non-employers provide early information about changes in increasingly prevalent self-employment activity in the U.S. economy.
View Full
Paper PDF
-
Building the Census Bureau Index of Economic Activity (IDEA)
March 2023
Working Paper Number:
CES-23-15
The Census Bureau Index of Economic Activity (IDEA) is constructed from 15 of the Census Bureau's primary monthly economic time series. The index is intended to provide a single time series reflecting, to the extent possible, the variation over time in the whole set of component series. The component series provide monthly measures of activity in retail and wholesale trade, manufacturing, construction, international trade, and business formations. Most of the input series are Principal Federal Economic Indicators. The index is constructed by applying the method of principal components analysis (PCA) to the time series of monthly growth rates of the seasonally adjusted component series, after standardizing the growth rates to series with mean zero and variance 1. Similar PCA approaches have been used for the construction of other economic indices, including the Chicago Fed National Activity Index issued by the Federal Reserve Bank of Chicago, and the Weekly Economic Index issued by the Federal Reserve Bank of New York. While the IDEA is constructed from time series of monthly data, it is calculated and published every business day, and so is updated whenever a new monthly value is released for any of its component series. Since release dates of data values for a given month vary across the component series, with slight variations in the monthly release date for any one component series, updates to the index are frequent. It is unavoidably the case that, at almost all updates, some of the component series lack observations for the current (most recent) data month. To address this situation, component series that are one month behind are predicted (nowcast) for the current index month, using a multivariate autoregressive time series model. This report discusses the input series to the index, the construction of the index by PCA, and the nowcasting procedure used. The report then examines some properties of the index and its relation to quarterly U.S. Gross Domestic Product and to some monthly non-Census Bureau economic indicators.
View Full
Paper PDF
-
Estimating A Multivariate Arma Model with Mixed-Frequency Data: An Application to Forecasting U.S. GNP at Monthly Intervals
July 1990
Working Paper Number:
CES-90-05
This paper develops and applies a method for directly estimating a multivariate, autoregressive moving-average (ARMA) model with mixed-frequency, time-series data. Unlike standard, single-frequency methods, the method does not require the data to be transformed to a single frequency (by temporally aggregating higher-frequency data to lower frequencies for interpolating lower-frequency data to higher frequencies) or the model to be restricted by frequency. Subject to computational constraints, the method can handle any number of variable and frequencies. In addition, variable can be treated as temporally aggregated and observed with errors and delays. The key to the method is to view lower-frequency data as periodically missing and to use the missing-data variant of the Kalman filter.
In the application, a bivariate, ARMA model is estimated with monthly observations on total employment and quarterly observations on real GNP, in the U.S., for January 1958 to December 1978. The estimated model is, then, used to compute monthly forecasts of the variables for 1 to 12 months ahead, for January 1979 to December 1988. Compared with GNP forecasts, in particular, for similar periods produced by established econometric and time series models, present GNP forecasts are generally more accurate for 1 to 4 months ahead and about equally or slightly less accurate for 5 to 12 months ahead. The application, thus, shows that the present method is tractable and able to effectively exploit cross-frequency sample information, in ARMA estimate and forecasting, which standard methods cannot exploit at all.
View Full
Paper PDF
-
JOB-TO-JOB (J2J) Flows: New Labor Market Statistics From Linked Employer-Employee Data
September 2014
Working Paper Number:
CES-14-34
Flows of workers across jobs are a principal mechanism by which labor markets allocate workers to optimize productivity. While these job flows are both large and economically important, they represent a significant gap in available economic statistics. A soon to be released data product from the U.S. Census Bureau will fill this gap. The Job-to-Job (J2J) flow statistics provide estimates of worker flows across jobs, across different geographic labor markets, by worker and firm characteristics, including direct job-to-job flows as well as job changes with intervening nonemployment. In this paper, we describe the creation of the public-use data product on job-to-job flows. The data underlying the statistics are the matched employer-employee data from the U.S. Census Bureau's Longitudinal Employer-Household Dynamics program. We describe definitional issues and the identification strategy for tracing worker movements between employers in administrative data. We then compare our data with related series and discuss similarities and differences. Lastly, we describe disclosure avoidance techniques for the public use file, and our methodology for estimating national statistics when there is partially missing geography.
View Full
Paper PDF
-
High-Growth Firms in the United States: Key Trends and New Data Opportunities
March 2024
Working Paper Number:
CES-24-11
Using administrative data from the U.S. Census Bureau, we introduce a new public-use database that tracks activities across firm growth distributions over time and by firm and establishment characteristics. With these new data, we uncover several key trends on high-growth firms'critical engines of innovation and economic growth. First, the share of firms that are high-growth has steadily decreased over the past four decades, driven not only by falling firm entry rates but also languishing growth among existing firms. Second, this decline is particularly pronounced among young and small firms, while the share of high-growth firms has been relatively stable among large and old firms. Third, the decline in high-growth firms is found in all sectors, but the information sector has shown a modest rebound beginning in 2010. Fourth, there is significant variation in high-growth firm activity across states, with California, Texas, and Florida having high shares of high-growth firms. We highlight several areas for future research enabled by these new data.
View Full
Paper PDF
-
Incorporating Administrative Data in Survey Weights for the Basic Monthly Current Population Survey
January 2024
Working Paper Number:
CES-24-02
Response rates to the Current Population Survey (CPS) have declined over time, raising the potential for nonresponse bias in key population statistics. A potential solution is to leverage administrative data from government agencies and third-party data providers when constructing survey weights. In this paper, we take two approaches. First, we use administrative data to build a non-parametric nonresponse adjustment step while leaving the calibration to population estimates unchanged. Second, we use administratively linked data in the calibration process, matching income data from the Internal Return Service and state agencies, demographic data from the Social Security Administration and the decennial census, and industry data from the Census Bureau's Business Register to both responding and nonresponding households. We use the matched data in the household nonresponse adjustment of the CPS weighting algorithm, which changes the weights of respondents to account for differential nonresponse rates among subpopulations.
After running the experimental weighting algorithm, we compare estimates of the unemployment rate and labor force participation rate between the experimental weights and the production weights. Before March 2020, estimates of the labor force participation rates using the experimental weights are 0.2 percentage points higher than the original estimates, with minimal effect on unemployment rate. After March 2020, the new labor force participation rates are similar, but the unemployment rate is about 0.2 percentage points higher in some months during the height of COVID-related interviewing restrictions. These results are suggestive that if there is any nonresponse bias present in the CPS, the magnitude is comparable to the typical margin of error of the unemployment rate estimate. Additionally, the results are overall similar across demographic groups and states, as well as using alternative weighting methodology. Finally, we discuss how our estimates compare to those from earlier papers that calculate estimates of bias in key CPS labor force statistics.
This paper is for research purposes only. No changes to production are being implemented at this time.
View Full
Paper PDF
-
Business Formation: A Tale of Two Recessions
January 2021
Working Paper Number:
CES-21-01
The trajectory of new business applications and transitions to employer businesses differ markedly during the Great Recession and COVID-19 Recession. Both applications and transitions to employer startups decreased slowly but persistently in the post-Lehman crisis period of the Great Recession. In contrast, during the COVID-19 Recession new applications initially declined but have since sharply rebounded, resulting in a surge in applications during 2020. Projected transitions to employer businesses also rise but this is dampened by a change in the composition of applications in 2020 towards applications that are more likely to be nonemployers.
View Full
Paper PDF
-
Starting Up AI
March 2024
Working Paper Number:
CES-24-09R
Using comprehensive administrative data on business applications over the period 2004- 2023, we study business applications (ideas) and the resulting startups that aim to develop AI technologies or produce goods or services that use, integrate, or rely on AI. The annual number of new AI-related business applications is stable between 2004 and 2011, but begins to rise in 2012 with further increases from 2016 onward into the Covid-19 pandemic and beyond, with a large, discrete jump in 2023. The distribution of these applications is highly uneven across states and sectors. AI business applications have a higher likelihood of becoming employer startups compared to other applications. Moreover, businesses originating from these applications exhibit higher revenue, average wage, and labor share, but similar labor productivity and lower survival rate, compared to other businesses. While it is still early in the diffusion of AI, the rapid rise in AI business applications, combined with the better performance of resulting businesses in several key outcomes, suggests a growing contribution from AI-related business formation to business dynamism.
View Full
Paper PDF
-
Punctuated Entrepreneurship (Among Women)
May 2018
Working Paper Number:
CES-18-26
The gender gap in entrepreneurship may be explained in part by employee non-compete agreements. Exploiting exogenous state-level variation in non-compete policy, I find that women more strictly subject to non-competes are 11-17% more likely to start companies after their employers dissolve. This result is not explained by the incidence of non-competes or lawsuits; however, women face higher relative costs in defending against potential litigation and in returning to paid employment after abandoning their ventures. Thus entrepreneurship among women may be 'punctuated' in that would-be female founders are throttled by non-competes, their potential unleashed only by the failure of their employers.
View Full
Paper PDF
-
Dynamically Consistent Noise Infusion and Partially Synthetic Data as Confidentiality Protection Measures for Related Time Series
July 2012
Working Paper Number:
CES-12-13
The Census Bureau's Quarterly Workforce Indicators (QWI) provide detailed quarterly statistics on employment measures such as worker and job flows, tabulated by worker characteristics in various combinations. The data are released for several levels of NAICS industries and geography, the lowest aggregation of the latter being counties. Disclosure avoidance methods are required to protect the information about individuals and businesses that contribute to the underlying data. The QWI disclosure avoidance mechanism we describe here relies heavily on the use of noise infusion through a permanent multiplicative noise distortion factor, used for magnitudes, counts, differences and ratios. There is minimal suppression and no complementary suppressions. To our knowledge, the release in 2003 of the QWI was the first large-scale use of noise infusion in any official statistical product. We show that the released statistics are analytically valid along several critical dimensions { measures are unbiased and time series properties are preserved. We provide an analysis of the degree to which confidentiality is protected. Furthermore, we show how the judicious use of synthetic data, injected into the tabulation process, can completely eliminate suppressions, maintain analytical validity, and increase the protection of the underlying confidential data.
View Full
Paper PDF