Towards Unrestricted Public Use Business Microdata: The Synthetic Longitudinal Business Database

In most countries, national statistical agencies do not release establishment‐level business microdata, because doing so represents too large a risk to establishments’ confidentiality. One approach with the potential for overcoming these risks is to release synthetic data; that is, the released establishment data are simulated from statistical models designed to mimic the distributions of the underlying real microdata. In this article, we describe an application of this strategy to create a public use file for the Longitudinal Business Database, an annual economic census of establishments in the United States comprising more than 20 million records dating back to 1976. The U.S. Bureau of the Census and the Internal Revenue Service recently approved the release of these synthetic microdata for public use, making the synthetic Longitudinal Business Database the first‐ever business microdata set publicly released in the United States. We describe how we created the synthetic data, evaluated analytical validity, and assessed disclosure risk.     

Re-identification in the Absence of Common Variables for Matching

A basic concern in statistical disclosure limitation is the re-identification of individuals in anonymised microdata. Linking against a second dataset that contains identifying information can result in a breach of confidentiality. Almost all linkage approaches are based on comparing the values of variables that are common to both datasets. It is tempting to think that if datasets contain no common variables, then there can be no risk of re-identification. However, linkage has been attempted between such datasets via the extraction of structural information using ordered weighted averaging (OWA) operators. Although this approach has been shown to perform better than randomly pairing records, it is debatable whether it demonstrates a practically significant disclosure risk. This paper reviews some of the main aspects of statistical disclosure limitation. It then goes on to show that a relatively simple, supervised Bayesian approach can consistently outperform OWA linkage. Furthermore, the Bayesian approach demonstrates a significant risk of re-identification for the types of data considered in the OWA record linkage literature.     

Choice of functional form for hedonic price equations

The appropriate functional form for a hedonic price equation cannot in general be specified on theoretical grounds. In this paper, a statistical procedure for the choice of functional form is proposed. A highly general functional form is specified that yields all other functional forms of interest as special cases. Likelihood ratio tests are used to test the appropriateness of alternative forms. The procedure is illustrated using cross section microdata for housing. For the case considered, the functional forms most commonly used in previous studies are strongly rejected.     

Disclosure protection of microdata: problems and solutions

National statistical offices meet an increasing demand for the dissemination of data sets containing individual data on respondents. This trend may cause problems with respect to the care taken not to disclose data about individuals. Therefore the risk of disclosure of each microdata set to be released should be assessed. If this risk is too high, measures have to be taken to protect the data set. This contribution describes the disclosure problem, and explains why it is a real problem. Using the concept of uniqueness, some theory is developed which may help to establish the risk of identification. It turns out that useful microdata sets can only be released if some of the disclosure risks are dealt with by legal arrangements, rather than by restrictions on the data to be released.     

Strategies for measuring risk in public use microdata files

Statistical agencies have the responsibility to design data release strategies which will not violate pledges of nondisclosure either through intent or neglect. In addition to ethical and legal concerns, statistical offices must be mindful that violating pledges of confidentiality may undermine an agency's ability to collect data due to loss of public trust. Statistical organizations also have the obligation to make information available to a variety of individuals and institutions to allow for informed discussion from differing perspectives on a range of issues. However, it is through fine, accurate detail on a file that risks of disclosure arise. In this report we discuss strategies for controlling risk in the release of public use microdata files. The equivalence class structure of a microdata file is defined and we show how the classic entropy function can be employed on the equivalence class structure to provide a measure of relative risk.     

A METHOD TO CREATE CARBON FOOTPRINT ESTIMATES CONSISTENT WITH NATIONAL ACCOUNTS

Although multiregional input–output (MRIO) databases use data from national statistical offices, the reconciliation of various data sources results in significantly altered country data. This makes it problematic to use MRIO-based footprints for national policy-making. This paper develops a potential solution using the Netherlands as case study. The method ensures that the footprint is derived from an MRIO dataset (in our case the World Input–Output Database (WIOD)) that is made consistent with Dutch National accounts data. Furthermore, usage of microdata allows us to separate re-exports at the company level. The adjustment results in a foreign footprint in 2009 that is 22% lower than the original WIOD estimates and a significantly altered country allocation. We demonstrate that already in the data preparation phase due to the treatment of re-exports and margins, large differences arise with Dutch national statistics, which may help explain the variation in footprint estimates across MRIO databases.     

Hospital Reimbursement Incentives: An Empirical Analysis

Reimbursement systems for health-care providers are very complex, like the production systems that they regulate. This complexity has led to some important misperceptions about the incentive consequences of major reimbursement reforms. One example is the prospective payment system (PPS), developed to provide "high-powered" incentives through fixed prices for hospital admissions for the US elderly. In fact, various features of the DRG system allow reimbursement to vary with actual treatment decisions during an admission, and so are not prospective. This paper develops a general method for measuring actual reimbursement incentives in complex regulated price systems. The method uses regression techniques with variance decompositions to quantify the effects of particular features of the payment system on prospective and retrospective cost sharing, as well as overall generosity of payments. I apply this method to microdata on 20 percent of Medicare hospital admissions in 1987 and 1990 to summarize the incentives created by PPS in practice, and how the incentives are evolving over time. I show that PPS involves limited and decreasing cost sharing with hospitals, most of which is not prospective. The reimbursement incentives vary substantially across diagnoses, demographic groups, and types of intensive treatments, possibly with important implications for hospital behavior and medical expenditure growth. The techniques developed here can be used to analyze a broad range of provider reimbursement mechanisms.     

Labour supply in Italy: an empirical analysis of joint household decisions,with taxes and quantity constraints

This study applies an econometric framework which allows for complex non-convex budget sets, highly non-linear labour supply curves and imperfect markets with institutional constraints. A married couple's version of the model is estimated on Italian microdata. The empirical results show that male labour supply is rather inelastic while labour supply among females, especially participation, is considerably more elastic. The elasticities depend strongly on household income. The largest elasticities are found for females living in poor households. The results of the tax simulations suggest that there are only modest labour supply responses from replacing the 1987 system by proportional taxes. Copyright © 1999 John Wiley & Sons, Ltd.     

Accounting for labor demand effects in structural labor supply models

When assessing the effects of policy reforms on the labor market, most studies only focus on labor supply. The interaction of supply and demand is not explicitly modeled, which might lead to biased estimates of potential labor market outcomes. This paper proposes a straightforward method to remedy this shortcoming. We use information on firms' labor demand behavior and feed them into a structural labor supply model, completing the partial analysis of the labor market on the microdata level. We show the performance and relevance of our extension by introducing a pure labor supply side reform, the workfare concept, in Germany and simulating the labor market outcome of the reform. We find that demand effects offset about 25% of the positive labor supply effect of the policy reform.     

Do macroeconomic forecasters use macroeconomics to forecast?

Faced with an uncertain future, forecasters often rely on textbook relationships to build a coherent narrative for their macroeconomic forecasts. We focus on two cornerstones of modern macroeconomics – Okun’s law and the Phillips curve – and examine whether or not professionals forecast in a way that is consistent with these. Using microdata from the US, Euro Area, and UK surveys of professional forecasters, we examine forecasts over the period 1981-2017 at the level of the individual and across different time horizons. Our findings show that the majority of forecasters produce their forecasts in a manner that is consistent with macroeconomic theory.     

Learning from Trade through Innovation

This paper explores the idea that firms learn from trade by introducing either new products or processes influenced by their trade links with foreign markets. By exploring microdata for Spain, including data on innovation and trade, we find a robust relationship between imports, exports and innovation. The results suggest that firms learn primarily from import links, which enable them to innovate and to ‘dress up’ for starting to export. This sequencing between trade and innovation, however, is shown to be more pronounced for small firms only and technologically advanced firms.     

Bootstrap inference using estimating equations and data that are linked with complex probabilistic algorithms

Probabilistic record linkage is the act of bringing together records that are believed to belong to the same unit (e.g., person or business) from two or more files. It is a common way to enhance dimensions such as time and breadth or depth of detail. Probabilistic record linkage is not an error-free process and link records that do not belong to the same unit. Naively treating such a linked file as if it is linked without errors can lead to biased inferences. This paper develops a method of making inference with estimating equations when records are linked using algorithms that are widely used in practice. Previous methods for dealing with this problem cannot accommodate such linking algorithms. This paper develops a parametric bootstrap approach to inference in which each bootstrap replicate involves applying the said linking algorithm. This paper demonstrates the effectiveness of the method in simulations and in real applications.     

Topics of statistical theory for register‐based statistics and data integration

Official statistics production based on a combination of data sources, including sample survey, census and administrative registers, is becoming more and more common. Reduction of response burden, gains of production cost efficiency as well as potentials for detailed spatial‐demographic and longitudinal statistics are some of the major advantages associated with the use of integrated statistical data. Data integration has always been an essential feature associated with the use of administrative register data. But survey and census data should also be integrated, so as to widen their scope and improve the quality. There are many new and difficult challenges here that are beyond the traditional topics of survey sampling and data integration. In this article we consider statistical theory for data integration on a conceptual level. In particular, we present a two‐phase life‐cycle model for integrated statistical microdata, which provides a framework for the various potential error sources, and outline some concepts and topics for quality assessment beyond the ideal of error‐free data. A shared understanding of these issues will hopefully help us to collocate and coordinate efforts in future research and development.     

Disclosure risks for microdata

In this paper a model is developed for assesing disclosure risks of a microdata set. It is an extension of the one presented in Bethlehem et al. (1988). It is used to calculate (an upper bound of) the risk that an investigator is able to reidentify at least one individual in an anonimyzed data set, and hence discloses some sensitive information about him. This risk is shown to depend on, among other things, two variables which can be controlled by the statistical office which is disseminating such a data set: the 'coarseness' of the key variables and the size of the data set. The model yields guidelines as to the usage of these two instruments to control the disclosure risk.     

The Effects of Union and Nonunion Forms of Employee Representation on High‐Performance Work Systems: New Evidence from French Microdata

This article investigates the effects of union and nonunion employee representation (ER) on the use of high‐performance work systems (HPWSs) in the French context. We use microdata from a nationally representative survey (REPONSE 2010–11) and estimate models dealing with the potential endogeneity of ER. After controlling for endogeneity and for a range of workplace characteristics, regression analyses suggest that neither union nor nonunion representatives are inherently against the use of HPWSs. Moreover, these forms of ER cannot be regarded as substitutes for one another. Results vary, depending on what type of bundle of practices is considered. © 2016 Wiley Periodicals, Inc.     

Brighter prospects? Assessing the franchise advantage using census data

This paper uses Census microdata to examine how starting a business as a franchise rather than an independent business affects its survival and growth prospects. We assess factors that influence the decision to become a franchisee and use various empirical approaches to correct for selection bias in our performance analyses. We find that franchised businesses on average exhibit higher survival rates than independent businesses; but importantly, the difference is small compared with claims in the trade press. The effect is also short lived: conditional on surviving a year or two, we no longer find survival (or growth) differences. We then explore two potential sources for this small survival advantage, namely franchisors’ screening process and the benefits arising from the brand and business know‐how provided by franchisors. We find evidence that both of the sources contribute to the franchising advantage.     

Incentives to innovate in response to competition: The role of agency costs

This paper investigates the role of managerial ownership and incentive payment as potential drivers of innovation decisions by firms and as shifters of the competition-innovation link in the Russian manufacturing industry, where poorly protected property rights and a path-dependent market structure (typical for many transition economies) lead to a variety of outcomes. We use recent survey-based microdata for nearly 2000 non-listed companies in Russia. Our results suggest that managerial ownership, which initially evolved as a means of protecting against and resisting dysfunctional institutions, may stimulate decisions to undertake R&D and risky product innovations. Further, managerial ownership and competition are complementary motivations for R&D and innovation. Incentive payment to hired managers is a positive commitment instrument but has no impact on the competition-innovation link.     

What drives racial segregation? New evidence using Census microdata

Using restricted Census microdata that link households to the Census block in which they live, this paper re-examines the question of whether racial differences in sociodemographic characteristics can explain observed levels of racial segregation. We develop a simple measurement framework designed to make use of the rich joint distribution of individual and neighborhood characteristics that these data provide, analyzing segregation patterns in the San Francisco Bay Area. The results indicate that racial differences in the collective set of characteristics we consider do have the potential to explain a considerable amount of the observed segregation, although more so for Asians and especially Hispanics than whites and blacks. Different sociodemographic factors emerge as potentially important for each race.     

Decentralized environmental regulations and plant‐level productivity

Using the framework provided by the Porter hypothesis, we study the impact of environmental regulations and enforcement policies on plant‐level green total factor productivity (TFP) growth and its components related to efficiency change and technical change. The detailed microdata we use are from Sweden and for the pulp and paper industry. This industry is the source of significant amounts of water and air pollution and is one of the most heavily environmentally regulated manufacturing industries. Sweden has a unique decentralized regulatory structure where the manufacturing plants have to comply with plant‐specific regulatory standards stipulated at the national level, as well as decentralized local supervision and enforcement. Our empirical results point to beneficial impacts of the environmental policies on plants' green TFP growth and sustainable production practices. We also find that political economy considerations are important, as the presence of the Green Party and aspects like plant size (with corresponding local and regional economic effects) matter in enforcement of the standards.     

Economic insecurity,conservatism, and the crisis of environmentalism: 30 years of evidence

There is an ongoing scientific debate about how environmental concern develops in a population, and under which circumstances it might decline at some point. In this paper, by analysing thirty years of microdata from the German Socio-Economic Panel (SOEP), I investigate the role of socioeconomic factors and political preferences in altering and addressing environmental perceptions in Germany, Europe's green leader. Results from correlated random-effects models reveal, inter alia, that economic insecurity and support for right-wing political parties significantly contribute to lowering environmental concern. Hence, policymakers ought to focus on economic insecurity in order to increase citizens' environmental concern and, eventually, bring an enhancement in pro-environmental behaviour as a result of the voters' political will.