Measurement error in US regional economic data

Abstract

Estimates of publicly suppressed and unrevised regional economic data produce error and potentially bias statistical inference. This article estimates measurement error in suppressed cell estimated data sets (SCEDs) relative to unsuppressed federal administrative data. In a cross section, coefficient estimates based on relatively aggregated SCEDs are attenuated by about 31%, increasing to 50% in panel data. Coefficient estimates based on less-aggregated SCEDs are not generally reliable. A review of the limited options for identification emphasizes both the systemic limitations suppression places on research inference and the value of retrospective revisions in confidential data, highlighting the

So Close, Yet So Far: The Benefits and Limits of Rural–Urban Industry Linkages

“…(The US Census Bureau's County Business Patterns (discussed in the Methods subsection Data) underwent greater noise infusion starting in 2017, leading to unreliable estimates of employment and establishments. See also [32]). Deller et al showed that higher initial (t − 1) values lead to slower growth as places converge to their equilibrium levels of y, p, and e.…”

“…Employment data came from WholeData, a data series that estimates suppressed employment values from the US Census' County Business Patterns (CBP) [37]. Carpenter et al (2021) [32] showed that WholeData has less measurement error and thereby generally reduces regression coefficient bias over common alternatives. Table 2 presents the summary statistics for nonmetropolitan counties in 2009 and data sources for the dependent and independent variables.…”

Empirical methods in business location research

“…5 Given the unknown error with suppressed-cell estimated datasets (SCEDs), there is a history of efforts to improve SCEDs over the midpoint technique (Isserman & Westervelt, 2006;Orr & Buongiorno, 1989). Researchers who use SCEDs often do not document methods used to produce the estimates (Anselin et al, 1997;Gerking & Isserman, 1981;Isserman, 2001;Markusen, 1996), which may substantially bias coefficient estimates (Carpenter et al, 2021b). Additionally, researchers often resort to using relatively aggregated industries due to disclosure limitations, which can cause aggregation bias (Carpenter et al, 2021c).…”

“…1 Following Carpenter et al (2021b), this article uses 'establishment' to refer to a physical location or 'address' where business activity takes place, and 'firm' is used to refer to a collection of establishments under a common ownership structure. 2 'Firm demography' is variously termed industrial demography, demography of the firm or enterprise, or economic demography (McCann, 2002a;van Wissen, 2005).…”

“…2 'Firm demography' is variously termed industrial demography, demography of the firm or enterprise, or economic demography (McCann, 2002a;van Wissen, 2005). 3 Nonetheless, Carpenter et al (2021b) uses manufacturing industries to develop the 'profit pool' concept, which takes advantage of the microfoundations of profit maximization, but maintains the demand threshold framework's intuitive nature. 4 If multiple industries are included, then the Poisson estimation should include an industry indicator variable for interpretation comparability with the conditional logit model.…”