Declining Survey Response and Government Data

Key Takeaways

• CPS response rate dropped from ~90% to ~60% in a decade
• Business surveys like CES and CPI-C&S also see steep participation declines
• Administrative data offers accuracy but faces privacy and integration challenges
• Incentives such as gift cards may boost response, but cost effectiveness varies
• Survey data remains essential for measuring unemployment and forward‑looking firm expectations

Pulse Analysis

The federal statistical system has long relied on household and business surveys to calibrate macroeconomic indicators such as employment, inflation, and consumer spending. Over the past ten years, response rates have eroded dramatically— the Current Population Survey fell from nearly 90 % to roughly 60 %, while key business instruments like the Current Employment Statistics survey report comparable drops. This attrition threatens the timeliness and representativeness of core data that regulators, investors, and researchers depend on, raising concerns about bias and the reliability of policy‑making metrics.

Policymakers are turning to administrative records as a partial remedy. Tax filings, Social Security earnings, and program payment data can supply high‑frequency, granular information without the burden of respondent outreach, and they often boast superior accuracy. Yet these sources were not designed for research; mismatched classifications, delayed releases, and strict privacy safeguards limit their immediate utility. Parallel efforts to harvest price and vacancy data from web crawlers or to employ AI‑driven estimation show promise, but they still grapple with coverage gaps, quality control, and legal constraints.

To stem the decline, agencies are experimenting with modest incentives—gift cards, prize‑draw entries, and even lottery tickets—to reward participation, a tactic that can raise response rates without inflating survey length. Simultaneously, questionnaire design is being streamlined, focusing on questions that cannot be answered through administrative or digital means, such as labor‑force status and forward‑looking hiring intentions. A hybrid model that blends targeted surveys, administrative linkages, and vetted web‑derived metrics could preserve the statistical backbone while curbing costs, ensuring that policymakers retain a reliable pulse on the economy.