Impacts of the COVID-19 Pandemic on Labor Market Surveys at the German Institute for Employment Research

Joseph W. Sakshaug, Jonas Beste, Mustafa Çoban, Tanja Fendel, Georg‐Christoph Haas, Sebastian Hülle, Yuliya Kosyakova, Corinna König, Frauke Kreuter, Benjamin Küfner, Bettina Müller, Christopher Osiander, Silvia Schwanhäuser, Gesine Stephan, Ehsan Vallizadeh, Marieke Volkert, Claudia Wenzig, Christian Westermeier, Cordula Zabel, Stefan ZinsMADOC 2020 · DOI

Abstract

The COVID-19 pandemic and associated government lockdown restrictions have fueled a high demand for survey data on how individuals and establishments are coping with the restrictions. However, the pandemic has also dramatically affected surveys themselves, forcing research institutes to adapt their fieldwork operations to the uncertain and evolving conditions. This paper documents the specific impacts of the pandemic on several ongoing surveys at the Institute for Employment Research in Germany, and describes how these impacts have been addressed. In addition, this paper summarizes efforts to rapidly develop new Covid-19 surveys to help meet the increasing demand for timely data about the effects of the Corona-Crisis on the labor market.

Introduction

The COVID-19 pandemic created a unique "data paradox": at the exact moment policymakers required high-frequency labor market data to manage the economic fallout, the lockdowns paralyzed traditional face-to-face (F2F) survey methods. In early 2020, the German labor market faced unprecedented shocks, including mass layoffs, the rapid expansion of Kurzarbeit (short-time work), and a sudden transition to remote work. To craft effective policy, the German government required granular, high-quality data.

As Germany’s primary labor market research body, the Institute for Employment Research (IAB) faced a dual challenge. Their gold-standard longitudinal surveys—many of which relied on F2F interviews—became legally or logistically impossible to conduct overnight. This paper details how the IAB pivoted its survey infrastructure by transitioning to digital and telephone modes, implementing mode-effect experiments, and launching high-frequency online panels to capture real-time economic shifts.

Methodological Adaptations

To maintain data continuity while ensuring the safety of respondents and interviewers, the IAB employed several adaptive strategies across its survey portfolio:

  • Sequential Mixed-Mode Design: For the IAB Establishment Panel (EP), the institute moved from F2F interviews to a sequence of Web → Paper → Telephone. This "web-first" approach was designed to minimize the response rate drop-off typically associated with the loss of personal interviewer contact.
  • Mode Effect Experiments: A primary concern in survey methodology is whether the medium of the interview affects the data (e.g., respondents answering differently on a phone than in person). The IAB conducted experiments by randomly assigning "refreshment" samples to different modes to isolate these effects from actual pandemic-driven changes in behavior.
  • Rapid Questionnaire Iteration: Ongoing surveys, such as the Job Vacancy Survey (JVS) and the Panel Study Labour Market and Social Security (PASS), were modified mid-fieldwork. New modules were integrated to capture data on quarantine status, childcare struggles, and the perceived risk of business closure.
  • Decentralized Interviewing: Fieldwork operations transitioned interviewers to private home offices. This shift necessitated new training protocols and more intensive quality control (QC) procedures, including increased re-contacting of respondents for verification purposes.

The IAB Survey Portfolio

The IAB adapted four major ongoing surveys and established two new benchmarks to address the crisis:

Ongoing Surveys

  1. IAB Job Vacancy Survey (JVS): A quarterly establishment survey with a sample of approximately 110,000.
  2. IAB Establishment Panel (EP): An annual longitudinal survey of roughly 16,000 establishments.
  3. Quality of Life and Social Participation (QLSP): A panel focusing on long-term unemployed individuals (N~16,000).
  4. Labour Market and Social Security (PASS): A household panel of 10,000 households focusing on welfare recipients.

New High-Frequency Instruments

  1. IAB High-Frequency Online Personal Panel (HOPP): Launched in May 2020, this bi-monthly online panel utilized a random sample of 200,000 individuals drawn from administrative registers to provide a "pulse" of the labor market.
  2. IAB-BAMF-SOEP COVID-19 Survey of Refugees (CSR): A specialized telephone survey targeting the unique vulnerabilities of refugee populations during the lockdown.

Key Insights and Strategic Advantages

The IAB’s response highlighted several strategic advantages in their research design. Because several cohorts (such as the JVS) were recruited in late 2019, researchers benefited from a "natural experiment" setup, allowing for direct comparisons of the same establishments immediately before and during the lockdown.

Furthermore, the IAB found that interviewer familiarity played a crucial role. In the PASS survey, households that had previous contact with specific interviewers in F2F waves showed higher resilience in response rates when transitioning to telephone-only modes. Finally, a major strength was the integration of survey data with administrative register data from the Federal Employment Agency (BA). This allowed researchers to validate subjective survey responses against objective employment biographies.

Results and Operational Continuity

Despite the operational chaos of 2020, the JVS successfully transitioned to telephone interviewing by mid-May without a total shutdown of data collection. To maintain statistical precision against anticipated lower response rates, the 2nd quarter sample was enlarged with supplementary samples from previous years. The HOPP panel moved from concept to a 200,000-person invitation in just weeks, providing a high-frequency data stream that traditional annual surveys could not match.

Limitations and Future Research

While the pivot was successful, it introduced new methodological risks. The move to online or telephone-only modes may lead to selection bias, potentially excluding "hard-to-reach" populations with limited digital literacy. Additionally, disentangling "mode effects" from reality remains a challenge; for instance, determining if reported stress increases are due to the pandemic or the increased anonymity of a web survey. Future work will focus on longitudinal research to determine if these mode changes cause significant long-term panel attrition.

Summary updated

Abstract

The COVID-19 pandemic and associated government lockdown restrictions have fueled a high demand for survey data on how individuals and establishments are coping with the restrictions. However, the pandemic has also dramatically affected surveys themselves, forcing research institutes to adapt their fieldwork operations to the uncertain and evolving conditions. This paper documents the specific impacts of the pandemic on several ongoing surveys at the Institute for Employment Research in Germany, and describes how these impacts have been addressed. In addition, this paper summarizes efforts to rapidly develop new Covid-19 surveys to help meet the increasing demand for timely data about the effects of the Corona-Crisis on the labor market.
BibTeX
@inproceedings{sakshaug2020impacts,
  author = {Joseph W. Sakshaug and Jonas Beste and Mustafa Çoban and Tanja Fendel and Georg‐Christoph Haas and Sebastian Hülle and Yuliya Kosyakova and Corinna König and Frauke Kreuter and Benjamin Küfner and Bettina Müller and Christopher Osiander and Silvia Schwanhäuser and Gesine Stephan and Ehsan Vallizadeh and Marieke Volkert and Claudia Wenzig and Christian Westermeier and Cordula Zabel and Stefan Zins},
  title = {{Impacts of the COVID-19 Pandemic on Labor Market Surveys at the German Institute for Employment Research}},
  booktitle = {MADOC},
  year = {2020},
  doi = {10.18148/srm/2020.v14i2.7743},
}
Figure