The Research Data Centre of the German Federal Employment Agency at the Institute for Employment Research (RDC-IAB)

Antoni, Manfred; Schmucker, Alexandra

doi:10.23889/ijpds.v4i2.1141

Cited by 5 publications

(5 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For a comprehensive documentation of the Integrated Employment Biographies (IEB), one of the main IAB data products we also use, see e.g. Antoni et al (2019).…”

Section: Methodsmentioning

confidence: 99%

First step and last resort: One-Euro-Jobs after the reform

Harrer¹,

Stockinger²

2021

J. Soc. Pol.

View full text Add to dashboard Cite

Job creation programmes aim at increasing the employability of hard-to-place unemployed, and eventually integrating them into employment. Yet, previous evaluation studies have been pessimistic about their efficacy. For One-Euro-Jobs, a job creation programme for welfare benefit recipients in Germany, previous evaluations found unfavourable effects particularly for easier-to-place participants. Thus, in 2012 the legislator reformed the programme in order to target the hard-to-place more accurately. This study is the first post-reform evaluation of One-Euro-Jobs. We find that, although the programme is indeed better targeted than before, One-Euro-Jobs decrease participants’ employment chances within three years after programme entry. These outcomes are worse than those found for pre-reform participants. We cannot conclude with certainty whether the reform decreased the programme’s efficacy, but we identify channels through which the reform and other contemporaneous changes may have done so. These channels include changes in programme design features, changes in business-cycle conditions, and prolonged lock-in effects due to “programme careers”. To substantiate the latter explanation, we also provide novel evidence that One-Euro-Jobs seem to initiate programme careers.

show abstract

“…For a comprehensive documentation of the Integrated Employment Biographies (IEB), one of the main IAB data products we also use, see e.g. Antoni et al (2019).…”

Section: Methodsmentioning

confidence: 99%

First step and last resort: One-Euro-Jobs after the reform

Harrer¹,

Stockinger²

2021

J. Soc. Pol.

View full text Add to dashboard Cite

show abstract

“…We demonstrate the use of SSDDR with reject option with an example of algorithmic profiling of job seekers. We use a large, anonymized sample of administrative labor market records provided by the German Institute for Labour Market and Employment Research (IAB, [1]), which enables us to model a realistic use case of public profiling. Decisions on the allocation of resources are made using either the predicted duration of unemployment (T , measured in months) or long term unemployment (Y LT U ), a binary version of the duration using 12 months as a threshold.…”

Section: Uncertainty Aware Profiling Of Job Seekersmentioning

confidence: 99%

“…Administrative data We use a 2% random sample of German administrative labor market records, called Sample of Integrated Employment Biographies (SIAB, [1]). The data combine information from various sources such as employment information, unemployment information and unemployment benefits receipt.…”

Section: A Appendixmentioning

confidence: 99%

“…Technological developments have led to an automation of decision making processes in various contexts, including employee recruitment [34], personalized medicine [45], jurisdiction [24], finance [17], and disaster management [44]. Automated decision-making (ADM) 1 systems aim to advance timeliness, efficiency, quality, and transparency in the allocation of resources and interventions [36].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Uncertainty-aware predictive modeling for fair data-driven decisions

Kaiser¹,

Kern²,

Rügamer³

2022

Preprint

View full text Add to dashboard Cite

Both industry and academia have made considerable progress in developing trustworthy and responsible machine learning (ML) systems. While critical concepts like fairness and explainability are often addressed, the safety of systems is typically not sufficiently taken into account. By viewing data-driven decision systems as socio-technical systems, we draw on the uncertainty in ML literature to show how fairML systems can also be safeML systems. We posit that a fair model needs to be an uncertainty-aware model, e.g. by drawing on distributional regression. For fair decisions, we argue that a safe fail option should be used for individuals with uncertain categorization. We introduce semi-structured deep distributional regression as a modeling framework which addresses multiple concerns brought against standard ML models and show its use in a real-world example of algorithmic profiling of job seekers.

show abstract

“…A rule-based matching approach using both entity name and address similarity can be found in such domains as the investigation of healthrelated behaviours dependent on living environments (Hirsch et al, 2020;Mendez et al, 2014), validating names and addresses of transportation and logistic entities (Guermazi et al, 2020), matching observations across financial datasets (Cohen et al, 2018;Burdick et al, 2015), identifying same entities in patent files (Medvedev and Ulanov, 2011;Magnani and Montesi, 2007). Most of them conclude that domain expert knowledge integration improves or would improve matching results (Pilania and Kumaran, 2019;Cohen et al, 2018;Choi et al, 2017;Antoni et al, 2018;Schild, 2016;Mendez et al, 2014;Magnani and Montesi, 2007).…”

Section: Related Workmentioning

confidence: 99%