Markus Froelich and Michael Lechner, "Combining Matching and Nonparametric IV Estimation: Theory and an Application to the Evaluation of Active Labour Market Policies", Journal of Applied Econometrics, Vol. 30, No. 5, 2015, pp. 718-738. The individual level data used in this paper are confidential and require permission from the Swiss State Secretariat for Economic Affairs SECO. The data used in this paper have been obtained from the Swiss State Secretariat for Economic Affairs SECO, contact person Thomas Ragni (thomas.ragni [AT] seco.admin.ch). The individual data on the unemployed persons stem from the Swiss unemployment insurance system (AVAM/ASAL) matched with social security and pension information from the AHV. The administrative data are confidential and have been matched by Jonathan Gast (SECO), using the unique personal identifier (social security number). The data sets have been provided to us by the SECO after destruction of the unique person identifier. Data were obtained for all of the approximately 180,000 persons registered unemployed on December 31, 1997. The first data source is the information system for placement and labour market statistics (AVAM) and the unemployment offices payment system (ASAL). These data provide detailed information about the unemployment history, ALMP participation and personal characteristics for the period 1996 to 1999. These data were merged with the social security records (AHV) for a random subsample of about 80,000 individuals. These social security data contain monthly information on the individual labour market histories and earnings (including self-employment) on a monthly basis for the period 1988 to 1999. These social security data allow us to examine the effects of active labour market programmes on employment as well as on earnings. Sample selection: In order to focus on the mobile population, various sample selection rules were applied. We deleted all students, apprentices, and homeworkers. We deleted all foreigners with less than yearly work permit. We restricted the sample to the age range 25-55. We deleted observations with very low previous earnings (insured monthly earnings < 1000 CHF). We deleted observations who were partly employed. We deleted persons who were disabled or had exhausted unemployment benefits (ausgesteuert). The final sample contains 66,713 observations. For being able to estimate the long-term effects of ALMP, the dataset was extended for this population to cover all years until 2006. Again, the extended data were provided by the Swiss State Secretariat for Economic Affairs SECO. To obtain access to the dataset, please ask the contact person Thomas Ragni at the Swiss State Secretariat for Economic Affairs SECO.