The Comparative Unemployment Benefit Conditions & Sanctions Dataset


Many believe that the unemployed are too 'lazy' or 'choosy', and that unemployment rates could be lowered if the unemployed would be given more 'tough love'. Others counter that the unemployed are already treated too harshly, and that they should rather receive more help and support.

For a long time, though, it was not easy to say exactly how strictly or leniently the unemployed in a given country were really treated: how freely they could choose between available jobs, how stringently their job-search efforts were checked, and what penalties could be imposed if they would not comply with their obligations. As a result, it was also not fully clear whether countries with tough rules really do have lower unemployment rates — and therefore whether putting the unemployed under pressure really is a good way to reduce unemployment.

The Comparative Unemployment Benefit Conditions & Sanctions Dataset was created to solve this problem. It provides systematic information and quantitative indicators of the strictess of job-search and reporting requirements, the definition of 'suitable work', and unemployment benefit sanction rules in 21 economically advanced democracies in Europe, North America and Australasia between 1980 and 2012.

These data have already provided us with a clearer picture of how strict rules for the unemployed are related to the performance of the labor market. The two interactive graphs below show that countries that had in the period between 1980 and 2012 required the unemployed to seek work more actively and be available for a wider range of jobs tended to have lower unemployment rates (graph on the left), but only slightly so. Also visible is that countries that had in the same period imposed tougher benefit sanctions on the unemployed tended to have higher unemployment rates (graph on the right).

More detailed analyses (published here and here) revealed that putting greater pressure on the unemployed to seek and accept work does indeed improve labor market performance; stricter sanction rules, on the other hand, are a result of high unemployment: Sanction rules tend to be made stricter after economic downturns. But the data do not show that stricter sanction rules then have an effect on unemployment rates (but they do have other negative effects).

The creation of this dataset was made possible by a generous grant from the Crafoord Foundation (Crafoordska stiftelsen) as well as lots of help, advice, and support from numerous contributors, country experts, and interested scholars. The work on this dataset was conducted at the Department of Political Science at Lund University.

Access & explore the data

You can access the data dashboard here. (Please view this website from a device with a wider screen to see the embedded dashboard and other visualizations.)

Directly load the dataset into R

R users can directly import the main dataset or selected subsets of it using a single convenience function, getCondData(). The function itself can also be directly sourced into R:



Arguments: The function has three optional arguments, which are used to subset the data:

  • indicators: Selects one or more of the synthetic summary indicators included in the dataset (see also "What the data measure" in the dashboard above or the Codebook for further details). Specified via a string (if selecting a single indicator) or a string vector (if selecting multiple):
    • conditionality: The indicator for the overall conditionality of unemployment benefits;
    • conditions: The indicator for the strictness of job-search and availability requirements;
    • sanctions: The indicator for the overall strictness of benefit sanction rules;
    • occup: The indicator for the degree of occupational protection ("can unemployed workers be forced to change occupations?");
    • wage: The indicator for the degree of wage protection ("can unemployed workers be forced to accept lower wages?");
    • oth: The indicator for the extent to which there are 'other' valid reasons (e.g. caring responsibilities, religious beliefs) for refusing an offer of employment;
    • jsr: The indicator for the strictness of job-search requirements;
    • vol: The indicator for the strictness of sanctions for voluntary unemployment;
    • ref: The indicator for the strictness of sanctions for an initial refusal of an offer of employment;
    • rep: The indicator for the strictness of sanctions for repeated refusals of offers of employment;
    • fail: The indicator for the strictness of sanctions for failures to report job-search activities;
  • time: Subsets the data to one or more years between 1980 and 2012. Specified as a numeric value (e.g. 1991) or a numeric vector;
  • space: Subsets the data to one or more countries. Specified as a string (if selecting a single country) or a string vector (if selecting multiple countries). Must be one or more of: Australia, Austria, Belgium, Canada, Denmark, Finland, France, Germany, Greece, Ireland, Italy, Japan, Korea, Netherlands, New Zealand, Norway, Portugal, Spain, Sweden, Switzerland, United Kingdom;


Get the entire dataset:


Get all data for Sweden & Australia:


Get only the overall sanctions indicator, but for all countries and years:


Get only the overall indicator for Austria in five year intervals:

getCondData(indicators="conditionality", space="Austria", time=seq(1980,2012,5))

Publications using the dataset

Published research using this dataset includes:

If you have also used this dataset in your research and would like to have any resulting publications listed here, please do let us know.


Carlo Knotz

Carlo worked on the Benefit Conditions & Sanctions Dataset during his time as a doctoral student at the Department of Political Science at Lund University. His dissertation, which draws extensively on the Benefit Conditions & Sanctions Dataset, was awarded the 2017 Research Prize by the Swedish Federation of Unemployment Insurance Funds (Arbetslöshetskassornas Samorganisation/Sveriges A-Kassor). He now holds a position as Associate Professor of Political Science at the University of Stavanger. Previously, he worked as a Postdoctoral Researcher at the Swiss Graduate School of Public Administration (IDHEAP) in Lausanne and as a Postdoctoral Fellow at the Bremen International Graduate School of Social Sciences (BIGSSS).

Moira Nelson

Moira is Assistant Professor at the Department of Political Science at Lund University. Her research examines the determinants and consequences of social policies, including the electoral incentives of governments to introduce welfare state reforms and the employment consequences of social policies.

The following persons contributed to this dataset by collecting data on individual countries:

  • Alberjan Tollenaar & Gijsbert Vonk collected the data on the Netherlands
  • Alejandra Keidel Fernandez assisted with the collection of the data on Spain
  • Luis Gonzales da Silva collected the data on Portugal
  • Stefano Sacchi & Patrik Vesan collected the data on Italy
  • Takeshi Yanagisawa collected the data on Japan
  • Varvara Lalioti collected the data on Greece

To ensure that our data are complete and free of errors, we asked a number of scholars who are experts in this field to review them and, where necessary, suggest corrections:

  • Australia: Agnieszka Nelson (Australian National University & FaHCSIA) and David Stanton (Australian National University)
  • Austria: Sarah Bruckner and Günter Krapf (Arbeiterkammer Wien)
  • Belgium: Ive Marx (Herman Deleeck Centre for Social Policy, Antwerp University)
  • Canada: Ann Porter (York University)
  • Denmark: Thomas Bredgaard (Aalborg University)
  • Finland: Heikki Räisänen (Ministry of Employment Finland & University of Tampere)
  • France: Jean-Claude Barbier (Université Paris 1 Sorbonne)
  • Germany: Jochen Clasen (University of Edinburgh) and Regina Konle-Seidl (Institute for Employment Research, IAB)
  • Greece: Dimitris Karantinos (National Centre for Social Research, EKKE)
  • Ireland: Mel Cousins (Trinity College Dublin)
  • Italy: Stefano Sacchi (University of Milan & Collego Carlo Alberto, Turin) and Patrik Vesan (Università della Valle d'Aosta)
  • Japan: Takeshi Yanagisawa (Meijo University)
  • Netherlands: Frans Penning (Utrecht University) and Albertjan Tollenaar (University of Groningen)
  • New Zealand: Alex McKenzie (Ministry of Social Development New Zealand)
  • Norway: Ivar Lødemel (Oslo and Akershus University of Applied Sciences)
  • Portugal: Luis Gonzales da Silva (Instituto do Direito do Trabalho)
  • South Korea: Deok Soon Hwang (Korea Labour Institute)
  • Spain: David Rueda (University of Oxford)
  • Sweden: Ola Sjöberg (SOFI & Stockholm University)
  • Switzerland: Carla Togni (EESP Lausanne)
  • United Kingdom: Jochen Clasen (University of Edinburgh)


No dataset is ever perfect or complete.

Therefore, and firstly, please do let us know if you spot any errors or problems in the materials we provide!

There are also some ways in which the data could still be further improved. We ourselves are lacking the time and resources to do so, but would be more than grateful for input from others interested in contributing. We will also provide all the help and assistance we can.

Examples of opportunities for further development include:

  • The data could be updated to reflect more recent policy changes. A series of OECD studies has already provided updated data for selected years, and these could be used as a starting point.
  • The coded data on Italy measure availability conditions and sanction rules for the Unemployment Benefit scheme. We have also collected, but not coded yet, qualitative data on the Italian Mobility Allowance and Short-term Work benefit schemes. Any interest in coding these schemes as well would be greatly appreciated.
  • The United States are, due to the devolved nature of the U.S. Unemployment Benefit scheme, not covered by this dataset yet. We do, however, have data on State-level rules that are not coded yet, and additional historical information on legislative rules and their changes in U.S. States may also be available from the U.S. Social Security Administration's Social Security Bulletin (see e.g. vol. 16(2) of 1953) or various articles in the U.S. Bureau of Labor Statistics' Monthly Labor Review (see e.g. March 1981).
  • We were not able to collect data on any of the countries of Central and Eastern Europe, although they are now commonly included in related data collection projects by the OECD and others (as they should be). Anyone who is interested in contributing historical data on unemployment benefit eligibility rules and sanctions in any of these countries is more than welcome to contact us.

Use policy & Disclaimer

You can download and use the data, documents, and other materials provided here free of charge. If you do, we kindly ask you to reference this dataset as:

  • Knotz, Carlo and Moira Nelson. 2019. The Comparative Unemployment Benefit Conditions & Sanctions Dataset. Lund: Department of Political Science, Lund University.

You may not sell any of the materials obtained here or parts thereof. You may also not incoporate our data and materials or parts thereof in any other dataset (other than replication datasets) without our explicit permission. You may furthermore not distribute any of the materials you obtained from us if they have been subsequently altered in any way, shape, or form.

We have taken great care to produce high-quality data, but we can unfortunately not give any warranty that the dataset and all the accompanying files are entirely free of errors. Should you detect an error in any of the files or would like to suggest any other modifications, please feel free to contact us. We reject all liability for any damages incurred while using any of the materials provided by us.