In Germany, all employed workers pay into a mandatory unemployment insurance scheme. By a reform in 2006, a voluntary unemployment insurance scheme was introduced. Under the new scheme, individuals who work abroad, as caregivers at home or who start-up as self-employed workers can insure themselves against unemployment.
Yet, little is known about this labor market policy tool partly because administrative data on the voluntarily insured workers were incomplete in the past. More specifically, only the selective sample of those voluntarily insured workers who became unemployed was observed in the data.
Aiming to close this gap and to contribute to a more complete picture on this labor market policy tool, we present new administrative data on the contribution payments to voluntary unemployment insurance. These data became available very recently. Before we describe the data and how they can be accessed, we explain the institutional setting.
2 The voluntary unemployment insurance scheme in Germany
In addition to the mandatory unemployment insurance scheme, a voluntary unemployment insurance scheme (Versicherungspflichtverhältnis auf Antrag, Section 28a Social Code III) was established in 2006 in Germany to provide a safety net for three groups of individuals: a) persons who work outside the European Economic Area and Switzerland, b) those giving care to relatives, and c) self-employed individuals. Individuals of these groups are entitled to voluntarily insure themselves against unemployment if they work at least 15 hours per week in the respective activity (14 hours for caregivers).
The entitlement to voluntary insurance is restricted to those who fulfil at least one of the three criteria regarding their previous employment history: when starting the voluntarily insured employment activity the individual either a) was mandatorily insured for one year within the previous two years, or b) s/he previously received unemployment benefits, or c) s/he previously participated in a job-creation scheme that interrupted mandatory insurance or receipt of unemployment benefits (see Table 1 and Bundesagentur für Arbeit, 2015, for more details).
|– One year of mandatory insurance in the two years before starting the activity|
|– Receipt of unemployment benefits immediately before starting the activity|
|– Participation in a job-creation scheme that interrupted mandatory insurance or receipt of unemployment benefits immediately before starting the activity|
Source: Section 28a Social Code III.
Notes: Persons need to fulfil one of these three criteria to insure voluntarily against unemployment.
Individuals cannot insure themselves voluntarily against unemployment on top of mandatory insurance. For instance, a caregiver who holds a part-time job (in addition to the caring activity) cannot sign up for voluntary insurance, if the part-time job comes with mandatory insurance.  Some groups are excluded from unemployment insurance, because they are exempted from the mandatory unemployment insurance scheme: for example, civil servants and old age pensioners cannot sign up for voluntary insurance.
Finally, individuals can only sign up for voluntary unemployment insurance within the first three months of the respective (self-)employment activity.  When signing up for voluntary unemployment insurance, individuals have to proof that they took up the respective activity and work at least the minimum hours.
Insurance contributions vary between employment activities. As of 2015, the monthly contribution ranges from 8.51 (7.25) euro in West (East) Germany for caregivers to 85.05 euro for persons working abroad. Self-employed workers pay 42.53 (36.23) euro per month during an initial period and 85.05 (72.45) euro later on. 
Upon entry into unemployment, the maximum benefit entitlement duration and the contribution requirements are the same as for mandatorily and voluntarily insured persons: depending on age at entry into unemployment and the prior labor market history, the maximum benefit entitlement duration ranges from six months to two years. 
The calculation of the benefit amounts is slightly more complicated in the case of voluntary insurance: the benefits are calculated based on the previous employment income (like for mandatorily insured individuals), only if the voluntarily insured person has been mandatorily insured for at least 150 days during the two years before becoming unemployed. Otherwise, the benefits are calculated from virtual income, which depends mainly on qualifications.
Table 2 gives the monthly benefits derived from virtual income by qualification level as of 2015. Importantly, self-employed workers can only interrupt their self-employment twice and receive unemployment benefits unless they acquire a new entitlement to unemployment benefits.
|Qualification||Monthly benefits (in euro)|
|West Germany||East Germany|
|Master craftsmen certificate or professional degree||1,230.00||1,087.50|
|No vocational training||806.10||687.00|
Source: Bundesagentur für Arbeit (2015).
Notes: The numbers refer to persons in tax category III without children. Note that these numbers are reference values only and actual benefits may differ depending on the individual situation.
The voluntary unemployment insurance is suspended when an individual either becomes mandatorily insured or exempted from insurance as both rules out voluntary insurance. Individuals drop out of the unemployment insurance if they do not pay their contributions for three months or terminate the activity for which they insured. After five years of voluntary insurance, persons can furthermore cancel the insurance with three months’ notice. Note that self-employed persons can remain in the insurance when they have employees.
3.1 Data source
The data on the voluntary unemployment insurance contribution (VUIC, henceforth) payments were drawn from the Federal Employment Agency’s (FEA) internal IT system in the course of a research project on the insurance scheme for self-employed persons (IAB-Projekt 1803 “Info-Treatment Freiwillige Arbeitslosenversicherung”). It is the caseworkers who store the information on VUIC payments.
The data we present in this article is based on the inflow into voluntary unemployment insurance in 2013. In 2013, around 21,500 individuals signed up for voluntary unemployment insurance and started to pay into the UI system. Our data on the contribution payments are censored at May 2015. In other words, the end dates of those spells that end after May 2015 are not observed in the currently available data.
The VUIC data contain the following variables: gender, date of birth, type of insured person (self-employed, caregiver, working abroad), date of application, date of insurance start, date of insurance end (missing if ongoing at censoring date), date of drop out and IEB-linkage identifier. We merged data from the Integrated Employment Histories (IEB) to these inflows. The IEB variables are described, for example, in Jacobebbinghaus and Seth (2007). Among other information, the IEB variables cover days in employment (excluding self-employment and civil servant employment) and days in unemployment. Moreover, the data contain detailed information on participation in active labor market policy measures including start-up subsidies. In the current version, these IEB data are censored at December 31, 2014.
3.4 Data quality
Given the administrative nature of the data, its quality is high compared to survey data. We conducted some data quality checks and report the results. We merged gender and date of birth between VUIC and IEB data and found that they match in 99.5 (gender), and respectively, 99.4 (date of birth) percent. In the raw data of around 21,500 individuals, we find that 1.3 percent have more than one VUIC spell starting between January 1, 2013 and April 30, 2015. These may be due to breaks between two periods of payment or overlapping spells. To calculate the descriptive statistics that we present in the next section, we keep only the first VUIC spell per individual, and respectively, the one that ends earlier in case of same start days.
3.5 Descriptive statistics
Figure 1 depicts the distribution of inflows into VUIC payments over the calendar year 2013.
February, November and December were the months with fewest VUIC payment starts and in January and April most VUIC payments started. The median duration of VUIC payments is 577 days and by April 30, 2015, 55.3 percent of the VUIC spells are censored. 66.6 percent of the VUIC individuals are male and 33.3 percent female (Table 3). 88.4 percent are self-employed, 1.9 percent are caregivers, 9.1 percent work abroad and in 0.6 percent the information is missing. On average, VUIC individuals were 45 years old at VUIC payment start.
|Median duration VUIC payments in days||577|
|Censored in %||55.3|
|Male in %||66.6|
|Female in %||33.3|
|Type of employment|
|Self-employed in %||88.4|
|Caregiver in %||1.9|
|Abroad in %||9.1|
|Age in years||45|
|Employment history in days:|
|− Year before VUIC start|
|− Year after VUIC start|
Source: Own calculations based on VUIC data and IEB data.
Notes: Number of observations 21,518.
Employed: any type of employment with social contribution payments without self-employment (including marginal employment).
Based on the IEB data, we built the number of days that individuals were employed (any employment), and respectively, unemployed (receiving any unemployment benefits) within the year prior, and respectively, after the start of the (first) VUIC spells. On average, individuals were 177 days in employment, and respectively, 120 days in unemployment in the year prior to VUIC start. Within the year after VUIC start, they spent, on average, 58 days in (not self-employed) employment and 20 days in unemployment.
4 Prospects and data access
The data are available to researchers under strict data protection in cooperation with researchers from the IAB. For information on alternative ways of access, interested researchers may contact the authors of the article. If necessary for specific projects, the IAB will try to update the data and to access a longer observation period than the one presented above.
We thank Claus Schnabel and Gesine Stephan for comments.
Bundesagentur für Arbeit (2015), Hinweise zum Versicherungspflichtverhältnis auf Antrag in der Arbeitslosenversicherung. Available at: https://www.arbeitsagentur.de/web/wcm/idc/groups/public/documents/webdatei/mdaw/mdk5/~edisp/l6019022dstbai392135.pdf?_ba.sid=L6019022DSTBAI392138 [Accessed on April 26, 2016).Search in Google Scholar
Jacobebbinghaus, Peter; Seth, Stefan (2007): The German integrated employment biographies sample IEBS. In: Schmollers Jahrbuch. Zeitschrift für Wirtschafts- und Sozialwissenschaften, 127/2: 335–342.Search in Google Scholar
©2016 by Hofmann et al., published by De Gruyter
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.