How Sortition works

Two principles guided the random selection of citizens for the Citizens' Assembly "Germany's role in the world":

  1. Since the Citizens' Assembly "Germany's role in the world" is a nationwide procedure, the concept for the random selection of the citizens had to ensure that all eligible participants, i.e. all residents in Germany from the age of 16 who have German citizenship, are theoretically able to participate.
  2. The Citizens' Assembly is intended to represent Germany on a smaller scale. Among the 160 participants, gender, federal states, town sizes, educational qualifications, and migration experiences of the participants should be represented.

These principles were implemented through a random selection of municipalities, citizens and the subsequent compilation of the group of participants from the registrations according to the above-mentioned criteria.

1. Representative random selection

In Germany, it is possible to draw random samples from municipal registers of residents. This procedure is usually used for lot-based participation procedures in Germany and will also be used for the Citizens' Assembly "Germany's role in the world". Compared with a random sample from the telephone directory, drawing from the population register offers the advantage that households are not selected, which can include several people. Moreover, it is also possible to reach those who do not have a landline connection.

Since the Citizens' Assembly "Germany's role in the world" is a nationwide procedure, the concept for drawing the sample had to ensure that all those entitled to participate, i.e. all residents with German citizenship from the age of 16 years, were theoretically able to participate. In order to ensure this, a random selection was carried out.

First of all, municipalities were drawn at random from the official list of municipalities in all federal states. The residents' registration offices in these municipalities then received a request to draw a random sample of citizens from their register of residents and to submit the list to the Citizens' Assembly "Germany's role in the world" for the invitation of the participants.

Technically, the selection of the municipalities as well as the citizens is done using algorithms. When selecting the municipalities, the nexus Institute used the "random area" function in an Excel file for all municipalities, sorted by federal state and staggered according to size classes. The registration offices, which supply the address data, also used algorithms for randomly drawing from their registers.

2. Size of the sample

In order to reach the desired number of 160 participants, many more citizens had to be invited than can participate in the Citizens' Assembly. This is due to the fact that many randomly selected persons will not participate in the Citizens' Assembly for various reasons. The most common reasons are illness and age, travel difficulties, family celebrations, holidays, and professional or other commitments. There are also people who are not interested in the topic or in discussion events in general. The aim is to ensure that the invitations attract more than 160 people to participate. This will make it possible for those groups of people who are usually poorly represented in democratic processes to be represented in the Citizens' Assembly "Germany's Role in the World". Furthermore, we also want to form a standby group if someone is prevented from coming on short notice. To ensure that the Citizens' Assembly will eventually bring together 160 people, 4,365 people throughout Germany were contacted.

3. Distribution of the sample among the federal states

The objective of mapping Germany on a small scale means that the composition of the membership of the Citizens' Assembly must reflect the proportion of the population of the 16 federal states in relation to the total population. For this reason, a different number of people from the different federal states were invited to the Citizens' Assembly on the basis of their share of the total population.

Federal State Population Percentage of Population Size of the sample Target number of participants

Baden-Württemberg

11 069 533

13%

638

21

Bavaria

13 076 721

16%

728

25

Berlin

3 644 826

4%

228

7

Brandenburg

2 511 917

3%

157

5

Bremen

  682 986

1%

43

1

Hamburg

1 841 179

2%

115

4

Hesse

6 265 809

8%

413

12

Mecklenburg-Western Pomerania

1 609 675

2%

63

3

Lower Saxony

7 982 448

10%

500

15

North Rhine-Westphalia

17 932 651

22%

993

35

Rhineland-Palatinate

4 084 844

5%

240

8

Saarland

990 509

1%

29

2

Saxony

4 077 937

5%

202

8

Saxony-Anhalt           2 208 321 3% 96 4
Schleswig-Holstein 2 896 712 3% 104 6
Thuringia 2 143 145 3% 114 4
Total 83 019 213 100% 4365 160

4. Selection of municipalities by size class

The selection was based on the official list of all politically independent municipalities (with a municipal association) in Germany (as of 31 December 2018), which is published by the Federal Statistical Office (www.destatis.de).

In the public debate on disenchantment with politics, especially when analysing the feeling of "being left behind", the focus is often placed on the geographical contrast between urban and rural. In order to better reflect the different living conditions in the sample of participants, we considered the size of the municipality in the first stage of the random selection.

Therefore, we were guided by existing systems for classifying municipalities according to size in terms of inhabitants, such as those used by the “Deutscher Städtetag” (Association of German Cities) to record municipalities. To improve the manageability of the random sample, we have worked with four size classes.

Municipality type Population Size class
Major City > 100.0000 I
Midsize City 20.000 - under 100.000 II
Small Town 5.000 - under 20.000 III
Rural Community under 5.000 IV

In the city-states, the random sample was drawn at the state level; in all regional states, the participating municipalities were selected at random from these four size classes. In the next step, a random sample from the population register was requested in these municipalities.

4.1 Selection of municipalities

The municipalities were selected from the official list of municipalities using the "random number" function of Excel. In each size class, a selection range between 1 and 5,000 was given, so that all municipalities in the size class were assigned a random number. The municipalities were then sorted in ascending order. The municipality that received the lowest random number was selected.

The share of municipalities of one size class in the total sample of the federal state corresponds to the share that municipalities of this size class have in the total of all municipalities of the federal state.

If the number of those to be invited to the procedure from the size class in the federal state was higher than 2% of the population, the municipality with the next largest random number was included in the selection of municipalities in this size class. The 2% limit was set so that families or groups of friends do not by chance represent the size of a municipality in a federal state. Since there are very small independent municipalities in some federal states, up to nine municipalities were selected in the fourth size class according to this procedure (Rhineland-Palatinate).

5. Mini Publics: Germany downscaled through sortition

Lot-based public participation procedures are also called “mini publics”. A term that draws attention to the fact that society is represented in miniature through sortition. To ensure that Germany is accurately represented, the Citizens' Assembly "Germany's Role in the World" looks at other criteria for the composition of the 160 members of the Citizens' Assembly in a further stage, in addition to the spatial distribution of participants.

5.1 Composition of the participants according to further criteria

The random selection guarantees a diversity of participants that is not achieved in other procedures, such as those based on public invitations. Therefore, participation procedures based on sortition are particularly inclusive. However, even with this method, there are groups of people who are over-represented compared to their share of the total population: This is particularly true, for example, of the highly educated group. To counteract this, the participants are put together from the positive responses in such a way that they represent Germany as well as possible on a small scale.

The decisive factor here is that the distribution of socio-demographic characteristics of the participants in the Citizens' Assembly corresponds as closely as possible to the distribution in the population as a whole. In addition to the spatial distribution, the following characteristics are taken into account: gender, age group, educational level, and migration background.

Some of this data was already included in the data of the residents' register, and some of it was collected with the re-registration card that all the invitees received.

The following table shows the statistical values on which the composition of the Citizens’ Assembly "Germany's role in the world" is based.

Gender Percentage of population aged 15 and over
Male 49,3 %
Female 50,7 %
Age Percentage of population aged 14 and over
14 - 24 12,8 %
25 - 39 21,9 %
39 - 65 40,4 %
65 and older 24,9 % 
Level of education Percentage of population aged 15 and over in %
Current high school student 3,6 %
No high school degree 4,0 %
Hauptschule 29,6 %
Intermediate level of education, including polytechnic secondary school 29,9 %
Technical/general higher education entrance qualification 14,3 %
University degree, including doctorate 17,6 %
Migration background Percentage of population aged 15 and over in %
German citizens with firsthand migrant background or parents with migrant background 12,1 %