Earnings Disclosure by Politicians

We plan to run a conjoint experiment. We present each respondent 3 profiles of hypothetical MPs and each profile is rated along various dimension (see below for the outcomes).

2023-12-07

2024-01-31

We consider five different outcomes that are based on voters’ evaluation of MPs. Voters evaluation of MPs whether (1) they represent the interests of their voters or (2) their own or third interests as well as whether they are (3) competent or (4) hard-working. All outcomes are measured on a 1-7 Likert scale and the order of questions is randomized for each respondent. We also plan to run heterogeneity analysis e.g. by self-reported political preference.

First, we run a standard attention check. Then, each respondent is presented with 3 MP profiles. Each MP is randomly assigned the following characteristics: gender (male/female), marital status (single/married), party (center-left (SPD), conservative-right (CDU/CSU), liberal (FDP), green (Grüne) or left-wing (Die Linke), tenure in parliament (1,2, 3 or more legislative periods), mandate (directly elected or elected via party list), and published outside activities (no activities, lawyer, self-employed/entrepreneur, member of supervisory or advisory board, and speeches/journalistic work) and respective earnings (no earnings, earnings between 1,000-3,500 EUR, 3,500-7,000 EUR and 7,000 EUR and more). The ordering of the characteristics is randomized for each respondent. We are primarily interested in the effect of the published outside earnings and activities on voters’ preferences.

The characteristics of the MPs are randomized by the survey provider by a computer. The characteristics of the MPs are randomized by the survey provider by a computer. We apply the following quotas: (1) gender (66% male, 35% female), (2) marital status (34% single, 66% married), (3) party (center-left (SPD) 32%, conservative-right (CDU/CSU) 30%, liberal (FDP) 14%, green (Die Grünen) 18% or left-wing (Die Linke) 6%, (4) tenure in parliament (1 (25%),2 (26%), 3 or more legislative periods) (49%), (5) mandate (directly elected (47%) or elected via party list (53%)), and (6) published outside activities (no activities and no earnings (25%), all remaining combinations of activities (lawyer, self-employed/entrepreneur, member of supervisory or advisory board, and speeches/journalistic work) and respective earnings (earnings between 1,000€ - 3,500€, 3,500€ - 7,000€ and 7,000€ and more) have a probability of 6.25%. The percentages of the demographic characteristics of MPs reflect the composition of the German Bundestag in the legislative period starting from 2021 and the earning categories refer to the offical brackets in which outside earnings are reported.

Each individual respondent is presented with 3 independently randomized profiles of MPs.

No

We plan to cluster our standard errors on the level of the individual respondent (N=4,000).

We plan to have 4,000 respondents that are representative of German voters based on the latest federal election in the year 2021 along the following characteristics: age group, gender, East vs. West Germany, and party preference. Each respondent is presented with 3 independently randomized profiles of MPs. This gives us 12,000 observations.

We have four different treatment arms that are of primary interest for us. First, no outside earnings (sample size of 25% of # respondents = 1,000), outside earnings between 1000€ - 3500€ (sample size of 25% of # respondents = 1,000), outside earnings between 3500€ - 7000€ (sample size of 25% of # respondents = 1,000), outside earnings above 7000€ (sample size of 25% of # respondents = 1,000).