Conflict-of-Interest Disclosure and Decision Quality in Public Policy Drafting

Participants will be public sector policy drafters, legal officers, or legislative staff. The intervention concerns the level of mandatory disclosure of potential financial or relational conflicts of interest prior to performing a standardized policy drafting task.

Participants will be randomly assigned to one of three groups:
1. Control Group (No Disclosure):
Participants receive no instruction to disclose personal or financial interests before completing the task.
2. Internal Disclosure Group:
Participants are required to complete a confidential disclosure form listing any potential financial or relational interests that could influence their decision-making. The disclosure is submitted only to a supervising authority and remains non-public.
3. Public Disclosure Group:
Participants complete the same disclosure form but are informed that their declaration will be made publicly available in a simulated online transparency register accessible to external observers.

In all groups, participants are given identical case materials describing a regulatory or policy scenario involving a potential conflict of interest (e.g., drafting a recommendation affecting a specific company or sector). They are instructed to prepare a short written policy memorandum within a defined time frame.

The intervention tests whether increasing the transparency and publicity of financial interest disclosures affects the quality, impartiality, and ethical awareness of policy drafting behavior.

2025-12-18

2025-12-19

Quality and neutrality of policy drafts (expert scoring).
Analytical accuracy and evidentiary consistency of recommendations.

The proposed randomized controlled field experiment aims to empirically examine how varying levels of mandatory financial interest disclosure affect the objectivity and quality of decision-making among public sector policy drafters. Despite the centrality of transparency and integrity in modern governance, there remains limited causal evidence on whether disclosure requirements genuinely enhance decision quality or merely produce formal compliance.

Participants will be civil servants, legal officers, or legislative staff involved in policy drafting or regulatory analysis. They will be randomly assigned to one of three experimental conditions that differ in the extent of disclosure obligations:
1. Control group (No Disclosure): participants receive no instructions to disclose personal or financial interests.
2. Internal Disclosure group: participants must confidentially declare potential financial or relational conflicts of interest to a supervising official prior to performing their assigned task.
3. Public Disclosure group: participants must disclose the same information under the assumption that it will be made publicly available in a transparency register accessible to external observers.

All participants will be tasked with preparing a short policy memorandum or regulatory recommendation based on a standardized factual case involving a potential economic stakeholder (e.g., a company subject to regulation). Independent evaluators, blinded to treatment conditions, will assess the resulting documents based on predefined quality indicators—such as analytical accuracy, impartiality, and evidentiary consistency.

The primary outcome will be the measured quality and neutrality of the produced policy drafts, operationalized through expert scoring. Secondary outcomes will include the time required to complete the task, the degree of self-reported ethical awareness, and the participants’ perceived decision autonomy.

By linking behavioral integrity mechanisms to measurable decision outputs, this experiment aims to contribute to the empirical understanding of how disclosure regimes influence bureaucratic behavior and whether transparency requirements genuinely mitigate conflicts of interest or primarily serve a symbolic function. The results may inform future legislative or institutional reforms on ethics and compliance systems within public administrations.

by computer random persons who match criteria

The persons will be people who work within the state. Primarily those working for the government (Verwaltung).

No

200.

200

50 BaFin
50 Bundeskartellamt
50 Bundesfinanzministerium
50 Bundeswirtschaftsministerium