Mine, theirs or ours? An online experiment in citizens motivations to invest in mental health

Mental health spillovers have been documented in many settings, but it is unclear whether citizens are aware of their existence. This awareness is crucial, as perceiving mental health benefits as purely private can lead to systematic underinvestment in one’s own mental health. To address this, we design an experiment to assess different motivations for investing in mental health, accounting for participants’ beliefs. The goal is to understand how to better incentivize mental health investments.

Participants are exposed to one of three randomized information treatments:
1. Mental Health (Only): Provides general information about mental health issues to raise awareness.
2. Mental Health (Private): Frames mental health as a private issue, emphasizing personal benefits from mental health investments.
3. Mental Health (Social): Frames mental health as a societal issue, highlighting collective benefits from addressing mental health challenges.

Each treatment is supported by visual aids, such as images and graphs, to reinforce its message. These interventions are designed to evaluate how different framings affect participants’ willingness to cooperate in public goods production and donate to mental health initiatives.

2024-12-09

2024-12-18

Primary outcomes will be measured using a public good game and a donation game. These games identify two key outcomes: individuals’ willingness to cooperate for the production of public good and their preferences towards mental health programs, as proxied by donations
to mental health organizations.

Secondary outcomes will be measured using a list experiment and survey questions. The list experiment identifies potential stigma individuals attach to mental health. The survey questions will capture individuals’ self-reported first- and second-order beliefs, levels of social and institutional trust, social media usage, social capital, and perceptions of responsibility towards themselves and the others.

Participants are randomly assigned to receive information treatments (private, neutral, or societal perspectives), supported by images and graphs. The experiment consists of four main phases:
1. Public Goods Game: Participants, grouped in fours, allocate an initial endowment of 100 tokens in a standard Public Goods Game.
2. Donation Phase: Participants decide how much to donate from a 100-token endowment to projects focused on environmental protection, healthcare, education, and mental health.
3. List Experiment: To investigate potential stigma surrounding mental health, list experiments include sensitive items. Randomization ensures balanced exposure to control and treatment lists, allowing for analysis of individual variations.
4. Survey: A post-experiment survey gathers data on factors influencing outcomes, such as social trust, social capital, and perceptions of stigma.

The study seeks to examine whether providing information about mental health issues (Mental Health (Only) treatment) enhances awareness and increases preferences for cooperation in the production of public goods, as well as donations to a mental health project.

It also explores how these attitudes may shift depending on whether investments in mitigating mental health issues are framed as personal, private investments generating private benefits (Mental Health (Private) treatment) or as collective investments generating societal benefits (Mental Health (Social) treatment).

The experiment begins with randomized information treatments, designed to present mental health from three perspectives: private, neutral, and societal, along with an active control condition in which participants receive information unrelated to mental health. Each treatment includes a supporting image and a graph that visually highlights the key message of the information provided. Following the information treatment, the experiment proceeds through four distinct phases:
• Public Goods Game: The game adopts the standard Public Good Game structure, without framing the public good around mental health issues. Each participant is provided with an initial endowment of 100 tokens and grouped with three other participants.
• Donation phase: This phase follows the structure of a standard Charity Dictator Game. Participants are provided with a 100-token endowment and are asked to decide how much they wish to donate to each of the listed projects. These projects include initiatives focused on environmental protection, healthcare, educational support, and mental health.
• List Experiment: Mental health can be a sensitive topic and may reveal underlying stigma; therefore, we design a List Experiment to include a sensitive item related to mental health issues. The experiment employs a list-based design, with the control list containing four items and the treatment list including an additional sensitive item. To minimize non-strategic errors related to the number of items, a placebo list is introduced. This placebo list builds on the control group items by adding a placebo item (Riambau and Ostwald 2021).
Together, these three lists constitute group A. In addition, we generate a corresponding set of three lists, referred to as group B, which are comparable to the treatment, control, and placebo lists in group A, while keeping the sensitive and placebo items unchanged. During the experiment, each participant is always presented with the treatment list and either a placebo or a control list.
Initially, the participant is randomly assigned to one list from group A. The treatment list is presented with a probability of 50%, while the control and placebo lists each occur with a probability of 25%. Subsequently, the participant is assigned to a group B list based on their prior exposure. If they previously encountered a treatment list, they are now assigned either a control or a placebo list (with the same probability). Conversely, if they previously encountered a control or a placebo list, they are now assigned to the treatment list.
This approach ensures balanced exposure to both treatment and non-treatment conditions. This design allows for analysis of within-individual variation, ensuring that the findings are not influenced by the specific content of the lists. Since the outcome might be affected itself by the information treatments, we schedule this phase before them.
• Survey: The post-experimental survey includes a set of questions designed to gather self-reported information on various factors that may contribute to heterogeneous effects on the primary outcomes.

Randomization by computer, within Qualtrics.

Individual level.

No

We will include 8260 individuals in our randomization, in 7 countries:France, Germany, Italy, Latvia, Slovakia, Spain,
Sweden.

We will include 8260 individuals in our randomization.

We planned to balance the control and treatment groups.

Our design consists of individuals from seven countries. Since we expect small effects from our treatments, with a target sample size of 1315 individuals for France, Germany, Italy and Spain, and 1000 for Latvia, Slovakia and Sweden (i.e., 8260 in total), we will be able to detect a minimum effect size ME=0.071 (7.1% of a standard deviation) at power p = 0.8 and alpha α = 0.05 over standardized outcomes.