Experimental Design Details
This experiment employs a randomized controlled design to examine how tax policy changes influence public expectations of key macroeconomic indicators.
Pre-Treatment Measures
Before any policy intervention, participants first provide their prior expectations regarding key macroeconomic variables, including inflation, interest rates, GDP growth, debt-to-GDP ratio, and deficit levels. These forecasts are elicited for two different time horizons: one year and three years into the future.
Next, participants answer a series of questions assessing their trust in government. This section includes measures of political affiliation and their confidence in government policies, allowing us to analyze whether political beliefs shape expectation formation in response to tax policy changes.
Information Treatment
Participants are then randomly assigned to one of four groups: a control group, which receives no information about tax policy, and three treatment groups, each of which is exposed to a different tax policy scenario. Some treatment groups receive information about a tax rate increase of different magnitudes, while others are provided with contextual details explaining the purpose of the tax increase.
Post-Treatment Measures
After receiving the intervention, participants update their forecasts for the same macroeconomic variables as in the pre-treatment phase. In addition to macroeconomic expectations, participants are also asked about their personal financial behaviors, including expected changes in savings and investment decisions.
To better understand the mechanisms behind expectation formation, respondents are further asked to explain their reasoning for their updated forecasts. Additionally, participants answer tax-related questions, designed to assess how their beliefs about taxation shape their broader economic outlook.
Demographic Information
Finally, participants complete a demographic questionnaire, collecting details on age, income, education, employment status, and other socioeconomic factors.