Public Debt and Entrepreneur Investment

How to cite

Paaso, Mikael; Torstila, Sami; Okat, Deniz; Blomkvist, Magnus. 2025. "Public Debt and Entrepreneur Investment." AEA RCT Registry. AEARCTR-0015646. https://doi.org/10.1257/rct.15646-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Mikael Paaso* , Sami Torstila , Deniz Okat , Magnus Blomkvist

* Primary investigator

Registry ID
AEARCTR-0015646
DOI
https://doi.org/10.1257/rct.15646-1.0
First registered
Last updated
Status
In development
Trial dates
2025-03-31 – 2027-01-01
Countries
Private
JEL codes
H63 Sovereign Debt, G40 Behavioral Finance
Registry record
http://www.socialscienceregistry.org/trials/15646

Abstract

We study whether entrepreneurs reduce their planned investment in response to higher levels of public debt. Ricardian equivalence (Barro, 1974) implies that entrepreneurs should anticipate higher future taxes when public debt is higher and reduce investment, but
entrepreneurs may be inattentive to the level of debt, myopic about the future or anticipate that they will not be the ones paying those taxes. We conduct two survey experiments, with results linked to administrative financial data, to test whether entrepreneurs report a lower willingness to invest when presented with exogenous shocks to the level of public debt.

We will conduct a survey of about 1000 Finnish SMEs (most Finnish firms are required to prepare public financial statements). We first test if firms that report being more concerned about public debt invest less in years during which public debt is more salient (as measured by mentions in parliamentary speeches). We then conduct two survey experiments: The first is a conjoint analysis, in which entrepreneurs are presented with 2 randomized “states of the future economy” in which attributes such as previous year GDP growth, inflation, government composition and government-debt-to-GDP are varied. We ask participants under which state they are more likely to invest. We
then conduct an information provision experiment in which all participants are presented with a forecast of public debt, but the nature of this forecast (optimistic vs. pessimistic) is randomized. We ask participants whether this forecast changes their investment plans and expectations about economic growth.

We also anticipate replicating our experiment with self-identified entrepreneurs on Prolific.

Experimental design

We will run two separate experiments within the project:

First, we will run a conjoint analysis. In this experiment, participants are presented with five scenarios in which they are presented with two possible states of the economy in 2027. These are presented as a table, with a row for prev. year GDP growth, GDP/capita, prev. year inflation, the interest rate, corporate tax rate, personal income tax rate (highest rate) as well as public debt/GDP. All of the parameters are randomized from a set of realistic options. We then ask participants which future state of the economy is more conducive to investment in their business.

Second, we will run an information provision experiment. Participants are presented with either a high or a low forecast of public debt to GDP in the future (all participants receive a forecast, i.e. there is an active control). Before this, we elicit their current expectations of public debt in 2030. We then test ask for their posterior estimate of public debt to GDP (after the forecast) as well as their investment plans (Likert scale of invest less or more compared to today) and perceptions of economic conditions. We plan to analyse this in two ways: First, by simply comparing expectations and investment plans for those who received the high vs the low forecast, and second by testing
whether expectations and investment plans change in the same direction as the public debt forecast (i.e. if posterior – prior is positive, we expect investment to decline).

Design details

Randomization method
Software (randomization done by the survey company we are using)
Randomization unit
Individual entrepreneur
Clusters
Aiming for 1000 individuals
Observations
1000
Arms
Information provision experiment: 500 per forecast
Primary outcomes
Conjoint experiment: Which "future state of the economy" do participants consider more favorable for investment? Information provision experiment: Likelihood of investing after forecast (or change in likelihood to invest)
Secondary outcomes
Information provision experiment: Change in perceived economic conditions

Cited by

No citation data available for this registration yet.

Keywords