The Widening Gap in Tax Attitudes: Role of Government Trust in the post COVID-19 period

Using Japanese data, this study reveals that the post-pandemic shift from unconditional support for tax-based redistribution to conditional altruism has created a widening gap where income redistribution now critically depends on trust in government, particularly among high-income individuals.

The Gist

Here is an explanation of the paper, translated into everyday language with some creative analogies.

The Big Idea: The "Social Contract" Got a New Clause

Imagine society is like a giant neighborhood potluck. Everyone brings a dish (taxes) to share with the community, especially to help those who can't cook or have nothing to bring (the poor).

Before the pandemic, this potluck ran on unspoken rules. Everyone just knew, "We help each other because that's who we are." It didn't matter if you trusted the person organizing the potluck; you brought your dish anyway because it was the "right thing to do."

This paper argues that COVID-19 changed the rules of the game. Now, people (especially the wealthy) are saying, "I'll only bring my dish if I really trust the organizer to do a good job." If they don't trust the organizer, they leave their dish at home.

---

The Story in Three Acts

Act 1: The "Before" Times (Unconditional Giving)

Before the virus hit, Japan had a very strong culture of helping each other.

• The Analogy: Think of it like a family dinner. You help set the table not because you trust your mom to cook perfectly, but because you love your family. It's automatic.
• The Finding: Rich people and poor people were both willing to pay higher taxes to help the needy, regardless of whether they trusted the government. The "social norm" (the unwritten rule of being a good citizen) was strong enough to keep the potluck going.

Act 2: The Shock (The Virus Hits)

When the pandemic hit, it was like a massive storm that blew through the neighborhood.

• The Result: The storm didn't just damage houses; it damaged the feeling of community. People stopped seeing each other face-to-face. The "automatic" desire to help each other faded.
• The Data: The researchers found that overall, people became less willing to pay taxes to help the poor after the pandemic. The baseline of generosity dropped.

Act 3: The "Widening Gap" (The New Reality)

This is the most important part of the study. The pandemic didn't just lower generosity; it split people into two camps based on Trust.

• The "High Trust" Group: These are people who still believe the government is doing a good job.
  • The Rich: Surprisingly, rich people who trust the government became more generous. They are willing to pay more taxes to help the poor because they believe the money will actually get there and do good.
• The "Low Trust" Group: These are people who think the government is incompetent or corrupt.
  • The Rich: Rich people who don't trust the government stopped giving almost entirely. They became "conditional cooperators." They said, "If I don't trust you to manage the potluck, I'm not bringing a dish."

The "Widening Gap":
Before the pandemic, the rich and the poor were on the same page about helping each other. Now, there is a huge gap.

• If a rich person trusts the government $\rightarrow$ They pay high taxes.
• If a rich person doesn't trust the government $\rightarrow$ They pay almost nothing.

The gap between these two groups of rich people has become massive.

---

Why Does This Matter? (The Takeaway)

The authors use a metaphor of a bridge.

• Before: The bridge (society) was held up by strong concrete pillars (social norms). It didn't matter if the weather was bad; the bridge stood.
• After: The concrete pillars cracked. Now, the bridge is held up by a single, fragile rope: Trust in the Government.

The Warning:
If the government loses the trust of the wealthy, the "bridge" collapses. The rich will stop paying taxes to help the poor, even though inequality is at its worst. The social contract is no longer about "we are all in this together"; it's now a transaction: "I will help you only if I trust you."

Summary in One Sentence

The pandemic broke the old, automatic habit of helping each other, and now, the rich will only help the poor if they are absolutely sure the government is trustworthy enough to handle the money.

Technical Summary

Here is a detailed technical summary of the paper "The Widening Gap in Tax Attitudes: Role of Government Trust in the post COVID-19 period" by Eiji Yamamura and Fumio Ohtake.

1. Problem Statement and Research Context

The paper investigates the structural transformation of citizens' willingness to support income redistribution (altruism) in the aftermath of the COVID-19 pandemic. While previous literature suggests that negative shocks (like natural disasters) can strengthen social capital and support for safety nets, empirical evidence regarding the pandemic is paradoxical. Some studies indicate increased inequality, while others suggest reduced solidarity.

The core problem addressed is the shift in the "social contract." Specifically, the authors examine whether the pandemic has eroded traditional, unconditional social norms that previously drove tax compliance and redistribution in Japan, replacing them with a system where support for redistribution is strictly conditional upon trust in government institutions. The study focuses on the micro-level dynamics of how income levels and institutional trust interact to shape acceptable tax rates in a post-crisis society.

2. Methodology and Data

Data Source

• Dataset: A large-scale, repeated cross-sectional dataset collected via internet surveys by Nikkei Research Company (NRC).
• Timeframe: 2018 (Pre-pandemic) and 2021, 2023, 2024, 2025 (Post-pandemic).
• Sample Size: Total $N = 18,181$ (5,929 pre-pandemic; 12,252 post-pandemic).
• Representativeness: Respondents were stratified by age (20–70) and prefecture to match Japan's national demographic structure.

Variables

• Dependent Variable: Tax Rate (Altruism). Measured by a hypothetical question asking respondents the maximum percentage of their income they would be willing to pay in taxes if the revenue were transferred directly to individuals earning significantly less than themselves (1–50% scale). This is a continuous variable censored at 1% and 50%.
• Key Independent Variables:
  • Trust: A 5-point Likert scale measuring trust in the government.
  • Income: Household income (continuous and categorized into quintiles).
• Control Variables: Gender, marital status, age (and age squared), years of schooling (and squared), and prefecture fixed effects (46 dummies).

Econometric Strategy

• Model: Two-limit Tobit model was employed to handle the censored nature of the dependent variable (bounded between 1% and 50%).
• Estimation Approach:
  1. Baseline: Separate estimations for pre- and post-pandemic sub-samples to compare coefficient shifts.
  2. Interaction Analysis: A full factorial interaction model was used to examine the nexus between income quintiles and government trust levels.
  3. Visualization: Coefficient plots and predicted marginal effects were generated to visualize the "widening gap" between high-income/high-trust and high-income/low-trust groups.

3. Key Hypotheses

1. Hypothesis 1 (Erosion of Unconditional Norms): Pre-pandemic, redistributive support was high and independent of trust levels (anchored in social norms). Post-pandemic, the baseline support has shifted downward.
2. Hypothesis 2 (Conditional Cooperation): Post-pandemic, support for redistribution has become "conditional." Individuals with high government trust maintain high support, while those with low trust significantly withdraw support, particularly among high-income earners.

4. Key Results

Descriptive Statistics

• Decline in Altruism: The mean acceptable tax rate dropped significantly from 13.2% (pre-pandemic) to 10.2% (post-pandemic).
• Stability of Trust: Mean levels of government trust remained relatively stable, suggesting the change in tax attitudes is not due to a drop in average trust, but a change in how trust functions.
• Inequality: The standard deviation of income increased, confirming rising income inequality during the pandemic.

Econometric Findings

• Shift in Income Effect: Pre-pandemic, income had no significant effect on tax willingness (suggesting a universal norm). Post-pandemic, a positive correlation emerged: higher-income individuals were willing to accept higher tax rates, but only under specific conditions.
• The Role of Trust:
  • Pre-Pandemic: The relationship between trust and tax willingness was flat; high trust did not significantly increase tax willingness compared to neutral trust.
  • Post-Pandemic: A steep positive slope emerged. High trust (levels 4–5) became a strong predictor of high tax willingness.
• The "Widening Gap" (Interaction Effect):
  • High-Income/High-Trust: This group showed the highest acceptable tax rates post-pandemic (approx. 12%), surpassing pre-pandemic levels.
  • High-Income/Low-Trust: This group showed a dramatic collapse in support, dropping to approx. 6% (a 5% decline from pre-pandemic levels).
  • Conclusion: The gap between high-trust and low-trust high-income earners has widened significantly. The affluent have become "conditional cooperators"—willing to redistribute only if they trust the state to manage it effectively.

5. Key Contributions

1. Measurement Innovation: Unlike previous studies using Likert scales for general redistribution preferences, this study uses a hypothetical tax rate directly linked to transfers to the poor. This provides a more concrete measure of "altruistic willingness to pay."
2. Structural Shift Identification: The paper provides empirical evidence that the pandemic acted as a structural break, transitioning the social contract from unconditional social norms (based on interpersonal harmony) to conditional institutional trust.
3. Mechanism of Polarization: It identifies that the erosion of face-to-face social capital (due to lockdowns) weakened the traditional Japanese norm of unconditional altruism, making the system reliant on institutional performance.
4. High-Income Sensitivity: It highlights that high-income earners are the most sensitive to institutional trust; they are the first to withdraw support when trust is low, posing a risk to the sustainability of welfare systems.

6. Significance and Policy Implications

• Social Contract Fragility: The findings suggest that in post-crisis societies, the "social bond" is no longer automatic. Redistribution relies heavily on the perceived efficacy and reliability of the state.
• Policy Risk: If government trust erodes, the wealthy (who have the capacity to fund redistribution) will withdraw support precisely when inequality is highest, creating a vicious cycle of instability.
• Restoring Solidarity: To rebuild the social contract, policymakers cannot rely solely on traditional norms. They must focus on institutional transparency and policy effectiveness to restore the "conditional cooperation" required to fund social safety nets.
• Global Relevance: While focused on Japan, the findings align with global trends of political polarization, suggesting that the shift from norm-based to trust-based altruism is a broader phenomenon in modern crisis management.

In summary, the paper argues that the COVID-19 pandemic did not just increase inequality; it fundamentally altered the psychological and institutional mechanisms of social solidarity, replacing a stable, norm-based system with a fragile, trust-dependent one.