When considering the burden of taxes and social insurance contributions, is it enough to look only at tax rates and deductions? Does government spending create more value than leaving that money with taxpayers? As a tax accountant, I believe I must also address that question.

I support smaller government. I want to reduce spending whose necessity or results cannot be explained and leave more money for households and businesses to allocate as they choose. But having a clear position is different from selecting only convenient figures. I will use international comparisons to examine whether countries with smaller governments really do have higher productivity.

How should we measure the size of government?

Here, I use general government expenditure as a share of GDP. General government includes the national government, local governments, and social security funds. Comparing national general-account budgets alone would be misleading because countries differ in how they account for local government and social security. Expenditure includes transfers such as pensions and benefits, so a spending ratio of 50% does not mean that government itself produces half of the economy’s output.

Government spending and tax revenue are also different. Under the OECD’s national-accounts framework, revenue includes taxes, social insurance contributions, and other income; the excess of expenditure over revenue is a fiscal deficit. A government that continually runs deficits spends more than the revenue it collects that year. Looking only at tax revenue understates the scale of its spending.

For productivity, I use GDP per hour worked, which measures value added per hour of work. The figures use purchasing power parity to adjust for differences in price levels. This is not a score for workers’ diligence or intelligence. Equipment, technology, industry mix, management, and institutions also affect it.

What do the figures show?

I aligned the figures for 2023 using published OECD data. Using only the latest available year would leave gaps for some countries or indicators. The chart shows 36 countries, excluding Türkiye, for which the spending and productivity figures did not align, and Chile, whose retrieved productivity figure appeared anomalous and remains unverified. The table highlights major countries used for comparison with Japan. Each country receives equal weight; this is not a comparison weighted by population or economic size.

Do smaller governments have higher productivity? 2023 · 36 comparable OECD countries / one point per country ↑ GDP per hour worked (USD, current PPP) 0 30 60 90 120 150 180 20% 30% 40% 50% 60% Denmark Finland France Germany Ireland Japan South Korea Luxembourg Norway Sweden Switzerland United States General government expenditure / GDP → Source: OECD. This compares country levels; it does not estimate the effect of spending cuts. Government statistics for Australia and NZ use fiscal years. Note multinational and industry effects in Ireland and elsewhere.

Japan and the United States stand out. Government spending is about 39% of GDP in both countries, but productivity is $59.2 per hour in Japan and $95.6 in the United States. Switzerland has a smaller government and higher productivity than Japan. Yet Denmark and Sweden have larger governments and higher productivity. Some countries combine a small government with high productivity; others combine a large government with high productivity.

The simple correlation coefficient across all countries in the chart is 0.02. It describes the strength of a linear relationship, not the causal effect of a policy. If Ireland, Luxembourg, and Norway are excluded for reference because multinational firms and the financial or resource industries have a large influence on their figures, the coefficient is 0.38. The result therefore changes with the set of countries. We cannot adopt only the more convenient figure as our conclusion.

Separate OECD tax statistics put taxes plus compulsory social insurance contributions at 33.7% of GDP in Japan and 25.6% in the United States. Similar spending shares do not mean that the burden collected or the composition of deficits is the same in that year. Those tax statistics already include social insurance contributions in total tax revenue, so adding them again would double count them. Japan’s tax statistics are on a fiscal-year basis; this ratio cannot simply be subtracted from calendar-year spending to calculate a deficit.

Sources: OECD general government statistics, OECD productivity statistics, and OECD tax revenue statistics. Retrieved September 8, 2026. Figures are rounded for display.

Correlation alone cannot answer a policy question

The chart compares countries’ positions in the same year. It does not estimate how much productivity subsequently rose in countries that reduced the size of government. The reverse explanation is also possible: wealthy, productive countries may be able to sustain large social security systems. If GDP falls in a recession, government spending as a share of GDP rises even if the amount spent is unchanged.

Furthermore, shifting publicly provided healthcare or pensions to the private sector may lower the government spending ratio without reducing households’ premiums or costs by the same amount. We need to ask whether a change in the system really reduces total costs and improves choice and services. The number of public employees, the strength of regulation, and the presence of state-owned enterprises also cannot be captured by the spending ratio alone.

Therefore, we cannot simply adopt another country’s government spending ratio as Japan’s target. A chart comparing the size of governments is a starting point for testing a hypothesis. To identify programs worth cutting, we must examine what they do.

What questions did Keynes and Hayek ask?

Chapter 3 of Keynes’s The General Theory of Employment, Interest and Money argues that when demand for goods and services is insufficient, enough jobs may not be created even for people who want to work. This concern leads to policies that use fiscal support when private demand weakens. It should not be taken as unconditional justification for permanent deficits or the continuation of every public works project.

Hayek’s “The Use of Knowledge in Society”, by contrast, argues that knowledge needed for economic activity is dispersed among people and workplaces, and that prices help coordinate it. Who knows which ventures will succeed next? How are mistaken judgments corrected? This perspective raises questions about governments continually choosing whom to support.

I give weight to Hayek’s perspective. Are people and capital being kept in place to protect incumbent companies and projects? Has passing a government review become more important than winning customers? To restore Japan’s economic strength, I want to revisit these questions. That does not mean dispensing with competition law enforcement or measures against monopolies.

Spending to cut and functions to retain

A 2016 study by Fournier and Johansson uses long-term data from OECD countries and associates larger government with lower growth, while finding that the relationship depends on government effectiveness. It offers support for my concern, but also shows why spending totals alone cannot settle the question.

Unlike the earlier chart of productivity per hour, the study analyzes potential GDP per person aged 15–74 and spending adjusted for the economic cycle. It makes adjustments for differences among countries, but it does not establish how much growth Japan would gain by cutting a particular number of trillions of yen today. In education, too, outcomes and efficiency matter alongside budget size, and individual programs need scrutiny.

The IMF’s 2025 statement on Japan recommends ending poorly targeted subsidies, particularly energy subsidies, while maintaining high-quality public investment. The statement also discusses revenue-raising measures, so it does not share my support for tax cuts. Still, its call to review the quality and purpose of spending, rather than protect it across the board, is instructive.

Judge public-private investment funds by their actual results

Alongside the question of whether the government can keep borrowing, we must ask what it spends the money on and what results it obtains.

Some government investments warrant review based on their record. According to a Ministry of Finance document published in July 2026, the Cool Japan Fund had a cumulative loss of ¥54.0 billion as of March 2026. That was ¥11.4 billion more than the planned loss of ¥42.6 billion, and it recorded a net loss of ¥15.7 billion in fiscal 2025 alone. The document states that new support decisions would be suspended and that a review would proceed on the premise of integration with another institution or abolition.

The Board of Audit’s 2025 report also found cumulative losses at the end of fiscal 2023 in 14 of the 23 funds whose profits and losses it examined. The other nine had cumulative profits; one of those nine had ended its operations in fiscal 2018. Because the funds were established at different times and are at different stages of recovering investments, the figure does not mean that “14 ultimately failed.” The report also notes inconsistent methods for recording performance indicators.

Moreover, JIC’s published record for the former INCJ reports profits exceeding ¥1 trillion across 144 investments. It would be wrong to say that no public-private fund has produced results. Yet profits and investment returns alone do not tell us whether government involvement achieved more than private investment would have. We need to examine profits and losses, whether the funds attracted additional private capital, and the need for a government role separately, rather than assume that programs repeatedly missing their plans should survive.

Make smaller government a reform that expands taxpayers’ choices

Reforming toward smaller government takes more than reducing spending. We need to review rules that block entry so new firms can more readily win customers and funding; put protected incumbents and unprotected new firms on fairer competitive terms; and help workers move to new jobs when businesses close. Only with these measures together does it become plausible that capital and talent will move toward creating new value.

Would cutting taxes first leave the government short of money and naturally force it to cut spending? A 2009 study by Romer and Romer examines policy changes in the postwar United States and finds little evidence for that account. It may not apply unchanged to Japan, but we cannot expect tax cuts alone to shrink spending automatically. We must decide specifically what to discontinue.

The same savings from spending cuts cannot be used in full for both tax cuts and deficit reduction. We need to specify how necessary services will be maintained and how debt sustainability and burden relief will be prioritized and apportioned. We also cannot ignore how to weather a short-term fall in demand. It would be inappropriate to promise that private investment will immediately rise by the same amount.

The government must explain not just how much it collects, but how its use of the money creates more value than leaving it in private hands. Spending that produces little benefit and obstructs private initiative should end. The burden of taxes and social insurance contributions can then be reduced, giving taxpayers more room to decide for themselves how to use their money. That is the smaller government I support. International comparisons alone do not prove this policy is right, but I believe comparisons and reviews of individual programs make it a reform worth pursuing concretely.