History’s lesson for taxing wealth today

Proposals for taxing wealth, reforming estate taxation, and better measurement of billionaire fortunes are back on the policy agenda. Yet these debates rarely draw on historical experiences that show what wealth taxation looked like at scale, how well it measured the assets of the wealthy, or why wealth taxation ultimately retreated. This learning is overdue and matters as much for advanced economies as it does for developing countries, where property taxes remain the most underutilized fiscal tool and both revenue needs and inequality issues are pressing. In a new paper, we build the first comprehensive, high-frequency annual wealth series spanning more than one hundred years using a largely overlooked source: the records of the General Property Tax in the United States. From 1800 until the Great Depression, this tax aspired to cover all private property, including land, buildings, livestock, financial assets, and, before emancipation, enslaved people. Compiled from thousands of state auditors’ reports and Census Bureau publications, our analysis offers a window into how wealth was measured, taxed, and documented at scale, with lessons applicable today. When Income Data Fails, Wealth Data Steps In For context, the General Property Tax generated substantial revenue, averaging 4 percent of GDP over 1850-1940, at a time when wealth accumulation was growing rapidly (fig. 1). Federal income taxation as we know it was only established in 1913, so governments were financing themselves almost entirely through taxing property at the time. High-frequency property tax records reveal that private wealth grew from roughly three times GDP in 1800 to five times by the eve of World War I, punctuated by a collapse to an all-time low of 195 per cent of GDP at the end of the Civil War. The all-time peak came in 1932, in the depths of the Great Depression: at 580 per cent of GDP, it reflects output collapsing faster than asset values, a reminder that this ratio can rise for bad reasons as well as good. Today, private wealth is again above five times GDP after four decades of asset-price growth close to all time-highs. Wealth at this scale and volatility is precisely what income-based tax systems are poorly equipped to measure. Today, policymakers concerned about the wealth of the ultra-rich face the same structural problem as back then: tax records only capture income flows and not economic capacity of the richest individuals. Capital gains can go unrealized, business income can be deferred, and much wealth today sits in assets that generate no reportable income from real estate to private equity. As a result, income-based systems understate the economic power of those at the top. A direct tax on the stock of wealth, assessed periodically at or near market value, sidesteps many of these difficulties. The historical American experience shows that such a system can be administered at scale, even with nineteenth-century institutions and information technology. The challenge, as we describe below, lies in doing it well. The Central Problem Is Accurately Measuring Wealth The most important historical lesson from taxing wealth is the gap between assessed and market values. Local assessors, often elected and untrained, valued property well below market price, reducing revenue collected and creating incentives for non-compliance. As a result, the national average assessment (the share of market value captured in assessments) fell from about 83 per cent in 1850 to roughly 40 per cent by the early twentieth century, and far lower in some states. Under-assessment remains one of the most persistent problems in property taxation today, with well-documented under-valuation of high-value real estate in the US, the UK, and even worse cases in developing countries. Under-valuation can be regressive: if costly properties are assessed at a lower share than cheaper ones, the effective rate on the wealthy is lower. Recovering market values from administrative data demands investment in independent valuation. The nineteenth-century Census Bureau sent agents into the field, surveyed thousands of experts, and checked assessments against sale prices. Without that infrastructure, the base is whatever assessors say, usually too low for the assets of the wealthy. Comprehensive Wealth Statistics Are a Public Good Comprehensive data collection was a precondition for General Property Tax, not a by-product. The decennial Census wealth publications documented property values, exemptions, assessment practices, and legislative change in every state, letting governments track wealth and hold assessors accountable for regional disparities. The parallel to today is direct. The OECD, the EU, and many economists now call for national wealth registers, and several Nordic countries already maintain them. One usual objection is cost. But the cost of not collecting the data is being unable to know how wealth is distributed, how fast it grows, and who pays their fair share. The data also reveals something the tax debate rarely confronts directly: the persistence and growth consequences of wealth concentration. Wealth Inequality Is Persistent and Hurts Prosperity Our long-run wealth data shows that spatial inequality across US counties was historically high and never converged: the wealthiest 10 per cent of counties held around 70 per cent of total US property by 1930, and rank-rank correlations between county wealth in 1870 and 1930 stay above 0.6. Places that were wealthy after the Civil War were still overwhelmingly likely to be wealthy sixty years later, with more populous counties both wealthier and faster-growing. This persistence is stronger in wealth than income data, so income convergence can mask stubborn inequality in the stock of assets. Inequality within places also tracks slower growth. Counties with wealth more concentrated in the top decile in 1870 grew significantly more slowly over the next sixty years, even after controlling for geography, demography, and economic structure. Human capital is a key channel: more unequal places accumulated less education. Replicated across hundreds of counties within one country, this marks wealth concentration as a drag on long-run prosperity. A Lesson for Developing Countries These lessons are particularly relevant for developing countries, where property taxes raise only around 0.1 per cent of GDP in emerging Asia and Africa, and with property tax the most under-utilized tax in developing countries. The binding constraint is rarely the tax rate but the same measurement problem the nineteenth-century United States faced, only larger: assessment rolls are incomplete or decades out of date, the assessed-market gap can dwarf anything in our series, and cadastral records may not exist at all. The returns to investing in valuation and wealth statistics are high, and this infrastructure is a precondition for domestic revenue, not a luxury for after countries grow rich. Takeaways for Policymakers Nineteenth-century America was no model of sophisticated governance. And yet it still assessed, taxed, and documented private wealth at a scale many modern systems have abandoned. The tax had real flaws: uneven administration, patchy coverage of intangibles, pervasive political pressure. Yet it raised substantial revenue and left an extraordinary record of how wealth changed. Three messages stand out. 1. Income taxation alone cannot reach the wealthiest; a direct tax on wealth stocks, assessed at market value, is feasible and has real advantages. 2. High-quality wealth assessment matters most and can be achieved: without independent valuation and the statistics behind it, a wealth tax is only as good as local assessors allow. 3. Comprehensive wealth data is a public investment that pays off beyond taxation, helping governments understand inequality, target policy, and hold economic power to account.