The Dynamic Effects of Wealth Taxes

California’s Proposition 40 would levy a 5% tax on the wealth of the state’s richest residents. The static arithmetic is straightforward. The dynamic arithmetic is not.

Bond yields have gone up worldwide due to many factors, but most importantly due to inflation concerns resulting from the across-the-board tariffs, the US-Israel-Iran war, and the fiscal deficits that most developed governments continue to run (at record levels, even when measured as a percentage of GDP). Higher yields mean higher borrowing costs for governments at exactly the moment when inflation-squeezed voters are demanding relief from local taxes. Governments are therefore hunting for new sources of revenue, and California has produced one of the most ambitious proposals: Proposition 40, a wealth tax.

Proposition 40 is one effort to mitigate the effects of inflation. California is running a substantial budget deficit, so the goal of the proposition is to reduce healthcare inflation for the average California taxpayer via redistribution rather than through additional deficit spending. The mechanics are simple: a one-time 5% tax on the wealth of the state’s richest residents, with the proceeds directed at healthcare costs. On paper, the state raises billions without borrowing a dollar.

While Proposition 40 is written as a one-time 5% tax, it will inevitably become a permanent one. This is the nearly universal experience with taxes that are introduced as temporary measures. The classic American example is the “temporary” telephone excise tax enacted in 1898 to fund the Spanish-American War; it was still being collected in 2006. A more directly relevant example is itself a wealth tax. Spain introduced its Solidarity Tax on Large Fortunes in 2022 as an explicitly temporary measure covering 2022 and 2023, then extended it indefinitely by royal decree at the end of 2023; Spanish practitioners now describe it as a structural feature of the tax system with no end date. Once a revenue stream exists and a budget has grown to depend on it, the political cost of letting it expire is far higher than the political cost of quietly extending it. Any analysis of the proposition should therefore treat it as what it will become: a recurring annual levy on wealth.

The tax only looks like a good solution, however, if one is myopic and does not consider the resulting moves that the individuals hit with the tax will make. Static analysis takes the tax base as fixed: multiply the stock of taxable wealth by 5% and collect the proceeds. But wealthy taxpayers are not a fixed tax base. They are the most mobile, most advised, and most tax-sensitive investors in the economy, and they will respond. It is these responses, the dynamic results of the tax, that determine whether the state actually ends up better off.

Much has been written about the wealthy moving out of California, and that is the response most commentators reach for first. But it is not the point of this article. California has natural advantages, its climate, its universities, its concentration of technology and entertainment, that will make many wealthy people stay even at a considerable tax cost. The more interesting question is what moves the individuals who stay would make. They cannot change their residence, but they can change their portfolios. And one portfolio change in particular has direct consequences for the state budget that the wealth tax is meant to repair.

Due to its budget deficit, California has issued a significant amount of municipal debt, approximately $600 billion, over half a trillion dollars, across the state and its municipalities and agencies. Like all states, California depends on continuous access to the municipal bond market to refinance maturing bonds and to fund new spending. The interest rate it pays on this debt is therefore not an abstraction; every basis point moves the budget.

Municipal debt is hugely tax advantaged. The interest is tax exempt at the federal level and, for California residents holding California bonds, exempt from state income tax as well. For wealthy individuals in the top brackets, that means a saving of over 50% on their taxes: 37% at the federal level and 13.3% at the state level. A California municipal bond yielding 3% is therefore equivalent to a taxable bond yielding roughly 6% for a top-bracket Californian. This is precisely why municipal yields can sit so far below Treasury yields and still find eager buyers.

But notice who those buyers are. Because the state tax exemption applies only to California residents, California debt is held disproportionately by Californians — a pattern documented across state municipal markets for four decades. And wealthy California residents get the most tax savings from holding California municipal debt because the exemption is worth the most to those facing the highest marginal rates. As a result California bonds are disproportionately held by wealthy individuals. The concentration is stark, and it has been getting starker. The most comprehensive study of household municipal bond ownership finds that the wealthiest 0.5% of households hold 42% of all municipal debt held by households, up from 24% a generation ago, while the share of households owning any municipal bonds at all has fallen from 4.6% to 2.4%. The buyer base had been narrowing toward the very wealthy long before anyone proposed taxing them. The marginal buyer of California’s debt, the buyer whose willingness to pay sets the price, is in most cases exactly the person Proposition 40 proposes to tax.

Exhibit 1

The buyer base has been narrowing for a generation

Bergstresser and Cohen, Hutchins Center Working Paper #20, Brookings Institution (2016), using Federal Reserve Survey of Consumer Finances data. Convexity computed the equivalent figures from the 2016, 2019 and 2022 surveys; the differences from 2013 fall within the survey's margin of error.

Now consider the arithmetic facing that buyer once the wealth tax is in place. If one’s wealth is being taxed at 5% per year, then one has to hold assets that provide greater than a 5% post-income-tax return simply to keep one’s wealth from shrinking. The average yield on California municipal debt is around 3% for a 10-year maturity post-tax (for California residents). Even the longest maturity California municipal debt, at 30 years, yields only 4.3%. Every California municipal bond on the curve yields less than the wealth tax takes. An asset class that was the single most tax-efficient holding for a wealthy Californian becomes, under a wealth tax, a guaranteed way to lose wealth every year.

Therefore, wealthy individuals will be disincentivized from holding California municipal debt. This includes the wealthy people who do move out of California, who lose the state tax exemption the moment they establish residency elsewhere and so have little reason to favor California paper over any other state’s; as well as the wealthy who remain in California, for whom the bonds now yield less than the tax consumes. The natural buyer base for California municipal debt, on both sides of the migration decision, shrinks at once. And with the wealthiest 0.5% of Californians holding nearly half of the municipal debt, even a small reduction in the buyer base will have an outsized impact on California muni bond yields.

It is worth being precise about what happens next, because nothing dramatic needs to occur for the effect to bite. The wealthy do not need to dump their existing holdings in a panic; they simply need to stop showing up for new issues and to let maturing positions roll off without reinvesting. California and its municipalities come to market constantly, refinancing old debt and funding new projects, and each auction now finds fewer of the buyers who were previously willing to accept the lowest yields. When the natural buyers of a bond step away, the bond does not go unsold; it gets repriced. Yields must rise far enough to attract the next tier of buyers, out-of-state individuals, funds, and institutions that receive less or none of the tax benefit and therefore demand more yield. This likely will result in California bond yields moving up by about 50 basis points. On California’s existing $600 billion of debt, that would mean an additional interest expense of about $3 billion per year for the state.

The market, it turns out, is not waiting for the tax to arrive. Since the start of the year, as Proposition 40 has moved toward the ballot, California municipal spreads have already widened by 38 to 45 basis points. In other words, most of the repricing described above has already happened before a single dollar of the tax has been collected. Markets price expected policy, not enacted policy, and bondholders are clearly assigning a high probability both to the proposition passing and to the tax becoming permanent. What was a forecast a paragraph ago is, for the most part, already an observation.

Now set that figure against the revenue side. According to the California Legislative Analyst’s Office, the wealth tax will bring in only about $3 billion per year. The additional interest expense caused by the tax is approximately equal to the revenue raised by the tax. The state would be collecting $3 billion with one hand and paying an incremental $3 billion to bondholders with the other, with nothing left over for the healthcare relief the proposition promises. Even if bond yields do not go up by as much as 50 basis points, they will go up substantially, and thereby offset at least some, if not all, of the gains from the wealth tax. Also there is an equally likely chance that yields go up by more than 50 basis points. And note the asymmetry: the wealth tax revenue depends on the wealthy staying put and holding still, while the interest expense arrives automatically, priced into every new bond the state sells.

Exhibit 2

Revenue in, interest out

Illustrative. The interest cost applies the repricing scenarios to California’s outstanding municipal debt; revenue figure per the California Legislative Analyst’s Office.

Even the $3 billion figure understates the full cost, because it covers only the existing stock of debt. Higher yields also reprice every bond California will sell in the future. Every school, water project, road, and transit line that the state and its municipalities finance from here forward will carry the higher rate for the life of the bond. The wealth tax therefore does not merely offset its own revenue; it raises the hurdle rate on all future public investment in the state, quietly shrinking what the government can build with any given budget.

There is also a compounding dynamic at work. Recall that Proposition 40 exists because California is running a substantial budget deficit. Higher interest expense widens that deficit, which increases the pressure for more revenue, which invites more taxes, which push yields higher still. A tax meant to close a budget hole can instead set off a spiral in which each round of revenue-raising increases the cost of servicing the very debt that made the revenue necessary in the first place. This is the fiscal version of a doom loop, and a state, unlike a country, cannot print its way out of one.

Exhibit 3

The doom loop

Schematic representation of the mechanism described above. No underlying data.

This is exactly how markets work. When marginal costs increase in one place, other prices adjust to the point where marginal costs equalize again. A wealth tax raises the cost of holding low-yielding assets; the price of those assets falls, and their yields rise, until the marginal holder is again indifferent. The municipal debt market and taxes are linked; when marginal costs go up via taxes, the marginal cost of municipal debt goes up with them.

California is not running this experiment blind; Europe ran it first. In 1990, twelve European countries levied an annual wealth tax. Today only three do. Austria, Denmark, Germany, the Netherlands, Finland, Iceland, Luxembourg, Sweden and France all abandoned theirs, and for the reasons on display here: the dynamic responses of the taxed shrank the base, valuation and collection costs consumed the proceeds, and actual revenues never approached the static estimates. France, the most studied case, repealed its wealth tax in 2018 after decades in which the outflow of taxpayers and capital was widely judged to have cost the state more than the tax ever raised. The countries that kept a wealth tax did so only at low rates, far below the 5% Proposition 40 contemplates.

Exhibit 4

Europe ran the experiment first

Perret, Fiscal Studies 42 (2021); OECD, The Role and Design of Net Wealth Taxes in the OECD (2018). Counting European countries; the OECD’s wider total of four includes Colombia, which joined in 2020.

The same lesson applies to the many other states watching California’s experiment. Every state that levies an income tax has created the same structure: a municipal bond market whose best customers are its own wealthy residents, held in place by a tax exemption that is valuable in exact proportion to the holder’s marginal rate. A wealth tax anywhere in that structure attacks the state’s own cost of capital. The larger the state’s debt load relative to the expected wealth tax revenue, the worse the trade becomes, and California, with roughly $600 billion outstanding against $3 billion in projected annual receipts, is running the experiment at two hundred times leverage.

None of this required exotic modeling to foresee, only the willingness to ask what the taxed individuals would do next. Static revenue estimates will always flatter a wealth tax, because the tax base appears large and captive. The dynamic effects tell a different story: the same wealthy residents who would pay the tax are the ones funding the state’s borrowing, and a tax that drives them out of the state’s own bonds sends the money back out the door as interest expense almost as fast as it comes in. That is the dynamic effect of a wealth tax, and it is the effect on which Proposition 40 should be judged.