Nearly every developed economy taxes corporations, and the design of these taxes affects economic efficiency and growth. The standard approach across countries is a tax on corporate profits after various deductions (notably, a deduction for interest expenses) are considered. However, there is no reason to think that this approach is optimal and, in particular, the interest deduction has been criticized on the grounds that it encourages excessive use of debt financing. This raises a question: how should corporate taxes be structured? Equivalently, which firms should bear the burden of corporate taxes?

In a world without financing constraints, money would flow freely across firms, so it would not matter which firms are taxed. In reality, some firms are constrained and have valuable investment opportunities that they are unable to undertake, while others have ample internal resources and have exhausted their valuable investment opportunities. If the burden of taxation falls on the constrained firms, this will further reduce their investment and overall economic growth. In contrast, if the burden falls on firms that are not constrained, it will have no effect on investment and growth. Thus, if the government seeks to minimize the impact of taxes on investment and growth, the corporate tax system should be designed to tax only unconstrained firms.

This, however, is difficult to achieve. The government does not know which firms are constrained and which are unconstrained. Firms will report that they are constrained if doing so will allow them to avoid taxes, regardless of whether they actually have valuable investment opportunities. A profit tax fails in this regard. Some firms will be highly profitable and at the same time need even more cash to take advantage of their investment opportunities. Other firms will have low profits and even worse investment opportunities. A profit tax would fall more heavily on the former than the latter, thus reducing investment.

However, these two types of firms differ in another respect: what they do with their cash. A highly profitable but constrained firm will reinvest all its cash and avoid paying dividends or conducting share buybacks. A less profitable and shrinking firm, in contrast, will return its cash to shareholders. The decision to pay dividends or repurchase shares reveals something that profits alone do not: whether a firm considers its cash better spent outside the business.

Financially constrained firms, facing profitable investment opportunities and lacking external financing, are unlikely to pay money out to shareholders. Those with more resources than they need are more willing to return money to shareholders. That choice—whether or not to make a payout—serves as a natural signal of whether a firm is financially constrained.

Thus, unlike a profit tax, a payout tax targets only unconstrained firms. Instead of taxing a firm on earnings regardless of how they are used, a tax is imposed only when funds are distributed. The burden falls, by design, on the firms best able to bear it. Importantly, the government does not need to discover which firms are constrained and which are not. That information is revealed through each firm’s behavior. 

To gauge the magnitude of the advantage of payout taxation over profit taxation, we compare two tax systems, each calibrated to raise the same amount of revenue. The difference is notable. A switch to a payout-based tax increases the total value of the corporate sector—accounting for both existing firms and new entrants—by approximately 7%. That increase reflects the higher productivity and profitability of firms operating under fewer financial distortions. Such a tax code also facilitates more entry, as it is more favorable to young firms that typically retain all their earnings in early stages.

The practical implementation of such a system does not require a radical departure from current practice. One way to approximate a payout-based system is to adjust the definition of taxable profits: retained earnings could be made deductible, meaning that only cash paid out to shareholders is included in the tax base. Firms would continue to report profits as usual, but only the portion they choose to distribute would be subject to tax.

It is essential, however, that the government maintain a constant rate of payout taxation over time. Taxing payouts today encourages firms to hoard cash; taxing payouts in the future encourages firms to payout cash today. Under a constant payout tax, these two effects offset, and firms will decide to payout their cash based only on whether it is needed for investment or precautionary reasons.

Intriguingly, this implementation justifies a deduction for interest expenses. Specifically, it points out that the mystery is not why interest is deductible, but instead why firms are taxed when they reinvest profits. Our theory justifies a deduction for interest expenses but not payouts to shareholders by highlighting a key distinction between debt and equity: debt payments are mandatory and made to investors who usually do not control the firm, whereas payouts to equity are optional and made to investors who control the firm. Only the latter serves as a signal of whether the firm is constrained; paying interest on the firm’s debt signals only that the firm is not bankrupt.

This implementation also treats all forms of payouts identically. Dividends and share repurchases face the same tax treatment. Otherwise, firms will have strong incentives to favor one method over the other, purely for tax reasons. Such avoidance behavior creates inefficiencies of its own. A well-designed system would apply a uniform tax to all cash distributions to shareholders, regardless of form.

Some broader considerations remain. One consideration is timing. Because firms can defer payouts, a tax on distributions may result in delayed revenue collection, especially if companies choose to retain earnings in the short run. Over time, however, distributions will resume once firms accumulate excess funds. The overall present value of tax revenue remains comparable, but the timing may shift, requiring additional borrowing on the part of the government as it shifts from profits-based to payout-based taxation.

Critically, the approach does not aim to correct other distortions in corporate behavior. For example, it does not address concerns about executive compensation, empire-building, or inefficient investment driven by managerial incentives. Those are separate policy challenges. The focus here is narrowly on raising revenue to minimize disruption to investment and entry, especially for the firms most in need of internal funds.

The broader message is that corporate taxation need not treat all firms equally to be fair or efficient. Financial conditions vary, and tax policy can be designed to reflect that variation without becoming complex. When firms voluntarily part with cash, it is usually because they have met their investment needs. Taxing distributions at that point captures surplus rather than squeezing essential capital. And when firms choose to hold onto cash, that decision serves as a signal that funds are still needed—whether for operations, expansion, or future uncertainty.

By relying on observable behavior rather than unobservable characteristics, a payout-based tax system offers a way to raise revenue while respecting the informational limits facing tax authorities. It also aligns well with how firms operate. Cash is precious when growth opportunities are high and financing is scarce. It becomes expendable only once a firm is confident that other uses have been exhausted.

Under the assumption that governments must raise tax revenue in order to perform their functions, how taxes are levied matters greatly. A system that focuses on what firms pay out—rather than what they earn on paper—can better align the tax burden with firms’ financial flexibility. By taxing cash that leaves the firm, rather than cash that might still be needed inside it, such an approach offers a more targeted and potentially less distortionary way for the government to raise revenue.