Following the COVID-19 pandemic, governments are confronted with the catch-22 of boosting their economy while raising money to avoid increasingly runaway deficits. More than ever, understanding the effects of fiscal policies on economic growth and aggregate investment is crucial. Increasing dividend taxes, at least temporarily, might be a solution to deal with both problems.
When President Bush adopted the Jobs and Growth Tax Relief Reconciliation Act of 2003, reducing the top federal tax rate on individual dividends from 38.6% to 15%, he claimed that the tax cut would provide “near-term support to investment” and “capital to build factories, to buy equipment, hire more people”.
By contrast, left-wing politicians often claim that the amount paid to shareholders is too high, and that this money could be used more productively for investment and to pay workers. Writing in the New York Times in 2019, Senators Sanders and Schumer wrote that “when corporations direct resources [to payouts], they restrain their capacity to reinvest profits more meaningfully in the company in terms of R&D, equipment, higher wages.”
At the heart of this debate is the question of how an increase in payout taxes affects firm behaviour and the allocation of capital across firms, something we know surprisingly little about from academic research.
In a recent working paper (Matray and Boissel 2020), my co-author and I contribute to the debate by studying one of the largest single increases in the dividend tax rate in any developed country. Looking at the effects of a steep dividend tax increase in France, we find that entrepreneurs affected by the tax increase substantially reduced dividend payouts, leaving them with more liquidity on their balance sheets. This tax-induced extra liquidity led to higher investment, firm sales and value-added growth, and a reduction in capital misallocation across firms.
The French policy change
Following its electoral victory, the French centre-left party decided to raise the dividend tax rate for entrepreneurs in 2013, explaining that: “it is fair and legitimate to reward patient and productive investment (...) We want to incentivize investment rather than dividend payouts.”
The reform led to an almost three-fold increase in the dividend tax rate, from 15.5% to 46% for firms with a specific legal status. Ultimately, the increase affected 75% of all French firms.
Using administrative tax files that cover the universe of firms, we can track firm outcomes over the period 2008–2017 and estimate the effect of the reform by implementing a series of difference-in-differences estimators which compare treated and control firms before and after the 2013 reform.
A salient and substantial shock
The tax hike was a salient and substantial shock for treated entrepreneurs, who decided to cut their dividends immediately. Figure 1 reports the amount of dividends paid relative to the firm capital pre-reform for treated firms relative to control firms. While treated and control firms behave similarly before the reform, treated firms swiftly cut their dividends by an average of 15% relative to the pre-reform mean.
Figure 1 Effect of increase in the dividend tax rate on dividend payment
Notes: This figure plots the yearly coefficient capturing the differential effect of the increase in the dividend tax-rate on the amount of dividends paid by treated firms (SARL) relative to firms not affected by the tax-hike from a difference-in-differences estimator. Dividends are normalized by the firm total capital. ‘0’ is normalized to be the year before the tax-hike.
A positive effect on investment and cash holding
The large drop in paid dividends left treated firms with more cash in hand, which they used to increase their investment, accumulate corporate savings, and extend credit to their customers.
For every dollar of tax-induced drop in paid dividends, treated firms increased investment by $0.3, corporate savings by $0.6, and customer credit by $0.1. We report the increase in investment in Figure 2.
The increase in investment is large and implies an elasticity with respect to the tax rate of 40%. In other words, for every increase of 1% in the dividend tax rate, treated French entrepreneurs increased their investment by 0.4%. Such an increase in investment, together with the expansion of customer credit, helped firms facing the dividend tax hike to grow faster.
Figure 2 Effect of the increase in the dividend tax rate on investment
Notes: This figure plots the yearly coefficient capturing the differential effect of the increase in the dividend tax-rate on the investment of treated firms (SARL) relative to firms not affected by the tax-hike from a difference-in-differences estimator. Investment is normalized by the firm total capital. “0” is normalized to be the year before the tax-hike.
Higher dividend taxes improved the allocation of capital
While the average treated firm increases its investment, the reform might still lead to a reduction in overall output in the long run if capital misallocation increases. This would happen, for instance, if a minority of firms, important for long-term dynamics, face heightened financial constraints (e.g. Gourio and Miao 2010) or if the increase in investment is concentrated in firms with low return on capital or firms with only wasteful investment opportunities. We offer three pieces of evidence suggesting that this is not the case and that if anything, the higher dividend tax rate reduces the misallocation of capital.
First, we show that the tax-induced increase in investment is concentrated among firms facing new investment opportunities. We proxy investment opportunities by computing the leave-one-out mean of sales growth in the firm industry-by-local labour market post-reform. We then sort this measure of investment opportunities into terciles and find that treated firms increase investment more only when they face large investment opportunities. By contrast, treated firms with limited investment opportunities do not invest more than control firms.
Second, we conduct a similar exercise by sorting firms according to their marginal return on capital pre-reform to estimate how capital misallocation evolves after the reform following Bau and Matray (2020). We find that the tax increase leads to an increase in the investment rate that is three times as large for firms in the highest tercile of ex-ante marginal return on capital relative to firms in the lowest tercile. We also find that within treated firms, higher dividend payers pre-reform increase investment more if they are in an industry with high levels of capital misallocation pre-reform.
Taken together, these results strongly suggest that the average tax-induced increase in investment does not come from entrepreneurs engaging in income shifting or wasteful investment. If that were the case, we should find a similar increase irrespective of the local investment opportunities or the ex-ante level of marginal revenue product of capital (MRPK). By contrast, we show that entrepreneurs who decrease their paid dividends substantially after the reform do not expand their investment more relative to low dividend payers when investment opportunities are low, or when the expected returns to investment are low. These two results imply that entrepreneurs are not willing to waste their undistributed earnings.
Did the reform lead any firm to decrease their investment?
While the reallocation of investment across firms shows that the reform had a positive effect on economic efficiency, one group of firms could still face heightened financial constraints – namely, equity-dependent firms, which might decrease efficiency in the long-run.
Indeed, some economic theories predict that higher dividend tax rates would reduce investment for equity-dependent firms, as these firms have to issue new equity to finance their investment (e.g. Poterba and Summers 1983). The cash flow generated by the new investments is then used to pay back the new shareholders with dividends. Therefore, any increase in the dividend tax rate will increase the returns demanded by potential shareholders.
The richness of our data and the large sample size allow us to compute multiple proxies for the degree of equity dependence and to re-estimate our investment regressions for the sub-sample of firms most likely to be equity dependent. Irrespective of the proxy chosen, we fail to find any significant negative effects. This suggests that even among those firms most likely to face an increase in their cost of capital after the dividend tax hike, the reform has no negative impact on their investment.
Implications for our understanding of the behaviour of entrepreneurs
Overall, our results indicate that entrepreneurs in equilibrium are financially constrained because they pay ‘too much’ in dividends, leaving them with not enough cash on their balance sheet when investment opportunities appear. This suggests that, unlike the conventional wisdom in most economic models where agents are on average correct in their foresights, entrepreneurs have a hard time predicting their future investment needs and may sacrifice long-term development for the short term.
Bau, N and A Matray (2020), “Misallocation and Capital Market Integration: Evidence From India”, NBER Working Paper, No. 27955 (see also the Vox column here).
Gourio, F and J Miao (2010), “Firm Heterogeneity and the Long-Run Effects of Dividend Tax Reform", American Economic Journal: Macroeconomics 2 (1): 131-168.
Matray, A and C Boissel (2020), “Higher Dividend Taxes, No Problem! Evidence from Taxing Entrepreneurs in France’’, Working Paper.
Poterba, J M and L H Summers (1983), “Dividend taxes, corporate investment, and `Q'”, Journal of Public Economics 22 (2): 135-167.