The election of François Hollande as President of the Republic in 2012 was followed by major reforms in the taxation of capital, aimed at bringing it closer to the taxation of earned income. Among these reforms, the dividend tax rate for LLC entrepreneurs was brought into line with the wage rate, namely 46%. The increase is significant, as previously dividends were taxed at a rate of 15.5%.
Not surprisingly, right-wing parties and employers’ unions predicted an economic disaster associated with the announced collapse of investment. Dividends should reward providers of capital and reward the risk associated with investments, raising the tax rate should increase the cost of capital (shareholders demand higher returns to offset tax increases) and discourage entrepreneurs from investing.
Seven years later, it is possible to sum up this reform, which we recently did in a research article with Charles Boissel, using French administrative data, which allows us to study many companies in France. The conclusion is clear: this increase in taxation of dividends has not had a negative impact on the economy. On the contrary, it has led to increased investment and employment!
How can we demonstrate this? Based on the fact that the reform only affected limited liability companies. Thus, comparing the behavior of SARL with companies with other legal statuses (in particular, simplified joint stock companies or AAC), before and after the reform, allows us to assess the “causal” effect of tripling the tax rate on dividends for LLC entrepreneurs.
Less dividends, more investments
This comparison shows us that since the reform, the SARL entrepreneurs have significantly reduced their dividend payments, by about 17%. This effect is the most expected. As the tax rate increases, entrepreneurs will choose to keep more cash in the company rather than pay out dividends.
What are companies doing with this new money? We found that, on average, entrepreneurs reinvest 0.3 euros for every euro not distributed as dividends. This means a significant increase in investment!
This increase in investment, driven by increased taxation of dividends, results in a better allocation of capital.
However, this increase may be partly counterproductive if these additional investments are focused on projects with low productivity.
If we take a closer look at which companies are reinvesting, on the contrary, we find that only companies with growth opportunities and the most productive companies invest and grow more. This increase in investment driven by the increase in dividend tax is thus reflected in a better distribution of capital: companies that are able to produce more wealth per euro of capital invest and grow faster than companies that are similarly affected by a higher dividend tax rate. but less productive.
Of course, it can be assumed that, despite these average positive effects, some companies suffered from the reform. Our ability to look at all companies in France allows us to pinpoint several subgroups that economic theory predicts will suffer: companies that are more likely to use equity financing, younger companies, more or less large ones. However, we do not see anywhere that an increase in the tax rate on dividends has led to a drop in investment.
If one third of retained dividends are reinvested, what do companies do with the remaining two thirds? We found that “tax avoidance” behavior, such as when a company pays for personal consumption (such as buying a “company” car for purely personal use), does not increase.
In fact, the remaining two-thirds of undistributed and uninvested dividends can be explained by an increase in cash and an increase in loans to the company’s customers. The latter no doubt partly explains why companies hit by a dividend tax increase end up growing faster than companies whose rate remains unchanged.
Our results refute the predictions of standard neoclassical models, but they can be explained by three reasons. First, entrepreneurs may underestimate future demand on average, especially during times of economic activity problems such as the sovereign debt crisis in Europe. As a result, they may keep cash levels too low to take advantage of investment opportunities when they arise.
Further, the relationship between tax increases and the efforts of economic agents is ambiguous. Indeed, if entrepreneurs want a fixed level of consumption despite tax increases, they will be forced to produce more and, in our particular case, invest more.
Finally, contrary to the standard assumption of most economic models, tax changes are not permanent, but are more likely to be perceived as transient. Political rotation has become a permanent feature of our democracies, and tax increases implemented by a government on the left are likely to be defeated by the next government on the right.
Consequently, entrepreneurs have an incentive to do what Anton Korinek and Joseph Stiglitz call “intertemporal arbitrage”: if I know that taxes on my dividends are high today but will decrease in the not too distant future, I am interested in lowering dividends and accumulating wealth within the company in during this period to pay me higher dividends when taxes are lowered again. This accumulation of wealth within the company can be achieved either by saving more cash or by investing when growth opportunities arise, which is what we are seeing.
Our results do not necessarily mean that higher taxes on dividends will lead to higher investments in the long run. Indeed, even if the next government does not reconsider a tax hike, entrepreneurs initially affected by the tax hike may get tired of waiting for tax cuts and decide to increase their dividends again, reducing the cash available for investment.
However, our study implies least that increased taxation of dividends does not harm investments and, conversely, can be used to support them during periods of economic downturn. And if, in the worst case, entrepreneurs simply decide to keep their dividends and the level of their investments constant, this will always generate additional income for the state to finance its public policy.