People respond to incentives. This is one of the most important economic principles, and in finance we add that companies respond to incentives. One major incentive for both companies and people is taxes. We think that people and companies will change their behavior if taxes change, because their incentives change.
While the idea makes sense, it is hard to prove. Ideally we would randomly select a group of companies, change their taxes, and change nothing else. Then we would have high confidence that any observed changes in behavior were caused by the tax changes.
But when governments actually change tax policy, they do not change taxes to help economists do research! Tax law changes are normally discussed, debated, and tweaked over a lengthy legislative process. Companies can see a tax change coming well in advance and change their behavior before anything is signed into law.
Furthermore, tax policy rarely makes changes for just a single group of companies and instead affects all companies, which makes a comparison to unaffected companies impossible. And to make matters even worse for us economists, the change usually affects taxation in hundreds or thousands of different ways. Linking any particular tax change to a change in company behavior is nearly impossible.
A Challenge Ahead
This creates a difficult challenge if we want to test our theory that companies will respond to change in tax policy. However, Professors Craig Doidge and Alexander Dyck of University of Toronto found a clever way to get around this problem with a natural experiment. They published it in their 2013 paper “Taxes and Corporate Policies: Evidence from a Quasi Natural Experiment.”
On October 31st 2006, the Canadian government made a surprising announcement. They were going to remove some tax advantages that a particular type of company called an income trust currently enjoyed under Canadian law. This was the only tax law change proposed, and it was surprising because it broke a key election promise of the newly elected government. Once announced, the tax change was enacted with alacrity; just one week later legislation was passed to remove the tax benefits.
The uniqueness of the situation gave Doidge and Dyck the perfect research opportunity. Instead of the tax change affecting all companies, only a specific group of companies saw their taxes change. Instead of thousands of tax changes, only a single change was signed into law. And instead of a long process that gave firms time to adapt ahead of time, it was a sudden massacre. When markets opened the next day after Halloween, all of the income trust companies saw their stock prices drop precipitously.
The companies’ response was also dramatic. At the time of the announcement in 2006, there were over 216 trusts in the Canadian market. By 2011, one year after the taxes went into effect, there were only 19 trusts left. Once the tax advantage was removed, no one wanted to structure a company as a trust.
Rapid Transition
Besides trust companies rapidly transitioning to different legal structures between 2006 and 2011, Doidge and Dyck also found that they began using other methods to reduce their tax liability as a substitute for the income trust status. For example, the companies increased their debt, because interest expense is deductible. The companies also made other adjustments to their dividends, cash holdings, and investments all likely in response to the new tax burden.
Ultimately, what Doidge and Dyck found suggested that companies do respond to tax incentives. While their work might have confirmed something we suspected to already be true, confirming our intuition is an important part of the scientific process.
So remember, remember the 31st of October, because incentives really do matter!
Eric McKee
Assistant Professor of Finance
West Texas A&M University
