An alternative to the capital gains change that incentivizes productive investment

Ending the tax deductibility of interest paid on residential real estate investment would kill many birds with one stone

Won’t somebody think of the rich people?

In their framing of the proposal to increase the capital gains inclusion rate to 67%, the government chose their target group well: in the current climate of populist political mud-slinging, soaring housing prices, and increasing wealth inequality, “Won’t somebody think of the rich people?” is not a very effective rallying cry.

This is a very deliberate distraction from the actual issues that are raised by a capital gains inclusion tax rate increase, issues that should be of concern to everyone, not just those who are immediately affected.

I am a believer in the idea that one should not criticize the ideas of others unless one has a better idea to propose. In this article, I explore an alternative to the proposed tax change that achieves many things at the same time: it incentivizes investment in productive Canadian assets, improving productivity in the long run; it disincentivizes investment in Canadian real estate and exerts some downward pressure on housing prices by eliminating an existing regressive tax policy that allows fully tax-deductible mortgage interest on investment properties; it leads to about the same level of new tax revenue as is expected of the proposed capital gains hike while only targeting people hoarding residential real estate; and it actually pushes toward the stated budget 2024 goal of economic equity.

This can all be achieved with a single policy change, instead of increasing the capital gains inclusion rate: end the tax deductibility of mortgage interest paid on pre-existing residential investment properties.

Capital gains and the housing market

It should not be news to anyone who reads this blog that Canada has both a problem with underinvestment in domestic commercialization of IP, and a problem with the price and availability of housing. These problems are closely related, since capital allocation across asset classes is a competitive process. A complete lack of incentives for investment in productive Canadian innovation makes real estate relatively attractive, with a much safer (and in recent years, higher) return on investment for those who want to keep their money in Canada.

This runs counter to one of the basic tenets of Adam Smith-style capitalism, in which nearly the entire role of government is to police and eliminate “rent-seeking behaviour”. It also diverts money away from productive investment, investment in building actual production capacity in the form of Canadian companies that can address our lagging economic performance. Not only does a capital gains tax increase disincentivize productive investment, it indirectly incentivizes further investment in residential real estate, since that asset class affords tax breaks in the form of deductible mortgage interest, tax breaks the compound over the lifetime of the asset and that will, on a long enough time horizon, compensate for higher capital gain tax on sale.

According to CMHC, Canada will need 3.5M units of housing above and beyond what can reasonably be built by 2030. The capacity to build that fast simply does not exist. But building need not be the only solution, at least not to the affordability issue. There are many units tied up in real estate investment, of which a significant portion could be recycled into the supply with the right push. While this will not help with the capacity issues, it would certainly help relieve upward price pressure in the meantime.

Not all investments add value

The introductory discussion suggests a natural division of investment into three broad categories that can and should be treated differently for tax purposes.

Type 1: Directly productive investment: Investments that fall into this category provide the means to grow Canadian production capacity directly to Canadian companies. This includes opening your own business, angel and venture capital investment, and direct purchase of shares at IPO. I would also include in this category investment made in building new homes, given the desperate need to increase the housing supply.

Type 2: Indirectly productive investment: Most investments fall into this category, including purchase of company stock on the open market or through things like ETFs. These investments do not go directly to companies nor do they necessarily stay in Canada, but they do provide liquidity to public markets that enable Type 1 investments to eventually see a return.

Type 3: Unproductive Investment: Investments that tie up capital in assets that do not increase production capacity fall into this category, including purchase of pre-existing residential real estate, investment in REITs, and investment in GICs or similar interest-bearing instruments.

In times of lagging productivity, sound tax policy will disincentivize unproductive investment in favor of productive investment. There is precedent for this, on both ends. In British Columbia, for example, direct investors in early stage companies receive a tax credit on their investment that is widely viewed as a critical driver of early stage investment there. On the other hand, interest paid on assets like GICs have always been taxed as income rather than capital gains. The seeds of the policy that we need are already there.

In this categorization, pre-existing real estate sticks out like a sore thumb. It’s an unproductive investment, but it has tax breaks everywhere. Sale of an investment property counts as capital gains rather than income, and the interest paid on the mortgage used to finance an investment property is fully tax-deductible. These deeply regressive policies combine to make real estate a disproportionately favored asset class compared to productive investments in Canada and are a significant driver of the wealth inequality that the current government claims to be trying to address.

There is a fairly intuitive possibility that arises from this breakdown of investment classes: set a different capital gains inclusion rate for each one. While this is a perfectly reasonable approach to take, it does not distinguish between real estate investors hoarding multiple unnecessary properties and individuals who just have a house and a cottage. I would prefer a more surgical approach to the problem that focuses the impact only on those who are active contributors to the problem.

To that end, let’s explore what would happen if we ended the tax deductibility of mortgage interest on investment properties, while leaving everything else unchanged.

Back of the envelope math

Let’s do some Fermi Estimation. Our task is to calculate an order of magnitude estimate of the amount of tax that Canada forgoes each year by making interest paid on investment in residential properties tax deductible.

We’ll start with individual private ownership. According to data here from the 2021 census, about 1.6M Canadians owned multiple residential properties in 2021. This number has certainly increased since then, but in the absence of updated data we will use this as our baseline. Note that the vast majority of these properties are single-family homes and condos, not large apartment buildings.

A bit of work in Excel using the data here suggests that among property owners that own at least two residential properties, 76% own exactly two, 16% own exactly three, and 8% own four or more. Assuming the percentage breakdown by property number generalizes across Canada and splitting the difference, we will use as our baseline 1,200,000 investment properties held by single-investment property investors (which is to say, they own two properties of which one is their primary residence and one is income generating), 400,000 held by investors with two income-generating properties beyond their primary residence, and 135,000 with three income-generating properties (or more, but we will ignore that). These numbers are in the right ballpark of every source I can find on this, leaning low, and will suffice to give us a conservative estimate of the number of units currently tied up in residential real estate investment.

We can now estimate the total value of those mortgages. Note that the values given in the dataset here are assessed values, which is typically quite a bit lower than sale price, so the number are are going to get at the end will be an underestimate. Using the bottom of the range for the value of those properties to deliberately underestimate and doing a weighted average across the ownership categories in Table 3, I find that BC homeowners with two properties are carrying an average value of about $1.2M in total, while residential real estate investors with three and four properties are carrying average values of about $1.8M and $2.8M, respectively. For Ontario, those numbers are $0.7M, $1.0M, and $1.3M respectively. Safe to assume that the rest of Canada will be a bit lower, but most of the real estate speculation is happening in these markets so these are the prices that matter. Let’s assume that each investment property was purchased for about $400,000 on average.

Real-estate investors usually operate at maximal leverage, since this is when the largest tax deduction is achieved. Assume 80% of the value is mortgaged. This leaves one-investment property owners carrying $384B in debt for which interest is deductible (assuming one of the properties in a non-deductible primary residence), while two- and three-investment property owners are carrying $256B and $130B, respectively.

A quick calculation of mortgage interest at an assumed average rate of 3% suggests that the tax-deductible interest on those mortgages for the three groups annually breaks down as ($11.5B, $7.7B, and, $3.9B).

At an average marginal tax rate of 30%, probably also an underestimate for real estate investors, this is a total of $6.9B in foregone taxes on investment that is not building production capacity in Canada.

We can corroborate these estimates more directly with the data in the table here, which has estimates for investment properties by ownership type for Ontario, BC, Manitoba, PEI, New Brunswick, and Nova Scotia in 2021, broken down by investment property status and property type. Limiting ourselves to single-detached houses, semi-detached houses, row-houses, condo apartments, and properties with multiple residential units that are owned by individuals, we find 1.6M units with a combined assessed value of $541B. Rescaling that to the total population of Canada in 2021, we estimate about $737B in mortgage debt on investment properties held by individuals, which translates to about $6.6B in foregone taxes just based on assessed value, using the same assumptions as in the previous analysis.

In getting to this point, we haven’t even considered the foregone taxes on properties that are being hoarded by REITs and corporations. Data here and here suggest that a an amount of mortgage debt equivalent to about half of the individual investor total is being held by corporations, but a lack of granularity in the data make it difficult to get a good estimate. In any case, the numbers we have already are already more than enough to make the necessary point.

In other words, using publicly available information from the 2021 census, we can reasonably estimate that Canada forgoes at least $7B in taxes every year by making investment mortgage interest tax-deductible, actively subsidizing investment in an asset class that is unproductive, inflationary, and a direct driver of wealth inequality.

Strategic taxation

The proposed capital gains inclusion tax rate is estimated by the government to bring in $21B over five years, or about $4B/year. If instead the government phased out tax deductibility for mortgage interest paid on residential investment properties, the numbers above indicate that all else being equal, we could reasonably expect to bring in more than that.

In reality, such a move would trigger the sale of at least a portion of these properties as overleveraged real estate investors are forced to divest. According to the numbers above, there are about 2.5M housing units tied up in multiple-residential ownership. Sale of even a fraction of these units would contribute to relieving short term supply pressures, with the largest effects being felt in precisely the real estate markets that have suffered the most due to speculative real estate investment (Vancouver and Toronto), targeted at people who own investment properties and who by definition have significantly more resources than most Canadians. In other words, not only does this policy bring in at least the required amount of tax revenue to offset the projected revenues from the capital gains inclusion rate change, it even does the geographic and socioeconomic optimization for us. People who own just one primary residence, and even those with a primary residence and non-income generating secondary one like a cottage, will not be affected. Doctors, dentists, and other professionals who operate small businesses, otherwise getting slammed with capital gains tax, will only be affected if they are also multiple-property landlords.

Targeting increased taxes not at investment as a whole, but specifically at the type of investment that does not build up the Canadian business ecosystem, is far preferable to a blanket capital gains tax hike. Reducing or removing tax deductibility for interest paid on investment mortgages would more than cover the expected revenues from the proposed capital gains change, exert sell pressure on those hoarding Canadian housing, and begin to relieve the cost of living issue that has been spiraling out of control over the last few years. It addresses several key issues facing the federal government with a single policy change while avoiding disincentivizing investment in the Canadian innovation ecosystem. In the long run, a lower cost of living would actually make it easier for Canadian companies to attract talent.

With the capital freed by sale of a subset of investment properties, Canadian investors will need somewhere to put it that provides returns. With unproductive investments disincentivized, this leaves directly and indirectly productive investment option, as defined above. Canada could take a page out of Israel’s book and provide direct incentives to reinvest in Canadian innovation, using the proceeds of the proposed change to mortgage interest, or consider extending the BC-based angel tax credit country-wide, to further incentivize productive investment.

To allow the market time to absorb the shock, I would suggest implementing the change only on new mortgages signed after the change comes into effect (or on old mortgages after five years, in case there are longer mortgage terms in play). This will have the effect of grandfathering in the updated rules over the course of the next five years and give real estate investors time to either plan a graceful exit, or prepare for a higher tax burden. I would also maintain tax deductibility on loans relating to new builds, given the desperate need for new housing supply.

As a final thought: a capital gains hike, in addition to having a cooling effect on business investment, will drive innovators, doctors, dentists, and a host of other critical professions out of Canada. In the long run, investment is elastic and can easily move elsewhere. A hike on taxes on real estate investment, on the other hand, taxes an asset that literally cannot leave. The owner either pays the additional tax, or sells.

Both outcomes serve the stated goal of the 2024 budget without causing harm to other economic sectors. A capital gains inclusion rate hike does not. But it appears that the current government would rather mortgage the future to pay for the mistakes of the past than reconsider a deeply flawed policy initiative.