Principles Of Canadian Municipal Finance (And Why A Land Value Tax Is Inferior)

The funding of Canadian municipal governments is not normally thought of as interesting topic; even Canadian fixed income investors are not particularly excited about it. However, there are two side issues that are of general interest. The first question is: what happens to Canadian municipalities if the housing bubble pops? (As a spoiler, not very much.) The second question is the feasibility of a Land Value Tax (LVT) which is a concept that gets some people on the internet very excited. I will then outline why a LVT is inferior to the Canadian property tax system (which is not that different than the American system for that discussion).

Canadian Municipal Basics

In interests of brevity, I will focus on how the Canadian municipal government system differs from that the United States. (Since I never saw opportunities to look at sub-sovereigns elsewhere, I will not discuss the comparison to other countries.)

The big difference for government finance between Canada and the United States is that Canadian provincial governments have much bigger economic footprints than American states. Canada has a larger welfare state, and it is largely implemented at the provincial level. Furthermore, Canadian provinces do not have balanced budget laws (which would make implementing the welfare state at the provincial level impossible). Modern Monetary Theorists would argue that the welfare should be implemented at the Federal level, but the Canadian Constitution trumps economic common sense.

When we drop to the municipal level, I believe that Canadian municipalities have less independence than is the case for American municipalities. For example, the Quebec government unilaterally decided to merge municipalities as a move towards cleaning up administration in the early 2000s. (The citizens revolted, and a number of those mergers were unwound.) I believe that most American states lack the authority to do anything similarly obnoxious.

Municipalities exist within a legal framework set by the province, and revenue raising has to be using specified means. (The British North American Act divided up the revenue pie between the Federal and Provincial levels of government.)

Details obviously vary across the country, but as in the United States, property taxes are a major source of municipal revenue. (Taxes for schools and utilities like water or garbage collection would be at this level. I think that school tax rates are linked to property tax rates, but might not be included in the property tax numbers in the chart I show below.) To distinguish a property tax from a Land Value Tax, a property tax is levied on the value of both the land, as well as the buildings on the land.

Property taxes are based on an assessed value. In my municipality, there is a periodic reassessment of property values (every few years) based on market pricing, plus you will get your assessment bumped up if you foolishly do major renovations (which I am about to do...). Since all property transactions are registered by the municipality, they have a very good idea what prices properties are transacted at. My feeling is that the assessed property values are pretty good estimates, but with a tendency to be slightly low (across the board). Having the assessed values being somewhat below the true market price is a useful way to keep the number of challenges to assessments limited. The only way the market value of your property is far below the assessed value is that you have some serious hidden defects. However, most homeowners know enough to not run to the entity that has the power to condemn properties as unfit for human habitation to argue that their assessment needs to be lowered because of mould on their  house's framework.

Since the assessment is close to where everyone feels the market price is, I'd argue that the tax is viewed as fair. (As a spoiler, this is no longer the case for a Land Value Tax.)

Less Cyclical than the United States

Canadian municipal finance rules are straightforward. The municipality just follows the given algorithm.

1. Determine the sum of all assessed property values in the city.
2. Figure out how much property tax the city wants to raise.
3. Divide (2) by (1) to get the tax rate.

The city will need to take into account property tax rates in nearby municipalities, but that is about it.

The flexibility of the tax rate means that changes in house prices alone have no effect on property tax revenue. (There are other channels, noted below.) If assessed house prices rise by 10%, they just cut the tax rate by 10%. This means that homeowners just need to look at the tax bill (and scream); the assessed property value is there for entertainment purposes only.

This means that Canadian municipalities are not going to be crushed by falling house prices by themselves (which is different than what was seen in some American municipalities). There would be some forced austerity at the local level, but it would be small potatoes. A housing development boom does help municipal finances, via new developments raising property values, as well as the Orwellian "Welcome Tax" (a tax paid when buying a house). Those positive factors would reverse, but it would not force widespread retrenchment. In a deep recession, the worry would be about the provinces, and whether they are "too big to fail."

(In any event, attempting to short Canadian municipals seems like a suicidal trade. Very few institutional investors trade the things, and so one would face extreme short squeeze risk.)

Is the Property Tax a Good Tax?

Before I get to the Land Value Tax, I want to first discuss the desirability of its close cousin: the property tax. My discussion here is extremely Canada-centric -- although I would argue that the situation is not that different in the United States. For other countries, there are differing institutions and political constraints. In interests of space, I am not going to discuss all issues with the tax, rather a more practical question: is it possible to greatly increase the property tax?

The first problem is constitutional, that appears applicable to both Canada and the United States. Property taxes are levied at the municipal level, and that is the wrong level of government to get a massive revenue boost. Us Keynesians who want "big government" don't refer to municipal governments -- as they do not provide much in the way of welfare state coverage. If we increase municipal governments, rich municipalities will get even more services relative to poorer ones. If your objective is to even out economic opportunities across the country, this makes the problem worse. (Another practical issue is that this would have to be implemented by municipal politicians. If you want to run as a municipal politician on a "let's raise property taxes" platform, expect to get less votes than the Marxist-Leninist candidates.)

The macroeconomic problem is that property taxes have no counter-cyclical properties. They tend to be flat across the cycle, and so they will not help counter-act recessions, nor are they particularly good for inflation control. Municipal governments are too small to apply Functional Finance principles to, but we need to worry about Functional Finance for aggregate behaviour.

Finally, the key problem is that property taxes are levied with no strong link with the capacity to pay. If you want to get theoretical about it, you are levying a tax on a stock, while incomes (and taxes) are flows. Believers in 19th century economic theory might believe that "economic rents" are constant over time, whereas casual knowledge of capital market behaviour tells us that it has to vary with time (assuming it exists..).

The basic law of taxation is that you can't get blood from a stone. No matter what your personal political animus against the stone is, it just ain't happening. Property taxes fight against that principle.

Insignificance of Property Taxes In General Revenues For Big Governments

The chart above shows percentage of "real property taxes" in total revenue for the General Government sector (which includes contributions to the Canada Pension Plan, which are indeed taxes). (As a disclaimer, this data set from Statistics Canada is new, and I am unsure about the details. It may be that some of the local government taxes are levied based on property values, and might be lumped in with property taxes.) As can be seen, it is a small number.

In order to help personalise the issue, I will discuss some back-of-the-envelope numbers for my local municipality. If you have a single earner household that has not bought an insanely large house, the property taxes would drop to about 10% of all taxes paid at an income level somewhere below $100,000. For those of you not familiar with our funny-money currency, $100,000 is a solid upper middle class salary -- more than a junior financial analyst, but probably less than what a competent plumber is making right now. (Since homeowners are often married, the ratio more generally would depend on how income is distributed.) Since all-in marginal income tax rates in the People's Republic of Quebec is around 50%, the importance of property taxes rapidly collapses as incomes rise above $100,000.

Let us imagine what happens if the property tax were doubled? For this hypothetical household, it's a 10% rise in the total tax bill. An annoyance, but survivable. (Since property taxes are not deducted at source -- and the typical Canadian household has almost no financial assets -- such a move would be psychologically painful in the short term for many. However, since many people roll their property taxes into their mortgage payment, the effect might get smoothed out.) 

However, this no longer the case for seniors. The Canadian system has dealt with the problem of senior poverty by letting them putter around in their houses. This is achieved by giving them a minimal Canada Pension Plan (Quebec Pension Plan) and Old Age Supplement (plus need-based transfers). They can survive on low pre-tax incomes since they pay little taxes (mainly the value-added tax). Raising property taxes is a direct assault on their limited fixed incomes.

In other words, raising property taxes will be a mosquito bite for the rich, and you will end up liquidating the class of senior citizens. "Why do you hate old people?" is the correct response to anyone advocating a massive increase in property taxes. Anyone with knowledge of Canadian seniors is that they are ornery, and taking them on politically is suicidal.

I lived in England, and I grasp the historical tribal class divide between land-owners and renters in European politics. From what I have observed in England, it seemed to me that this distinction was breaking down. In Canada (and the United States), that binary division does not fit the political facts on the ground.

I grabbed the following quote from the summary of this Statistics Canada study: Income from owner-occupied housing, 1969 to 2011

From 2006 to 2011, home ownership rates increased from 27% to 35% among 20- to 29-year-old households, and from 67% to 71% among households aged 70 or older.

Home ownership is not the mark of the ultra-rich; it's 71% of the senior population. Furthermore, there is no guarantee that the renters are poor; luxury apartments and old folks homes are a thing.  It may only be the well-to-do households that can afford to move out of their homes into rental dwellings once they have difficulties getting up and down the stairs. (Admittedly, a class division might be more obvious in the larger Canadian cities, but less of a factor in the smaller population centres.)

Furthermore, rural incomes tend to be lower than urban. However, land is cheap, and there are almost no building code enforcement. Anyone who has been working for most of their life should be able to afford a shack somewhere in rural Canada.

Class warriors might think that renters will be spared. It is likely that in the short term, property taxes will not be passed through into rents. However, it is only possible to squeeze landlord profit margins, but there will always be profits. There will be no rental supply forthcoming at a loss. Large increases in property taxes raise the cost of doing business for landlords, and they will be passed through to renters.

This is what happens when you impose taxes that are not linked to the capacity to pay. You have to go where the money is, and only income flows point us to that location. This is why modern welfare states are powered by income and value-added taxes.

(If one is a free marketeer that wishes to shrink government, a reliance on property taxes helps limit its size. This because you hit pain capacity constraints relatively quickly.)

Land Value Taxes: Suboptimal

In addition to the modern revival of Social Credit on the internet, Land Value Tax bugs have arisen from the ashes of dead economic theories. Fans of the proposals are very happy to send you 200 page PDF's that describe its merits. (As regular readers might guess, I did not bother reading any of those 200 page PDF's. I try not to waste my readers' time, nor my time.)

If you live in a jurisdiction without a competently-run property tax system, a Land Value Tax might be a good idea. It may also be easily implemented in ridiculous postage stamp countries like Singapore, or on targeted districts like Manhattan. My comments here refer to the merits versus a property tax, and for countries like Canada that are large enough so that swinging a cat around by its tail does not involve having its nose pass through several border controls.

Taxing land alone would be a straightforward change to the property tax system: we zero out the value associated with the building on the assessment, adjust tax rates accordingly, and send out the new property tax bills. However, no municipality in Canada is crazy enough to do that. Why?

The first problem is that it would be a steeply regressive change to the distribution of tax. In urban areas, plot sizes are relatively uniform within a district, while building values are not. Taxing land alone redistributes taxes from young people living in McMansions to old people who live in bungalows. "Why do you hate old people?"

Within city cores, you will be subsidising high rises (in case any of my readers are liberals, that includes Trump Tower) at the expense of the various low rise support buildings around them. Once again, the rich get richer.

Furthermore. this exacerbates the breakdown between the taxes imposed and the capacity to pay -- which was already a problem for property taxes. Since you are now squeezing the poorer income segments even harder, you will probably have to lower the aggregate tax take.

I had interlocutors on Twitter object to this. Might not development plans change? Sure, if you have a magic want that can costlessly restribute land and buildings. Even if there eventually was a new equilibrium that was in some sense reasonable, there is no assurance that we can get here from there.

There is no guarantee that the high density utopia that Land Value Tax devotees appear to favour would appear. More dense construction practices implies a greater need for local government services. (For example, a high rise with lots of apartments raises the number of tots that show up in the local school.) If the government does not get increased revenue as a result of the development, it is financially suicidal to approve the development plans.

LVT Administration Issues

The next problem with a Land Value Tax is that it breaks the administrative process. We know exactly what people pay for properties -- but the values for land are not observed. Since mortgages are based on property value, there is a large private sector infrastructure for home valuation. This infrastructure keeps property assessments in line with the facts on the ground.

Conversely, in urban areas, land values are purely hypothetical. Almost all transactions involve properties with buildings on them. In my district, my guesstimate is that there was exactly one land deal for sure over the past decade, and possibly a couple more where the houses involved burned down.

Furthermore, as soon as such a tax was imposed. everyone with a couple of brain cells to rub together would throw sand into the land valuation process. At the extreme, every single property transaction would involve a pair of (linked) transactions, with the land given a nominal value of $1, and the building getting the rest of the value. Since mortgages are assessed on both (and lower tax values raises the credit capacity of the borrower), everyone has an incentive to structure deals this way. Since these are the only observed prices, any market-based approach to valuation would have to use them. (Even if a plot of land had no building, people would just slap a pre-fabricated shack on the property to suck up the valuation.)

An idealist might say that the government should pass a law against such abusive tactics. However, the law is what a judge says the law is. Most judges own houses. The government would have its posterior handed to it in court.

The government is stuck with assessing taxes on a purely administrative basis, removed from market pricing discipline. (This will be just as efficient as the tax on windows.) In a big country like Canada, there are three options.

1. It is a flat tax imposed on a per-square-meter basis across the country. However, if one glances at a map, one realises either that the tax rate would be laughably small, or it would wipe out the farming industry.
2. It is administered solely at a local level. The problem is that if the local governments control it, the local governments involved would tell the LVT backers exactly where they could put their LVT, and it would not be implemented.
3. A central government (probably provincial in Canada, courtesy of the BNA Act) imposes guidelines based on broad categories (rural versus urban zoning, etc.)

The final option might attract some LVT fans, but it would be rejected by anyone with any knowledge of politics. The objection is straightforward. Imagine you are in the position of a Democrat living in a blue state. You should be very concerned about the possibility of handing the Republican-controlled Federal government the power to levy an arbitrary tax that can be targeted on the basis of ZIP codes. Unpopular minorities (like English-speaking Quebecers), tend to cluster geographically; giving central governments the power to target people on a geographic basis would raise civil disharmony.

In order to avoid obliterating farmers (a political group that is almost as ornery as seniors), farm taxes per square foot have to be lower than urban land taxes. This creates a massive problem on the border of almost every urban area. If you start of at the centre of a town or city and walk in selected directions, you will eventually hit a border line between a suburban and rural land. The typical case is that on one side of a road is a farm, and on the other side there are houses. Under a property tax system, those houses pay way more tax per square foot. If a LVT were assessed on the value of land, it would be the same as the farm on the other side of the road. In order to avoid obliterating farmers, the administrative rule would have to impose higher rates on the households -- which they could avoid be being zoned as farms. Since practically all suburbanites have tomatoes in their backyards, we would suddenly have pressure to have most of the country declared as rural properties. The central government would suddenly be facing off in court against most of its voters in swing ridings. Good luck with that.

Concluding Remarks

If your model suggests that a Land Value Tax is more efficient than a property tax, your model is wrong.

(c) Brian Romanchuk 2018