Should 888 Richmond Way pay higher property taxes than 13 Deadman Road?

In Canada, property taxes are typically based on "market value assessment". The assessment is arrived at by

• gathering information, for as many properties as possible, on:
  • observable characteristics such as lot size, age and size of the home, number of bathrooms, proximity to parks and the city centre and
  •  selling prices 
• estimating a relationship between the homes' characteristics and their selling prices, and then 
• predicting what each individual property would sell for. 

For example, after carrying out a statistical analysis of data on home sales, an assessor might arrive at an equation for determining the value of detached homes in Snowbank, Saskatchewan that looks something like this: 

estimated market value=0.2*lot size (in sq ft)+0.5*house size (in sq ft)+15,000*number of bathrooms+25,000*number of bedrooms-5,000*distance to city centre in km+500*presence of granite countertops

Each home owner's property taxes are then calculated by multiplying the estimated value of their home (in thousands of dollars) by the mill rate  - taxes owing per thousand dollars of assessed value.

There are three basic justifications for basing property taxes on market value assessment. First, people buy homes because they derive a flow of consumption services from those homes. A person's willingness to pay for a home is equal to the discounted present value of the consumption services that home yields. Therefore the market value of a home is the best possible estimate of the flow of consumption services it generates.

A second justification for market value assessment is that it is a way of taxing people's housing wealth, and the flow of implicit income from that wealth. This wealth and income would otherwise be untaxed in the Canadian tax system (except for capital gains on second homes and income from investment properties).

A final justification for market value assessment is that it is better than the alternatives, for examples, taxing homes on the basis of their purchase price, or on the number of windows that they have.

While the basic principle of market value assessment is uncontroversial, the implementation is not. For example, homes that are close to public transit typically have higher market values than less conveniently situated homes. Should that higher value be reflected in their property taxes? If the aim of property taxes is to tax the value of the home to the homeowner then, yes, people who live close to public transit should pay higher taxes. Yet if people pay higher taxes when they live near public transit, they will have less incentive to do so. Developers will have less incentive to build near transit infrastructure, making it harder to develop a well-functioning city.

Another example is the taxation of more subjective elements of property value. A recent paper by Nicole Fortin, Andrew Hill and Jeff Huang, using five years of data on the Greater Vancouver real estate market, found that:

in neighborhoods with a percentage of Chinese residents higher than the metropolitan average of 18%, houses with street numbers ending in four [considered unlucky in Chinese numerology] are sold at a 2.2% discount and houses with street numbers ending in eight [considered lucky] are sold at a 2.5% premium in comparison to houses with street numbers ending with any other digits.

In parts of the Lower Mainland, the market value of an 888 house number will be significantly higher than the value of a 444 house number. So should the owner of number 888 pay higher property taxes, and the owner of number 444 pay lower ones?

The purpose of property taxes is to tax the imputed income people receive from their homes, or the wealth that people hold in housing. If people are willing to pay a premium for house number 888, it must be because they receive consumption value from having a lucky house number. The consumption value of living, say, near a park is typically taxed in most market assessment systems. Is that any more arbitrary than taxing lucky numbers?

On the other hand, the residents of 2880 Kings Avenue might have bought their home 40 years ago, and have no use for numerology and other superstitions. Why should they have to pay higher taxes just because they happen to have a lucky number? What if they can't afford higher taxes?

While the "it's not of value to me" argument has some force, it applies equally well to unused bathtubs, bedrooms, garages, and so on. If the lucky number/extra bedroom/fancy bathroom is not of value to the owner, or he cannot afford the property taxes, he should sell the home. This would lead to a more efficient allocation of the housing stock. 

I'm not seriously advocating giving special tax breaks to people living in houses numbered 13, or including feng shui in market value assessment functions. Yet the arguments against taxing lucky numbers illuminate serious tensions faced by our present system of market value assessment.

Note: I've edited this post in response to comments from a professional tax assessor posted below. 

Comments

[hosertohoosier]

I think it's pretty clear that property values are a poor gauge for the direct use of services. Police services are one of the biggest budget items in any municipal budget and, if anything, rich people probably consume those services less. Ditto fire services.

However, there is a premium to living in a city with good public services. The bigger one's financial stake in the city, the more one benefits indirectly. If Toronto increased its police force and crime withers away to nothing, property values would soar (pending the tax implications of more police). However, because the residents of 888 Richmond Way have a more valuable home, they will gain more than the residents of 13 Deadman Road - assuming that both homes were in areas where crime rates drop by similar amounts (and anyway, even if one lives in a safe area, people do plenty of things in common space areas, like downtown).

[Determinant]

Part of your argument, Frances, is really aimed at the possibility under any property-value tax of being house-rich and income poor. Any stock or wealth-based tax runs this risk.

Second, under most market-value systems, including Ontario's, a municipality's tax rate must decrease to balance off the total value increase in a municipality after a reassessment. So you only get hit with a tax increase if the value of your home increased faster than the municipal average.

Any purely objective, quantitative system based on the number of windows, the number of bedrooms or the number of toilets and nothing else creates a market distortion and makes any other valuable addition essentially tax-free. So we have to take note of total value, which is a purchaser's best guess about the total value and features of the home, from location, style, "taste", while prioritizing nothing in particular. Do we really want a tax system that advantages homes with one window and one bathroom, though it have 10 bedrooms and an inground pool.

No, we just let the market give it a number based on similar houses and go from there.

[Determinant]

There is also the political question that the alternative, income taxes and VAT/Sales Taxes, have been strictly reserved to the provincial government in most if not all provinces.

Some US states do allow municipal income taxes though, but they haven't replaced property taxes, they just augment them.

[Frances Woolley]

hosertohoosier: "However, because the residents of 888 Richmond Way have a more valuable home, they will gain more than the residents of 13 Deadman Road"

No, the point here is that there is a price differential between homes are otherwise identical, except for the luckiness/unluckiness of their numbers. In this case, I don't see that increased policing would have more/less impact on one house than the other - unless the criminals carefully checked out the feng shui of the home before robbint it.

Determinant - "house-rich and income-poor"

As the baby boomers retire, this is going to become a huge issue.

"we just let the market give it a number based on similar houses and go from there" -

"Similar houses" in practice means ones with similar numbers of bedrooms/toilets etc. A market value assessment only differs from what you call a "quantitative system" system in that there is more flexibility in the weights attached to, say, bedrooms and bathrooms, and also which parameters enter the evaluation function. I suspect that reason people don't do more to game the evaluation system in Ontario is that the formula used to estimate a home's market value isn't widely known.

[Determinant]

As long as you have one parameter called "market" which is the cost of comparable houses, by location or similar features, the market, in all its generality, enters the equation. It's a random parameter with some sort of normal distribution.

As far as I know, MPAC in Ontario takes into account the selling prices of neighbouring homes when evaluating a particular house, or the selling price of the house in question if it is recent.

I witnessed an exchange between a township clerk and a cottage owner (I was waiting to get garbage tags) where the owner complained at length about his higher taxes due to reassessment. Yet he admitted that his purchase price was still higher than the assessed value. He lost the argument.

"House-rich and income poor?"

When a person loses their job, they are expected to sell their home if necessary. After voting for governments that sat on their hands for three decades while the employer-based pension system was rolled back, particularly in the private sector, why are we getting concerned about the baby-boomers now? Pure government retirement benefits are not overly generous in Canada, it was never a system designed to keep people in their valuable homes.

The rest of us who have had to endure stagnating incomes, income insecurity, volatility and prolonged unemployment are not amenable to protecting the accumulated assets of baby-boomers.

There nothing wrong with selling a house and purchasing a life annuity. Your micro colleagues have posted on this blog that that is exactly what could, should and will happen. Why are we shocked?

[Frances Woolley]

Determinant: "As far as I know, MPAC in Ontario takes into account the selling prices of neighbouring homes when evaluating a particular house, or the selling price of the house in question if it is recent."

This is done through what is termed in econometrics "fixed effects" (or dummy variables or indicator variables). So in the equation above there would be a variable equal to one for for people who lived in Shaugnessy, with a large positive coefficient, and a variable equal to one for people who lived near Main and Hastings, with a large negative coefficient. I'm not sure whether the fixed effects are done at a street/block/neighbourhood level, however.

"There nothing wrong with selling a house and purchasing a life annuity."

Absolutely. That's why I wrote above "This would lead to a more efficient allocation of the housing stock."

[Phil]

What you're saying is that the formula for market value is flawed, because it doesn't include "quiet street" or "888" consumption value.

Fair enough. Does that mean, though, that the entire system must be bad? If the system is *roughly* fair, isn't that good enough? I mean, the income tax system is only roughly fair, isn't it?

Also, in most cases, the buyer of the house on the quiet street isn't really getting a tax break, because the market value he paid for the house was higher than it would have been if taxed more accurately. That is, the previous owner extracted the value in the selling price.

That is, a house that provides X consumption value with an "888" would sell for more than a house that provides X consumption value because of an extra bathroom, because it's taxed less -- at least if you're concerned about fairness to current homeowners.

[Frances Woolley]

Phil - "Does that mean, though, that the entire system must be bad?"

No. But it does mean that - perhaps - it might be better if it took into account the impact of house numbers on house prices.

If taking into account the impact of house numbers on house prices *doesn't* improve the assessment system, the question then becomes: why?

"the buyer of the house on the quiet street isn't really getting a tax break"

Your point, that property taxes are, to some extent, capitalized in housing prices, is a reasonable one. However my gut feeling is that there is less than 100% capitalization, for the simple reason that there is a fair amount of uncertainty about the future direction of property taxes - this is the case where I live. As soon as a home sells everyone checks the selling price and tries to guess the impact on future property taxes.

[Frances Woolley]

Sent, via email, by a professional property assessor:

I was just reading your views on property assessment at Worthwhile Cdn
Initiatives. It's sort of correct but also incorrect.

You've forgotten to mention that assessments are also based on the
concept of the mono-centric city. Closer property is to where the
action is (like jobs and stores), the more valuable the land is. Most
people do not realize just how much the value of land is. In most
cities that are large enough to have old neighborhoods built let's say
back in the '30s or '40s, and newer neighborhoods built up around those
every decade. A lot of the times the land is 70% of the value, with the
improvements (homes and garages) being the rest. New neighborhoods it
can be the reverse. The home without depreciation to take away value
being the bulk of value.

Residential properties are suppose to be calculated through multiple
regression - mass-appraisal, using the sales approach. Sales of similar
homes being the core of valuation with details (like lot size, house
size, # of bathrooms, developed basement, granite counters, etc etc.)
used to personalize the value to that home. Only characteristics with
proven market value should be used. Calgary for example believes
garages add no value. It's not a variable used. Elsewhere we
believe a garage is an improvement to the land so is valued.

Commercial & rental properties are income properties so use the income
approach. The last approach is cost. The cost the build the improvement
plus the land value. Only used in special property cases, like
hospitals.

I don't know of quiet neighborhoods ever being a variable. There is a
rule in assessment - no discrimination! A noisy neighbor in theory can
move. Land penalties are given for traffic counts. Positive land
adjustments are made for being in more secluded areas, usually because
those are in a close or by a man-made lake (or slew) or surrounded by
green space. Being close to a major road or feeder road is usually
considered a good thing. Remember, mono-centric city theory.

The only problem with assessment economics is when politicians pass
laws which effect the market value, like California! Other bad things
include putting a roof on value. That won't change the mill rate. Only
budgets can change mill rates. People are taxed on the amount of equity
they have in the land and its "highest and best use" (another important
principal).

from a property assessor

[Determinant]

So you are asking, Frances, why Property Taxes aren't perfect. They have to be computed *relatively* fairly for 10 million or so files, fair enough for the complaints against the system to be manageable.

Speed matters in a mass-production system like property tax assessments.

Sometimes a rose really is just a rose, not an unfair rent.

[Stephen]

I'm the property assessor (aka, tax assessor) who wrote to Frances...

Couple things I should point out about property assessments since most people don't know how it's done. The methods used are suppose to be IAAO standards but not all provinces (or States) end up following it due to politicians getting their thumbs into something they don't understand (economics!). I'm going to use Alberta as my example since that's where I am. Other places should have something similar.

The value you see in your property assessment is a historic value. Here it's July 1st of the previous year. The sales used to determine the values are the sales in that market (like a specific city). Similar home 'models', 'qualities', and 'structures' are used for those specific MQS type homes. So the sales of a semi-custom 2-story homes built after 1984 while be used for that MQS. If there are not enough samples (sales) than appraising adjustments need to be made. The sales used are between a certain time (July 1st of 2 years ago to June 30th of a year, here in AB). Fair market value sales are used. NO foreclosures, homes sold from parents to kids, homes sold under duress, in-house sales (like what realtors do), or new home sales of newly built homes are used. A fair and open market sales only. Yes there are outliers and if a sales investigation can't solve that than it's an exception.

We need to keep our records on every home up to date. We door knock, we watch MLS, we send out questionnaires in hopes to be as accurate as possible. Like all statistics it is an educated guess. IAOO standards say the assessed value needs to be within +/- 10% of the market value. The value you see is a median value. Here in Alberta we use a 5% ASR, not a 10% one. We do get it wrong sometimes wich is why we have appeal times. If you feel your home is assessed wrong (over or under 10%) than for sure come and see us! It's our job.

One variable that is suppose to be illegal in valuing is on discrimination. A house in a "ghetto" neighborhood can not be penalized for that. Crime rates is not a variable. It's the size of land and its location to desirable things, like green spaces (parks), a close, near a major road. Also undesirable ones like heavy traffic, an oddly shaped plot of land, distance away from major commercial areas. This is why suburb land has less value than inner city land. Also the improvements to the land has value, like houses. "Ghetto" neighborhoods usually have poor MQS houses, they are small, they are old. It's this age factor which robs the improvement of its value usually due to depreciation. If you reno your house than it's like a face lift. Your home's effective age becomes less old than the actual age.

Homes in old inner city neighborhoods usually have around 70% of their value in the land. New homes are anything from 50/50% land/improvement or more value in the house than the land.

Mill rates come from budgets. Here it's 2/3 from the city budget and 1/3 from the province's education budget. So we here in Alberta got a big property tax increase because our government increased the education budget by $7 billion (she lied, she did raise our taxes). If your city is growing than city sprawl explains some of the budget increases. The rest are vanity city planner projects (like tearing up roads and sidewalks for bike lanes). Us public employees only get a pay raise equal to inflation (hopefully that's ok).

[Frances Woolley]

Stephen - I'm glad you've weighed in on this debate. Property tax assessment falls into the category of "things that lack a useful Wikipedia entry" which suggests that there's a need for more public debate/info.

On the "do nice neighbourhoods pay higher taxes" question. Nice neighbourhoods do tend to be close to green spaces, parks, good shopping areas (in part because residents of higher income neighbourhoods may be more successful in lobbying for parks - compare, e.g., the amount of green space in Vancouver's East and West side). A typical outside observer would just see higher taxes in nice neighbourhoods, and assume it's a neighbourhood effect and not a proximity to parks/shopping effect.

[BSF]

Frances: Are you following the Irish political debate on this?

http://www.irishtimes.com/newspaper/finance/2012/0904/1224323571988.html

http://www.irishtimes.com/newspaper/ireland/2012/0910/1224323797426.html

http://www.irishtimes.com/newspaper/ireland/2012/0904/1224323575035.html

[Frances Woolley]

BSF, no, I hadn't -I'm wondering what that's going to do to Irish property prices?

[Peter N]

This is from Dr. Housing Bubble. I don't know enough to have an opinion, but it sounds ugly. Do you or any of the others have an opinion about this?

http://www.doctorhousingbubble.com/

"Canadian housing bubble goes into full mania mode – Canadian debt-to-personal income ratio near 145% while US at peak of the housing bubble was at 125%.

As global real estate bubbles burst at differing intervals, those still engaged in the depths of mania fever find every convenient argument to justify the existence of the current inflated economic structure. We can debate the nature of the current US housing market but with the median nationwide home price at $151,600 from the most recent Zillow housing report and the median household income at roughly $50,000 prices look to be leveling out especially with the absurdly low interest rates. As we know, housing markets are regional so applying this nationwide trough to frothy markets may not be the best way to measure investment value. However, when we look at the Canadian housing market we realize how insane things have gotten. I’m amazed by how many of the debt rehab or home flipping shows have migrated to the Canadian market. Of course they rarely mention this thinking the American audience will mindlessly assume they are in some other US city to prime the consumption pump. Yet when we look at the metrics, Canada is poised for a deep and profound correction."

[Frances Woolley]

Peter N - interesting question, but somewhat off topic.

[Stephen Gordon]

Awesome: a post on tax assessment elicits thoughtful, constructive and instructive comments from a professional. Thanks, Stephen, and well done Frances for drawing him out.

[Peter N]

I thought this one would be in your area, since it involves determination of prices. Is a house in Vancouver worth 60% more on a sales to rental ratio than Honolulu at the peak in 2005? The Canadian sales to rental ratio is the highest in the developed world at around 175%.

This can't end well. Is anybody paying attention?

http://theeconomicanalyst.com/content/house-price-rent-ratios-canadian-cities-alarming-levels

[Frances Woolley]

Peter N - Let's put it this way: I'm not buying property right now.

[Pmu resultat rapport quinte]

I think the admin of this website is actually working hard
in support of his web page, since here every information is quality based material.

[Lindsay]

Something that has always bothered me is that if the property tax is a tax on the implicit income from wealth, why do we have circuit breakers and deferrals?

[Jacques René Giguère]

Here on the QC North Shore, 22-years-old mechanics are stil buying $ 1M house while iron ore prices have been halved in the last two months. I lived through the 80-81 crisis and I am already scoping a home for my retirement. There will bargains soon.

[Stephen]

Pmu: if the property tax is a tax on the implicit income from wealth

It's not exactly that (but more or less). It's your share of the equity within the city. Have a expensive lot or big lot - you pay more for having more land. Bigger house or fancier house which the market recognizes people will pay more for, you have more equity as compared to the owner of a tiny house on an oddly shaped lot next to a freeway (since most ppl don't want that and only do buy it because it's what they can afford).

The tax is your share of the city budget based on the amount of equity you have valued by the assessment. You have to remember that market value doesn't care if you can afford a house in the long-term. It just recognizes the market value of the home on that specific date of the assessment (in Alberta it's July 1st of the previous year).

More people need to judge what they can afford over what a bank says they're willing to lend them. I bought a condo this year. I could of bought one at twice the value I did buy mine at. I just like having spending money throughout the month over a twice as fancy newer condo (though I do have a good one for its year built - lots of renos from previous owners).

- Stephen