Low Density Development Undermines Infrastructure Funding — Even When Property Values are Very High
For many years lower density neighborhoods have been associated with higher quality of life — this was American Dream of the Baby Boomer generation. Consumer preference was so strong that in the early 1990s, David Rusk built his city elasticity theory in part on the idea that older urban centers could not grow population by adding infill density — because middle income Americans would categorically reject living in more dense neighborhoods.
This worldview — that limiting density is still the keystone principal for designing desirable neighborhoods — emerged during the Transform Baltimore Zoning Code rewrite process as a justification for rolling back reforms that promote dense, mixed-use development and prioritize housing, businesses, amenities over things like surface parking for automobiles (necessary in low-density environments).
Advocates argue that low-density places, under the current system of property tax assessment raise too little money to cover the cost of supporting infrastructure, creating a kind of public subsidy for the low-density lifestyle. It is easy to see that this is true where property values are similar — a $150,000 property on a larger lot in a neighborhood of large lots clearly covers less of its infrastructure cost than a $150,000 property on a small lot in a neighborhood of small lots. If common sense is not good enough for you, there is also a growing body of empirical evidence that lower densities create a drain on public finances.
However, there is a tendency to believe that certain large lot/low-density properties, by virtue of being located in elite neighborhoods where homes are high-end, generate such inordinate property tax income that no serious argument questioning these properties’ contribution to infrastructure and other funding could be made.
The below analysis indicates that this is not the case — even when a large lot property is valued among the top 1% of owner-occupied houses in Baltimore City, the inefficiency of low-density development to cover infrastructure costs comparative to more modest properties on smaller lots is not overcome.
The question is: is it fair or right that the property tax assessment system does not account for the cost of covering infrastructure expense such that a very high value property on a large lot is outperformed by a modest small dwelling on a small lot?
Lower Density means less tax per unit of area, even when property values are very high
Indications are that lower density places generate too little income from property taxes compared to their share of infrastructure — even when luxury improvements on a property, or location in an elite neighborhood result in a very high assessment value. Consider two properties located in neighborhoods just on the other side of Cold Spring Lane from each other. First, a Roland Park estate.
It is assessed at $1.12 million dollars and sits on 17,358 square feet, or roughly four tenths of an acre. This property is one of only 866 owner-occupied properties in Baltimore City worth more than $1 million (less than 1%). (2014 American Communities Survey). The home on the property is very large, about 5000 square feet, and $916,000 or 81% of the overall value. At Baltimore’s property rate of 2.248% the property pays $25,170 in taxes. That’s a value of $65 per foot or $1.45 in annual taxes per square foot.
Down the road is a property in the Hoes Heights neighborhood, a modest row home assessed at $139,000. The home is worth approximately $79,000 or 57% of the overall property value. This property is close to the median home value of $155,000 for owner occupied units in Baltimore City (2014 American Communities Survey).
It sits on 1116 square feet, or one-fortieth of an acre. That’s a value of $125 per foot, and at Baltimore’s 2.248% rate, the property pays $3124.72, or $2.79 per foot.
Calculated another way, the Hoes Heights property generates $125,000 in property tax per acre. The Roland Park mansion pays $62,500 in property tax per acre, or roughly only half as productive per acre as a common row home.
Lower Density Means Less Tax Available to Pay for Infrastructure (and Everything Else)
Roland Park Property
While the Roland Park property contributes approximately 8 times more taxes in real terms, the per acre productivity matters a lot, not only because city boundaries are unlikely to expand, but because the low density property is served by a much greater share of infrastructure that must be maintained and occasionally replaced. The Roland Park property’s large lot requires more feet of water and sewer pipe, and more road space per property.
The Roland Park estate in question has about 255 feet of road and sidewalk frontage, shared with two properties. Approximately 125 feet are directly attributable to the property (i.e., property only pays for its “side” of the street).
Some portion of the other neighborhood infrastructure should also be assigned to this property according to a ratio that takes into account miles of street and property area for the neighborhood served.
In the above graph, the total road area is approximately 1.7 miles (adding all streets) and the total property area is about 1.26 million square feet, divided across 50 properties. If everyone shares equally an amount of infrastructure beyond direct street frontage that is equal to the ratio of their property to total neighborhood land area, then the property in question should be responsible for an additional 1.4% of that 1.7 miles of infrastructure, including approximately 20 ft of Roland Avenue (a street that is of a different class than the other neighborhood streets).
All told the property in question is responsible for 255 feet of public infrastructure including 20 feet of a four lane street (more expensive)and 235 of neighborhood street. Because Roland Ave is 4 lanes, we will count the 20 feet twice (total 275 feet). This is approximately 705 ft of infrastructure per acre.
Hoes Heights Property
Now, let’s look at the Hoes Heights property. Because of the simple layout, we can just use property boundaries to estimate infrastructure. The boundaries are approximately 17 x 65. In the front and back, 34 feet is divided two ways, attributing 17 feet of infrastructure. Directly attributable infrastructure is 17 feet.
Next, at the neighborhood level the land area of the Hoes Heights property is about 0.5% of the total neighborhood land area. The neighborhood is served by approximately two-thirds of a mile of public infrastructure, about half of which is alley (cheaper than neighborhood street). All told, using the same formula as above, and counting the alley space as one half of a road, approximately 30 feet of public infrastructure should be attributed to the Hoes Heights property. This yields approximately 1162 feet of infrastructure per acre.
Tax Revenue Per Square Foot
All told, though the Hoes Heights property must pay for more infrastructure per acre, it also produces far more income tax per acre.
Calculating based on the numbers above, the Roland Park property generates $62,500 per acre to pay for 705 feet of infrastructure per acre, or $88 per foot per year. The Hoes Heights property generates $125,000 per acre to pay for 1162 feet of infrastructure per acre, or $108 per foot per year. The Hoes Heights therefore covers 22% more of its infrastructure cost than the Roland Park property does every year.
This comparison should raise questions about how we think about density related to cost of infrastructure in Baltimore. Compared to suburban locations or other large lot properties in the City, the Roland Park property does deceptively well due to the extremely high value of the home on the property.
Similarly, even a modest row home on a small lot outperforms the luxury large lot home; consider how much better luxury rowhomes and townhomes perform in places like Southeast and South Baltimore, where home values may exceed $300–400,000 for similarly sized (or smaller) units.
We want infrastructure costs to be shared equitably. A Roland Park homeowner contributes greatly to the tax base in real terms (not per square foot) and also by paying high income tax, it’s worth noting the deleterious effect of low-density neighborhoods, as opposed to, for example, the same high-income homeowner living in a penthouse suite or luxury townhome in the urban core.
Policy measures to improve infrastructure coverage need not be tax-based; increasing density of business and amenities in existing corridors, or allowing greater utilization of land through, for example, accessory dwelling units would all tend to improve infrastructure coverage.
It’s also important to consider that elsewhere in the City or in the counties large lot properties may host improvements worth far less. The math for properties must necessarily be far worse than the above comparison, examples of the Suburban Ponzi Scheme in action.
Finally, a modest rowhome is not even a paragon of density. Rowhome neighborhoods can add “missing middle” and mid-rise buildings to add tons of density without dramatically altering the character of neighborhoods. This comparison should illustrate just how financially powerful density is in terms of paying for infrastructure.