The Fiscal Realities of High Density Development

January 20, 2005

SUMMARY:  A look at the amount of residential real estate tax revenue raised by high density development suggests that such development will be revenue-negative. Therefore, advocates of high density growth are either working only in their own interest (developers come to mind), or are asking the current residents of the jurisdiction to subsidize this growth while these residents suffer from its impact on roads, schools, libraries, and other services.

The purpose of this article is to take a look at the fiscal costs of high density development to a locality. Local government must provide services to the residents within its jurisdiction. The most expensive service is providing free public K-12 education. Besides educating children, local governments must also provide their residents with police, fire, and ambulance services, social services, parks, recreation, libraries, public transit and public works, and of course pay the administrative costs of governing.

To pay for these services, local governments raise money through taxes on businesses, homeowners, workers, and shoppers. They also get funds from their state government and the federal government. For capital projects, they can issue bonds, but they must pay the interest and principal on the bonds from operating funds.

The local government must get the mix of its various taxes right. If taxes on businesses are too high, the businesses will relocate to low-tax jurisdictions. It's a rare jurisdiction that doesn't have to worry about this, although there are a few. For example, New York City has a lock on many types of financial service business, though the lock is less tight than it once was. On the other hand, if taxes on businesses are too low, homeowners must shoulder the burden in the form of higher real estate taxes. Since these homeowners are likely voters, there is a limit to how much a government can tax them as well. Workers are even harder to tax. A local income tax makes that location a less desirable place to work, and businesses will have to pay higher wages to compensate the workers. As a consequence, local income taxes are rare — and unpopular. Finally, localities may levy sales tax in addition to the state sales tax. Again, when this tax gets out of line with sales taxes in other localities, people will shop in the low tax locations and hurt businesses in the high tax localities.

Development creates additional revenue for the local government in the form of real estate taxes on the new homes. Therefore it's a fair question to ask whether new development will produce enough real estate tax revenue to compensate for the cost to the jurisdiction of the services that must be provided to the new residents. If the new development is revenue-negative, the other taxpayers of the jurisdiction will have to subsidize the new residents by making up the difference — either by paying more taxes or getting less service.

Calculating the revenue impact of new development is difficult and involves many assumptions. The fairest way to make the calculation is to determine what percentage of the locality's revenue comes from residential real estate taxes and then determine the amount that must come from each household. This assumes that the next year's budget will add services in a proportionate amount for the new residents. That may not be the case. The jurisdiction may attempt to hold the level of service the same or, more likely, will increase the level of service more than would be expected accounting for the population growth and inflation. However, we'll assume that the increase in the budget will be proportionate to the growth in population. If that's true, the new development should produce enough real estate tax to make up the difference.

The following table shows the income from real estate taxes and the expenditures on service for several development scenarios at different densities in my location, Fairfax County, in Northern Virginia. I've used two acres as the parcel size because it means we don't have to deal with fractions of houses for R-3 parcels. I've set up some house assessed values based on current market values in the various zoning areas. The houses in the R-3 neighborhood are at the low end of the scale, meaning that this is probably an older neighborhood. Home buyers know that new R-3 neighborhoods have houses that are in the $600,000 - $900,000 range. If I used those higher figures, they would be more supportive of my argument, so I've chosen to use older neighborhoods as a worse case. Yes, houses are expensive in Northern Virginia.

To determine the amount of real estate tax this development will produce, I've applied the County's 2005 tax rate of $1.00 per $100 of assessed value. In Virginia, houses must be assessed at market value, which results in an assessment rate of about 91-95% of actual market value. I've simplified here by assuming that the assessment rate was 100%, deliberately making my revenue numbers a bit high.

To calculate the cost of County services for a household, I've first taken the number of households in the County and adjusted it up to account for the anticipated population increase from January to July. I've then taken the total County budget and subtracted out the County's transfer to the schools, leaving a budget of $1.46 billion. Dividing that budget by the number of households produces a cost per household of $3,832. The next column in the table multiplies the number of homes in the development by that cost. Because residential real estate taxes are only called on to pay (just under) 50% of the cost per household, I've reduced the cost accordingly. Yes, it's no secret that businesses subsidize the services provided to households through the taxes those businesses pay.

To calculate the costs of education, I've taken the number of children in the school system and divided it by the number of households. That suggests that there will be about .433 children per household. Of course, you can't have a fractional child. I've heard the figure used that only 20% of households in the County have children, but of course many have more than one child. Our fraction of a child, when applied to a large development, will probably be fairly accurate in predicting the overall number of children who will live there. We'll discuss some caveats to this below.

It costs Fairfax County $9,462 per year to educate a child (not counting the contributions of the state and federal governments). Again, we'll reduce the cost to what the residential real estate tax will have to support. This gives us a cost per child of $4,597.

Finally, adding the costs of education to the cost of services, we get the total cost. That can be compared to the revenue raised. Here's the table. If you don't like my assumptions, I've attached the spreadsheet I created so that you can make your own assumptions and do what-ifs. The spreadsheet also includes the references for the numbers, most of which are taken from the County budget for FY 2006. The link is at the bottom of this page.

Residential Real Estate Tax Revenue and Budget Expenditures on 2 Acres of Residential Property
Zoning Number of Houses Assessed Value Per House Real Estate Tax Revenue County Services Paid by Real Estate Tax Education Cost Paid by Real Estate Tax Total Paid by Real Estate Tax
RE 1 $2,500,000 $25,000 $1,862 $1,992 $3,853
R-1 2 $1,500,000 $30,000 $3,723 $3,984 $7,707
R-3 6 $475,000 $28,500 $11,169 $11,951 $23,120
PDH-12 24 $375,000 $90,000 $44,677 $47,803 $92,480
PDH-18 36 $325,000 $117,000 $67,016 $71,704 $138,720

This is probably a good time to make some comments about the number of children in the various types of development. I know of no studies that address this point, particularly because it's probably location-dependent and related to house prices and incomes. Still, we can probably make some general assumptions. First, the million dollar houses will have fewer than average children per house; the owners are most likely older and most of their children have gone to college or have families of their own. The number of children in the R-3 and PDH-12 developments is probably higher than average. Those house prices reflect the fact that many of their owners are in their child-rearing years, and while they want more room, they're saving for college and can't afford mansions. However, given the recent home price escalation, it may be that many of the children in these houses are in their teenage years and will therefore spend fewer years in school. These are no longer starter homes, but increasingly move-up homes. Formerly, I would have said that the residents in the PDH-18 homes would be mostly singles and childless couples. However, the escalating home prices make those homes places to begin a family. They may also attract more single parents. So while the number of children in PDH-18 might be slightly less than the average, it's not likely to be as out of kilter as it may appear. It would be nice to have some actual data on this.

It's worth saying something at this point about high-density high rise condo projects. It has always been thought that these will be net revenue positive developments because, with owners who are singles or young married couples, there will be few children to educate. The problem, however, is that with the high cost of housing the comparatively low-cost condos may attract single parents. Also, couples with children may be financially unable to move up to the more expensive single family homes with yards as their families grow, raising the number of children per condo to a level much closer to average. Finally, in housing turndowns, condos have generally taken much greater price hits than single-family homes. That would mean that real estate tax revenue from such developments would decline (at least proportionately) much more than in single family neighborhoods. A "perfect storm" of these three possibilities would have a huge impact on the net real estate tax gain/loss they produce.

The table and the foregoing discussion clearly show that high density communities tend to be revenue-negative for local jurisdictions, and the more dense the development, the more revenue negative they become. There are exceptions, of course. Expensive high density housing catering to rich seniors would be strong revenue generators and, due to the dearth of children, would be low cost. On the other hand, inexpensive high density senior housing might be a totally different story. Most jurisdictions give a break on property taxes to low and moderate income seniors with low and moderate assets excluding the house itself. Thus if we were to fill high density housing with seniors, we might collect less than half of the projected real estate tax revenue. So, as a general rule, high density housing in the suburbs is often going to be very expensive for the jurisdiction when the real estate tax revenue is compared with the cost of the required services.

In some respects, the numbers I've used understate the problem. First, we've worked with the assumption that each household uses its proportionate share of County services. In fact, lower income households are likely to use more services than higher income ones, especially the higher income ones that don't have children. Second, we haven't figured capital costs for new schools, new roads, new libraries, and new government facilities, just the costs of operating and maintaining them (and not the cost of maintaining roads, which in Virginia is a responsibility of the state government for the most part). In densely built areas, there may be little land for these facilities, and government will have to take existing residential and business property at huge cost. Anyone who has driven in Northern Virginia is aware of the serious traffic problems, to which there are no inexpensive solutions, either by increasing mass transit, building new roads, or both. High density growth has a much greater negative impact on all these capital needs than does low intensity growth.

Of course, there may be good justifications for high intensity development in some small number of locations. However, advocates of high density growth are either working only in their own interest (developers), or are asking the rest of us to subsidize this growth while we suffer from its impact on our roads, schools, libraries, and other services.

Download the spreadsheet

This article can be found at

Go to the next article, "A Critique of Smart Growth"

Go to the index for the entire Smart Growth Series.

I appreciate well thought out comments on these articles, and will sometimes incorporate them into a future draft of the article.  I'll ignore abusive responses and ensure future correspondence from the sender goes to the mail filter that deletes it unread.

Send me an e-mail

Envelope: click here to e-mail me