Economics In A Small College Town
Photo: Wannapik.com. Creative Commons
By Gerald Friedman and Maral Asik
Introduction
The Town of Amherst is confronting development issues that will shape its future. These issues have become bitterly contentious in our town, with neighbors choosing sides along issues of perceptions of equity, respect, and aesthetics. But behind these debates are often unspoken economic assumptions, models of the relationship between town policy and economic outcomes, and between economic outcomes and our town’s ability to provide services.
I hope over the next few weeks to contribute to our dialogue by clarifying some of these economic issues. With a research assistant, Maral Asik, we will present a three-part series discussing the relationship between different forms of development and tax revenue and town expenditures, assessing the impact of alternative commercial development strategies on economic growth in the town, and concluding with some proposals for economic development that would build on our town’s unique resources.
In drafting these pieces, Maral and I would like to shed light and contribute to our town’s civil dialogue. Preferences for different policies involve judgments where the principle of de gustibus non est disputandum (no arguing with taste) may hold. Some may prefer modern glass building while others prefer traditional New England brick and wood. But behind each policy is an economic model; every judgment has economic ramifications, implications for our town’s prosperity and its tax base and its ability to provide services to our residents. We are all free to make choices. Maral and I would like to identify those choices, and suggest ways that we might make them less painful.
Part One: The Sad Facts Of Federalism And Local Government
All activity has costs and even commercially successful developments can reduce the town’s ability to fund services. This is because within the American federal system, Amherst is responsible for funding many services that would be a national or regional responsibility in other countries.
The greatest expense by far is education, where the town is responsible for about 60 percent of the cost of local schools. Many of us moved here to educate our children in excellent public schools. These schools are expensive, and a powerful argument for promoting commercial development is to fund these schools. Unfortunately, the logic of local funding is that measures that increase the tax base while also increasing the number of students may be counterproductive, raising the demand on the town’s educational services faster than they raise revenues.
Because residential development often brings children in need of education to the town, and schools account for about half of the town’s total budget, many promote commercial development as the best way to increase tax revenues. But even developments that provide additional property tax revenue require services. And the additional property tax revenues may come at a cost. New commercial development can undermine the tax base by driving away residents, existing businesses, lowering the value of existing property by enough to reduce property tax revenues.
These are not purely hypothetical suggestions. The most obvious case, of course, is where a new commercial structure replaces an existing one and one must subtract, from the gross tax revenue from the new building, the amount that was already being paid by the previous structure. Beyond this, there is the danger that competition from new businesses will drive existing ones out, into bankruptcy or out of town. Aesthetics matter. Poorly designed developments with ugly new buildings and poor pedestrian access and views can drive residents to leave town, taking with them their tax revenues. At worst, some without children will vacate their homes for younger families with children requiring expensive education.
Even where new buildings occupy new spaces, they can lower the value of existing property. Unless the new building brings in new tenants, competition can drive down rental prices on retail or office space, driving down the value of existing structures and their tax liability. Even worse is the case of particularly ugly structures. Attractive and interesting construction, such as the Pompidou Center in Paris or New York’s High Line, can contribute to the revitalization of neighborhoods, encouraging visitors and new businesses. Eyesores can make an entire community unattractive. Highways and elevated transit, of course, are prominent examples but others include ugly and barely designed buildings like the empty windswept plaza in front of New York’s Lincoln Center, or the ugly towers lining Sixth Avenue in the city’s Midtown district. Cases like these are a warning to town planners: tearing down a variety of business sites and replacing them with a single giant structure can destroy a successful and diverse community, leaving behind a wasteland.
Advocates of commercial development are on firmer ground in criticizing the construction of residential property that may lead to higher spending on town services. On average, taxes from residences do not cover the cost of the services their occupants consume (see Table 1). Of course, there are caveats here as well. Since the greatest expense is education, families with children are especially expensive while those without children usually pay more in taxes than they cost; and more expensive houses pay more in taxes making them less of a net burden on the town’s resources. On the other hand, large families cost more than those in less expensive houses paying less in property taxes.
The joint effect of family size and property values on the town’s net revenue from residential property is also illustrated in Table 1. The intuition here is simple. The town spends a certain amount on police and fire, an amount that rises as the town gets bigger and additional residences are added. Taking account of this, new residences generate enough property tax revenue to balance the cost of educating additional students only if there are very few students or very high value homes. Indeed, even the most expensive homes of almost $1 million cost the town more if there are two or more children to be educated. On the other hand, families who remain in their homes after their children leave school make net payments in taxes to support town services, especially if they live in expensive houses. Encouraging such families to remain in their homes helps the town budget. Should they vacate and sell to younger families, it will draw students to the town’s public schools.
| House value ($000s) | Tax revenue minus marginal town expenditures | ||||
| No children | 1 child | 2 children | 3 children | 4 children | |
| $ 150 | $ 1,793 | $ (7,949) | $ (17,691) | $ (27,433) | $ (37,176) |
| $ 250 | $ 3,925 | $ (5,817) | $ (15,559) | $ (25,301) | $ (35,044) |
| $ 350 | $ 6,057 | $ (3,685) | $ (13,427) | $ (23,169) | $ (32,912) |
| $ 450 | $ 8,189 | $ (1,553) | $ (11,295) | $ (21,037) | $ (30,780) |
| $ 550 | $ 10,321 | $ 579 | $ (9,163) | $ (18,905) | $ (28,648) |
| $ 650 | $ 12,453 | $ 2,711 | $ (7,031) | $ (16,773) | $ (26,516) |
| $ 750 | $ 14,585 | $ 4,843 | $ (4,899) | $ (14,641) | $ (24,384) |
| $ 850 | $ 16,717 | $ 6,975 | $ (2,767) | $ (12,509) | $ (22,252) |
| $ 950 | $ 18,849 | $ 9,107 | $ (635) | $ (10,377) | $ (20,120) |
| $ 1,050 | $ 20,981 | $ 11,239 | $ 1,497 | $ (8,245) | $ (17,988) |
By this metric, expensive student rental housing can be a good investment for the town because they pay high property taxes while requiring few public services — some police and fire, but they have few children requiring expensive education services. Although high-rent student housing accommodating high-income students brings new revenue for the town, this may not apply to low-rent and less expensive housing, such as some of the developments outside of downtown or homes that are converted to student residences. When filled with students, such housing, by definition, pays less in property taxes than more expensive housing, but this is not the major problem. More expense arises when relatively low-rent student housing attracts families to their affordable housing. Often attracted by the town’s good schools and safe environment, these families require an increase in the town’s education spending.
Where Does This Leave Us?
Commercial development provides tax revenue to help pay for town services only insofar as the development attracts more new business than it undermines.
Residential housing, by contrast, drains existing resources except where filled by college students or other families without children in town schools. When occupied by young families, housing demands more resources for educating children while almost always providing less in new taxes. This is especially the case with low-cost or affordable housing where less is paid in taxes. Residential housing can provide net resources to the town, however, where it is filled with high-rent students or family housing without children in the schools.
None of this determines policy. We could decide to promote more affordable housing and compensate for a limited tax base by raising taxes. The town could choose to promote high-rent student rental housing, housing with rents beyond the capacity of low-income families to afford, and commercial development to expand the tax base and lower tax rates. But these approaches risk backwash effects unless carefully planned. Student housing only provides net resources where it continues to rent to students, not to families with children. Commercial development only generates tax revenues where new development does not cannibalize existing businesses or undermine the town’s appeal to retail customers, new business development, or established residents without children in the schools.
There are no easy answers. What we need are commercial and residential developments that bring successful businesses to the town, without alienating families living here, and that pay taxes even when they no longer have children in the schools. This requires creative business development. Which we will discuss in the next two pieces.
Gerald Friedman is Professor of Economics at the University of Massachusetts. He graduated from Columbia College in 1977 and earned a Ph.D. in economics from Harvard University in 1986. His most recent book, The Case for Medicare for All, appeared in March 2020 with Polity Press. Follow him on twitter at @gfriedma
Maral Asik is originally from Longmeadow and is a freshman at UMass Amherst participating in the Economics Undergraduate Research Assistant Program. She is interested in economics, international relations, and environmental studies.
Why is it reasonable to use the average cost per public school student rather than the marginal cost of adding an additional student in this analysis? This seems as objectional as the State’s forcing the town to pay charter schools the average cost per student, rather than the marginal cost, when a student leaves pubic schools for a charter school. I’m not an economist, but this seems like an Econ 101 mistake to me.
I am in agreement with Steve George’s letter about employing the marginal cost for each student, rather than the average cost. This is particularly true in Amherst which has a large number of students who “choice in” to the Amherst schools each year (about 200). By State law, the sending school districts pay what it costs to educating a students in those towns, which is generally less than the average cost of educating students in Amherst. If we expand the number of students who reside in Amherst and live in residences that pay property taxes, there should be a net gain to the Town. Professor Friedman is either unaware of this or simply overlooks this complication in his analysis.
There are so many avenues to pursue here, and I thank Mr. Friedman and Ms. Asik for getting this rolling. And thank you to the Indy for providing this forum.
It would be helpful too, to get some insights about the ratio of the Town’s revenue mix – commercial versus residential tax base – relative to other communities and also over time.
Again, thank you for this. A great start.
I thank Steve George and John Hornik for their comments! They are, of course, absolutely right that when small changes in enrollment are considered, adding one or even a few students to our schools, it is more appropriate to use marginal costs to assess the cost of the additional students. We can accomodate a small change in enrollment without requiring additional teachers or classrooms so the marginal cost of educating those few students will be much below the average. But when larger changes are considered, such as new housing developments or zoning changes to allow more family housing, we are talking about adding many students requiring additional teachers and facilities. Changes like these would involve costs closer to the average than the marginal so it is appropriate to use average costs in assessing the impact on the town’s budget.
Thank you, Professor Friedman, for starting the ball rolling on this. I appreciate that you point out that there are costs to the Town associated with increased development. Some that spring to mind are parking (if it is not provided by the developer), road repair, increased water/sewer expenses, etc.
Re average spending per student, I’m curious what figure you used in the table. When I look on DESE’s website, in 2019 the Amherst (elementary district) Total Per Pupil Expenditure was $23,090 (In-District Per Pupil was $23,693).
On a property assessed at $550,000, with annual tax revenue of $11,726 ($550,000 x $21.32), wouldn’t it fall short of covering the average cost of educating even one child?
Also, I understand that there are economies of scale with student enrollment. A greater enrollment will lower the per pupil cost since there are many fixed costs that would be spread across more pupils (such as, facilities, utilities, principals, assistant principals, nurses, custodians, etc.) . And I think state aid is set at a rate per pupil (plus a portion of Special Education costs where appropriate) so more students brings in more state aid.
Lastly, with respect to John Hornik’s comment on School Choice, I understand that the revenue/expense has been fixed at $5,000 per child for many years, with an extra allowance to cover any special education services the child may require at their receiving district. In general, the district considers School Choice funds a net gain since in theory students are only added when doing so does not require additional classrooms/teachers. However, this may only apply to the first year they are admitted since shifts in enrollment happen each year at each grade as people move in and out of the Town. In subsequent years, the number of School Choice-In students in a given grade at a given school may actually tip the balance into requiring another classroom/teacher.
I appreciate Prof. Friedman’s reply to my comment – I see his point. It still might be useful to present the analysis with both average and marginal costs for comparison, since it seems like it would take unusually (for Amherst) large developments with many children for the average costs to apply.
Each year (2005-2008) that I served on the Amherst Select Board, when the time came to “set the tax rates for the fiscal year” (I’ll spare the details — in part because I don’t recall them, and they have likely changed by now — but essentially this was to determine whether commercial property would be taxed at a higher rate than residential property, and whether an exemption on the first N-dollars of assessed value might granted to owner-occupied properties) the issue was raised (by yours truly) whether to do either of those things.
While I tried to frame this issue in terms of real estate taxes being progressive or regressive, I always was disappointed that — despite being part of what was oft-called “the progressive majority” during that 3-year interval — I rarely got a “second” on motions that would have made our local taxes more progressive, and if I did (so there could at least be a discussion), I don’t recall ever having another Board member vote with me.
It would be greatly appreciated if Gerald Friedman and Maral Asik would address (perhaps in a future analysis) the inherently regressive nature of real estate taxes, and whether the above-mentioned — yet still rarely used, at least outside of Boston — mechanisms to make them less regressive, are really up to the task of equitably funding local government services, especially public education.
In effect,
There are several takeaways from Professor Friedman’s analysis. One is that older people, whose children are finished with their K-12 education, are very profitable for the town. I still pay the full property tax on our home, same as when I had children in the school system.
If you google “retire in a college town” you’ll see many articles about the advantages, and several cities and towns who value that important ingredient (seniors) in the mix.
In Amherst, I know of several people who are considering moving from here, because of high taxes, big tax increases on the horizon, and turned off by the steady transformation of Amherst Center into large dormitories with a minimal nod to ground floor businesses.
It is established fact that it’s much easier to satisfy and retain a current customer than to find a new one. I suggest to our town leadership that as you seek increased tax revenues from student housing in town, there is a growing hole in the bucket, of who’d hoped to “age in place” and are now thinking of where to move.
There are problems with property taxes when real-estate values change sharply possibly imposing burdens on homeowners where the property/income ratio rises sharply. There are also equity issues because some owners of equivalent property have different incomes while paying the same in property taxes. In general, however, property taxes are progressive, even highly progressive, because wealth and real-estate property increases with income.
Bryck is absolutely right about the advantages of keeping current residents in town. I would repeat his statement: “It is established fact that it’s much easier to satisfy and retain a current customer than to find a new one.”
Almost every real estate tax in the United States is at best a flat-rate tax, not a progressive-rate tax. While Gerald Friedman and I agree that higher income and higher wealth people (and corporations) tend to own more real estate, and thus pay more in real estate taxes overall, I hope he and would also agree that effective real estate tax RATE tends to DECREASE with income — and most people understand that to be the definition of a REGRESSIVE tax. On the other hand, exempting part of the of the averaged assessed value residential property for owner-occupied from the real estate tax — see, for example
https://www.mass.gov/doc/living-with-the-residential-exemption/download
for a discussion of how this provision of Massachusetts General Laws (MGL c. 59, sec. 5C) works, along with a list of municipalities in Massachusetts that do it and at what rate — makes it a more progressive tax. So I renew my request that he and his colleague evaluate this possibility for a community like Amherst.
Very interesting that people feel the burden of the high taxes in a town that has challenges, these numbers are from the town’s master plan from 2010:
“Dominant land uses include residential (23 percent of land area), protected agricultural (18 percent), conservation (18 percent), and land owned by Amherst’s three institutions of higher education (16 percent). The town has a relatively small amount of land (3.6 percent) designated as commercial, retail, or industrial zones. This breakdown shows that we are trying to generate enough capital to operate our town with 52% supplying little to no tax revenue. If you look at the balance of the “taxable” land use, 3.6% is designated to some form of commerce and 23% is for residential use. This ratio of 1:6.39 puts a significant burden on the latter to produce enough revenue to support investments in our town and its services.”
The way the article (economics in a small college town) is framed, it makes it seem like a burden to have families in town based on property taxes, and the cost of education but I would argue that there is a velocity of capital that exists when the family is spending money on clothes, recreation, food, education etc. Whereas retirees do have consumer behavior that parallels these actions, I am led to believe they are not buying clothes as frequently, have different recreational expenditures and there is a greater likelihood that they are traveling and spending their capital elsewhere in the world. Let us also take into consideration the value that young workers bring to a community when they are the ones willing to work the service industry jobs that add vibrancy to a community, i.e. entertainment, restaurants and retail.
Just my two cents.
@Rob Kusner. Why do you think that the real estate tax rate decreases with income?
Real estate is very inequitably held, with high-income households holding vastly more real-estate wealth than do low-income households https://www.federalreserve.gov/releases/z1/dataviz/dfa/distribute/chart/#quarter:125;series:Real%20estate;demographic:income;population:1,3,5,7,9,11;units:shares shows that the richest (in income) 1% of households own 12% of real estate wealth and the richest 20% of households own 56% of real estate wealth. The poorest 20% of households hold 4 percent of real-estate wealth and the poorest 40% of households hold 13.5% of real estate wealth. These figures are not out of line with the distribution of income shares (https://eml.berkeley.edu/~saez/saez-UStopincomes-2018.pdf). That is, there’s no indication that property taxation is regressive. In fact, if you think that wealth rather than income is a fairer basis of taxation, then property (real estate) tax may be more progressive than income tax (although not as progressive as a tax on financial wealth, where the big bucks lie).
For Amherst, life cycle is probably quite relevant — as Jerry Friedman suggested. It’s not clear why retired households that had high incomes during peak earning years and remain wealthy should not share in the obligation to pay taxes to maintain the town.
ASIDE: there’s no question that LOCAL taxation is regressive — Holyoke has and would have a very unfair shake relative to Amherst as long as the base, be it real estate, income, or, for that matter, financial assets, is taxed LOCALLY. That is a statewide disgrace but not relevant for this particular question.
SECOND ASIDE: An open question — I’ve read the literature and can’t find anything definitive — is how the property tax is split between renters and landlords. There’s good reason to believe that property owners bear a decent share of the property tax, i.e., they cannot pass all of it along to renters. If that’s the case, then the property tax is really progressive: renters in Amherst, who tend to be poorer than homeowners, may well pay a very small share of property taxes.
Hi Michael: I don’t “think that the real estate tax rate decreases with income” and in fact I wrote that it “is at best a flat-rate tax, not a progressive-rate tax” – and I included also information on the specific provisions of MGL that several Massachusetts municipalities (including Boston, Cambridge and Somerville) have adopted to make the real estate tax more progressive in their communities.
Your efforts to cite documents to support the position that “there’s no indication that property taxation is regressive” are appreciated, but to state “the richest (in income) 1% of households own 12% of real estate wealth” actually supports the point I made earlier: the “effective real estate tax RATE tends to DECREASE with income….” This is because (according to https://www.usatoday.com/story/money/2020/07/01/how-much-you-need-to-make-to-be-in-the-1-in-every-state/112002276/ ) the share of all income earned by the top 1% of US households by income was about 20% as of July 2020 (in Massachusetts it was even higher: 23.7%), and so their effective real estate tax is LESS THAN HALF that of the lowest 25% of US households (who earn less than 4% of the income, but own – according to your own figures above – more than that fraction of the real estate wealth).
So, please think more about what I wrote (and asked Jerry to consider), Rob
…and so their effective real estate tax *RATE*is LESS THAN HALF that of the lowest 25% of US households…
Or to say it another way, Michael (and Jerry and Maral): it appears that the FRACTION OF INCOME paid toward real-estate taxes by the lowest-income 25% of the US population, is more than DOUBLE the FRACTION OF INCOME paid toward real-estate taxes by the highest-income 1%. But however one cares to say it, doesn’t that sound like a regressive tax? And isn’t it important to understand why Amherst – unlike Boston and Cambridge and Somerville and … – does not also avail itself of a tool authorized by MGL to address this? (Maybe there’s a good reason? If so, let’s hear about and understand it….)
@Rob Kusner You don’t think you wrote it but you literally did: “I hope he and would also agree that effective real estate tax RATE tends to DECREASE with income” (see above).
I like the plan you cited and agree it would increase progressiveness. No argument there.
Nonetheless you should not go around declaring that the property tax is regressive without better evidence. The figures I cited indicate that property wealth generally tracks income quite closely. You have not accounted for (a) renters who literally pay no property tax in the sense of writing a check to Amherst but obviously pay some via pass through in rent. If you know how much, cite it; or (b) households with high lifetime income (and high wealth) but low income now, ie, well-to-do retirees. I would take consider relatively high property taxation of such household to be progressive based on lifetime income despite the high ratio of property taxation to current income.
The progressivity of the property tax in Amherst is an open question. There is no particular reason to think it is less progressive than an income tax. It is undoubtedly less progressive than a financial wealth tax.
As long real estate taxes remain the primary way we fund public services like K-12 education, the question of whether they are effectively regressive or progressive is important. I’m happy to see that my friend Michael Ash “like[s] the plan [I] cited and agree[s] it would increase progressiveness.” And while the additional words “effective” and “tend” show that I did not *literally* write what he claims I wrote, I concede that more evidence for what I claim to be true is needed. What can be easily find at the “macro” level, I’ve provided; getting more detailed, granular evidence – say at the level of a community like Amherst – would be valuable. And that’s what I’m asking the economists (like those writing columns here in the Indy) to do. As a concerned citizen, scientist, and mathematician, I’m even willing to help them with such a project….
I completely agree with Rob’s note. We need to understand if the property tax is progressive (tax paid relative to “ability to pay” increases with “ability to pay”) or regressive (tax paid relative to “ability to pay” decreases with “ability to pay”). Fill in “income,” “lifetime income,” “wealth,” etc., for “ability to pay” (and the choice of metric may matter). The public economics literature is not definitive on the question. More granular local data might help with Amherst in particular. The data demands are high because property valuation and income, let alone lifetime income or wealth, are not typically reported together. Theory is also needed (with respect to renters, landlords, and the property tax). So I don’t know the answer, and I am working on it. But it’s not prima facie obvious that the property tax is regressive within Amherst.
Maybe IRS data would have the property tax (claimed as a deduction) in tandem with current income data — but (1) that would be limited to itemized tax returns rather than standard deduction tax returns; (2) it would address only current income, not lifetime income or wealth; and (3) it can’t address the rent question. The American Community Survey Public Use Microdata has individual household-level household income and house value data but the geography is too broad (https://www2.census.gov/geo/maps/dc10map/PUMA_RefMap/st25_ma/puma2500200/DC10PUMA2500200_001.pdf) and problems (2) and (3) aren’t solved. This really can’t be done without data and theory. Anyway let me know if you have ideas.
ASIDE: The grotesque regressivity of LOCAL taxation can be addressed by increasing state contributions to local resources, i.e., relieving the burden on localities to generate their own revenue, and increasing the progressivity of state revenue collection. The upcoming “Fair Share Amendment” will do both.
I’ll keep Rob, Jerry, and the Indy posted.
Thanks, Michael — I really look forward to that!
While I hesitate to add a few “calculus-like” points, maybe they’ll enlighten some of our readers:
• what you call “tax paid relative to ‘ability to pay’” is what I called the “effective tax rate,”
• and we’re interested in whether this “effective tax rate” increases (progressive) or decreases (regressive) with “ability to pay.”
Another way to say and see this:
• “progressive” means the graph of “tax paid” bends upward as a function of “ability to pay”
• and “regressive” means the graph bends downward.
Measuring which way it bends now (or at any other time) is an econometric (or historical) question.
Deciding which way it should bend is a policy (or moral) question — a question of justice.
For those of you following this discussion you may be interested in this New York Times article on the systematic over-assessment of the value of houses in poor neighborhoods resulting in poorer people paying a disproportionate share of property tax. The excess burden generated for Black and Latinx homeowners is particularly high. The main reasons: (1) richer people systematically and successfully challenge their valuations; (2) periodic reassessments are infrequent and systematically pessimistic relative to sale-based assessments; and (3) comparing houses in disadvantaged neighborhoods to “comparables” in a range of neighborhoods (and vice versa) makes asymmetric mistakes to the disadvantage of people living in the disadvantaged neighborhoods (and vice versa).
https://www.nytimes.com/2021/04/03/opinion/sunday/property-taxes-housing-assessment-inequality.html
This phenomenon, if it applies in Amherst, would augment Rob’s case for regressivity in the property tax. A good project would be grabbing Zillow-type estimates or sales data and comparing them to recent assessor valuations for the same housing unit.