What can we learn from Providence's assessment data?

I scraped all of Providence's assessment data as found on its Vision page and then mapped the results here, finding that whites are more likely to own homes (especially single families and condominiums), that racial ownership is geographical, that we have far more vacant land than affordable housing, and that property wealth is concentrated on the East Side. These are all fairly predictable outcomes but arresting when visualized.

The map presents these outcomes and more, such as the location of Tax Stabilization Agreements and nonprofit land. Just click a radio button in the right-hand panel and observe. You also can search for addresses with the magnifying glass on the left-hand side. Note that, given the amount of data points, you shouldn't toggle too many features.

The appendix lists caveats, limitations, clarifications, and all that stuff. You can download the full spreadsheet here, if that's your thing.

Below I present what I consider some interesting findings, for those who don't want to explore. But you really should, as it's kind of fun; you can zoom around neighborhoods, click on properties for more information, etc. Just wait a couple of seconds for the map to load because it holds a lot of data points.

Where's the money at? The East Side, naturally?

Of course, where the white people live (see next question). The below figure shows Providence's overall assessment values, ranging from yellow (~$200,000) to purple ($1 million plus). You really want to live east of the Boulevard.

Who owns Providence? It's basically white people, right?

Yup.

Assessors obviously do not publish demographic information, so I used a name guessing model to estimate the probability of a property owner's race as Asian, Non-Hispanic White, Black, or Hispanic. Unsurprisingly, the model estimates that property owners are more likely to have Non-Hispanic White names for all property types but especially single family homes and condominiums. Hispanic names are more likely to be associated with multifamilies (but still less than white names). You can read more about this model, as well as discover some boxplots, in the appendix.

Anyway, two static maps follow below, where blue/purple means a property owner is very likely to be Non-Hispanic White, based on the first and last names, and red means the owner is very likely to not be Non-Hispanic White.

Jacob Binder wrote about similar segregation in The Brown Political Review, and he had even included what looks like a cool map based on 2010 Census Block data. So, I'm here corroborating what's already well-understood, though this time from the perspective of home ownership. We also see that whites, while associated with home ownership everywhere, are far less associated with multifamily home ownership outside of the East Side.

Single family homes, estimated probability of property owner being Non-Hispanic White


Multifamily homes, estimated probability of property owner being Non-Hispanic White


How much of the city's property wealth do white people own? One way to estimate this would be to calculate the expected monetary value for each property. Suppose that a property is worth $500,000 and our model believes the owner has a 90% chance of being Non-Hispanic White. Then we can calculate the Non-Hispanic White's assessment "share" as 500,000*0.90 for $450,000, with $50,000 divided among the other categories. Of course property values are not divvied up like this. A Non-Hispanic White person cannot own 90% of a home, in most circumstances. But expected value, especially across a large sample like here (over 30,000 residential properties), should be more or less useful, provided that the underlying name guessing model works.

Applying this calculation to all residences yields quite a disparity. I applied a 5% penalty to Non-Hispanic White names, shifting that amount instead to Black persons, on the assumption that the model over-estimated white ownership (as explained in the appendix).

White people have an expected ownership value of almost $5 billion, with Hispanics coming a distant second at $1.6 billion. A much larger penalty than 5% could be levied and white ownership would still dwarf the rest. Once again, not terribly surprising, though arresting. Non-Hispanic Whites are estimated to comprise around a third of the city's population. And they, at least according to this calculation, own around 63% of the city's property value wealth. Further, as the appendix explains, this only considers individual homeowners, not businesses, so if we were to factor in businesses then white ownership would be much higher.

But there's more affordable housing than vacant property?

No. There are 431 acres classified as vacant land (some addresses were very unclear so I removed them or they were reserved for the utility company) compared to 284 acres of affordable housing.

Below, black lots are vacant and purple are affordable housing. Obviously not all of these lots are developable, but they may also be an undercount (see appendix).

What did I consider affordable housing? I chose public housing and any property listed as accepting Section-8 vouchers or online as affordable/low income housing. So, to keep things simple, I lumped all of the housing together, even though they're not uniform; "affordable housing" often is not very affordable for low-income residents. This criteria might miss properties that aren't listed online or in Providence's database, but I may be overcounting housing because I assumed that all units on a property were affordable when that may not be the case.

Is this the city's fault? The lands are privately owned, after all. Even if we could negotiate with owners to extract their properties, building or repairing homes would cost a fortune. Some of the lots are on former industrial sites and nothing can ever be done with those areas, ever. Federal limits restrict the amount of housing we can build. And many of these vacancies are so small that no one would want to live on them.

Well, at some point, you know, you have to solve crises. You can't just keep marinating in them. I mean, sure, you can, and many politicians do, but if you care about solving problems and not ascending to higher and higher forms of office, then you have to suspend talking points and get to work. New York City tried to do something with its vacant land and properties; it didn't really work, and the city's leaders per usual rewarded for-profit lobbyists, but at least something happened.

Ok, what about the square footage of affordable housing?

Acreage doesn't much account for density; build a high-rise apartment on an acre and you certainly have much more living space than single family homes on five acres. So, I lumped square footage into the categories of affordable, luxury, and market value housing. I only included a property as "luxury" if it was advertised in some way as a luxury complex (apartments or condos), not if it was just really expensive.

The exact numbers are market = 68,404,021, affordable = 7,519,828, and luxury = 2,482,123.

Livable square footage in market value properties obviously dwarfs that of affordable units. I don't know what a proper ratio should be, but I doubt it looks very much like this for a city with Providence's poverty; we have about one third luxury space as we do affordable housing. This also doesn't include units designated as student housing, including the "luxury" types as found in, for example, Brown's River House.

Provided we have enough affordable housing, I don't care how much luxury space we have. But we don't have enough affordable housing. Not even close.

Do homeowners even live here?

I suspect it's better all around for homes to be owner-occupied: more investment, financially and emotionally, and fewer slum lords. It's not about single vs. multifamily homes; multifamily homes are great (I live in one), and I would think especially so when they're owner-occupied.

Even if I'm wrong and owner occupancy makes no difference in property management, Providence's owner-occupied homes are more likely to be minority owned. The below figure shows the median Non-Hispanic White percentage for each property type, separated by owner occupancy. As is clear, minority ownership's strongly associated with owner occupancy for all multifamily properties.

I compared the addresses of each property with that of its owner, and indicated that it was not owner-occupied when the streets did not match, on the assumption that some multifamilies have different numbers despite being the same property. Below is the percentage of owner-occupied homes by category as well as the percentage of owners who live in Providence.

So, if you walk past a five-family home, there's a very good chance it won't be owner-occupied, and almost a 50% chance the owner won't even live in Providence. A four-family is a coinflip of being occupied.

Are there high numbers or are they normal? I suppose it could be normal like it's normal nowadays for thousands of people to refuse ultra-potent vaccines and then die from a disease against which these same vaccines offer protection, but that doesn't mean it's OK. I would argue it is, at least potentially, problematic that so much of our city's owned by people who don't actually live here.

An obvious caveat is that the owner's address is a crude measure for who owns the property. Maybe an elderly man lives in "his" single family home but the registered owner is his out of town daughter. Someone could occupy a home but use a PO BOX, so the addresses don't "match up." On the other hand, sometimes the owner's address is not really the owner's address, so we may be overcounting the amount of owner-occupied properties. 66 Williams St., for example, is owned by an LLC that has its address listed as 66 Williams St., giving the impression that "it" lives there, but its business address (as filed with the Secretary of State) is 3 Davol Square. I also personally know a few properties that are definitely not owner-occupied but list the owner's address as the property's address.

And what about these business owners? Businesses own about 11% of two, three, four, and five family properties and 4% of single family homes, i.e., the owners are an LLC, partnership, inc., or other corporate sounding name. While spread throughout the city, they seem most concentrated in Elmhurst (student housing mostly) and sort of in Federal Hill and Fox Point.

So about those nonprofits.

I really don't have much to add to the conversation here, but if I didn't comment on nonprofit space then I'd be remiss. Nonprofit space accounts for 748 acres, of which 386 are classified as university property. The total assessed value of nonprofits is $4.83 billion, with universities making the largest share, as seen below.

Schools are private, and support relates to any nonprofit which seems to offer support services.

Obviously the greatest university landowner is Brown, with 146 acres assessed at $1.38 billion; this is probably an undercount, though, because we know Brown owns desirable real estate through at least one subsidiary (which I have included here).

By the way, Brown's holdings represent about 6.4% of the city's entire assessed value. This assumes, as well, that we're accurately assessing Brown's property (and that of large institutions generally), which seems unlikely given the variance between assessment values and real estate transactions. So, we can say that the most conservative estimates put Brown at owning around 6% of our city's assessed value.

General question. How old is Providence's housing?

As is to be expected, pretty old. Most of the homes were built in 1930 (5,403), and 90% of them were built before 1978, meaning that they very likely have lead paint. So don't feel too bad that your house has lead. Almost everyone else's does, too!

One positive to Providence's limited public housing stock is that it is relatively new. Almost of Providence's public housing came after 1990, in stark contrast to the rest of America, where the Urban Institute estimated that 42% of public housing properties were built prior to 1975.

The bad news is that we just don't have much affordable housing, as indicated above. And Section 8-eligible properties are much dolder.

Pretty random question but how much land does former mayor Joseph Paolino Jr. own through his property development firm?

Maybe you didn't know that we seem to rent the Doorley Municipal Building for over $700k a year to Joseph Paolino Jr.'s firm (I say seem because I've yet to have a city official confirm it). Maybe that sounds crazy to you, too, and maybe after finding this out you got curious about what else Paolino owned.

As these guys use lots of shells–I mean, auxiliaries and subsidiaries–you can't just search ownership records for their regular business names. However, they often use the same lawyers, mailing addresses, and directors, so you just search for those in the owner address field. The appendix has some R code to help facilitate this search.

Paolino owns at least 91 properties, across 22 acres, assessed at a total of $140,815,500. Not too shabby considering it's only what can be found through registered agents. With almost 7 acres and 390,976 square feet alone in parking lots (paid or decrepit), it's no wonder he so vehemently rejects any attempt to reduce car congestion.

What about that Walter Bronhard guy who GoLocalProv reported on? He's got 51 properties across 10 acres assessed at $41,597,500. Like, whoa, real estate guy.

Obviously the assessment data do not reflect acquisitions which have yet to be entered in the database or assessments on newly renovated properties, such as Paolino's Beatrice Hotel.

Tax Stabilization Agreements and My Shameless Editorializing

Providence has signed TSAs for 82 acres of property and $851,044,200 in assessed value. I don't know much about these specific agreements, so those who do would be more qualified to examine what's being received vs. what's being assessed vs. what's being produced. I do know the party line, though, which is that TSAs generate "growth" and "activity," that they lead to construction jobs, that they "revitalize" blighted areas, and, more poetically, that they're a beautiful marriage between the public and private sectors.

At some point you have to take stock of what your policies have done. You might think that all of these corporate and special interest handouts have failed–we have more vacant land than affordable housing, we're facing widespread discrimination in not only renting but also home ownership, and we're watching as a superminority of white people sit on the city's most valuable properties and complain about marginal tax increases.

Then you realize these policies are working exactly as intended. It's no accident that they have engendered major levels of discrimination. Providence considered selling off its water supply because it could not outrun its pension deficit; meanwhile, we give millions upon millions in tax breaks to downtown developers with vague promises that they'll pay off.

As the old saying goes, the purpose of a system is what it does.

And what, exactly, does our system do?

The River House apartment complex received over $9 million in TSA subsidies. Brown University, which you might naively but understandably believe is a university and not a hedge fund property developer, bought it for $75 million, possibly netting the developer $22 million; the same developer who profited from another real estate deal with Brown, a conflict of interest disclosed well after the fact.

These maps show you the Providence that they've built, and not just yesterday but across many generations, across many councilpersons and many administrations. Mayor Elorza calls it the City that Works.

We naturally are left asking, "Works for whom?"


Michael Carlozzi, 2022, contact = car lot see @ gmail dot com. You can use these data and maps wherever you'd like, pursuant to CC-BY-US. Maps created in R.