In my last post I sort-of-promised that that would be my final commentary on all matters Mozilla. I am breaking that promise, but will console myself with the thought that this post is only going out to the handful of subscribers to my Substack newsletter (which I use for less formal posts), and I won’t be tweeting about it or otherwise promoting it. But of course every post of this newsletter is publicly viewable and sharable, and many of you have Twitter accounts too, so that’s not much of a concession.

Anyway, on to Mozilla. One thing I regret not discussing in my prior posts was the sort-of-announcement of the extension of the Mozilla Corporation’s deal with Google. I say “sort of” because there was no formal press release or blog post. Instead the news was leaked to the press through back channels. To quote The Register: “Within hours of the browser maker laying off a quarter of its staff this week, a well-placed source told The Register Moz had signed a three-year agreement with Google.”

The key words here are “within hours” and “well-placed source.” This has all the hallmarks of Mozilla Corporation senior management trying to correct a public communications problem, after a layoff announcement that not only was silent about the fate of popular Mozilla-sponsored projects like Rust or MDN, but also by implication cast doubt on the long-term future of the Mozilla project as a whole.

The Register’s story goes on to say, “Our source told us Moz will likely pocket $400m to $450m a year between now and 2023 from the arrangement, citing internal discussions held earlier this year.” The key words here are “likely” and “internal discussions”. The reference to “internal discussions” (within the Mozilla Corporation? between the Mozilla Corporation and Google?) might imply that the deal was anticipated at press time but not yet signed. However as reported in the same story, a Mozilla spokesperson did confirm that the contract had been extended.

So, if the deal was done, why is the amount of revenue to Mozilla characterized as a “likely” range rather than a fixed amount? A clue here is in Mitchell Baker’s blog post announcing the layoffs: “Economic conditions resulting from the global pandemic have significantly impacted our revenue.” If the contract was for a fixed amount, payable no matter what, then COVID-19 should have had no impact. So the implication is that the contract value was dependent on certain metrics whose values could fluctuate, for example, based on the number of active Firefox users and/or their web browsing activity.

That implies that the estimated annual contract value of $400–450 million is dependent on maintaining Firefox market share or even possibly increasing it somewhat (which was possibly the main subject of the “internal discussions”). By comparison, from the 2018 Mozilla consolidated financial statement [PDF] we see that “royalties” (the category used for search engine revenue) were $430 million in 2018 compared to $539 million in the previous year (page 4, “Consolidated Statement of Activities and Change in Net Assets”).

However, royalties include more than just search engine revenue. Actual search engine revenues were about $391 million in 2018, 91% of overall royalties, down from $501 million or 93% of royalties in 2017 (page 25, note 12). So the estimated annual $400–450 million in search engine revenue for the next three years, while a conservative estimate compared to the peak year of 2017, is still relatively optimistic compared to the actual search engine revenue in 2018.

The change from $501 million to $391 million represents a 22% decline in search engine revenues from 2017 to 2018. Even assuming that the estimated revenues of $400-450 million are achieved going forward, that’s still a 10–20% decline from the peak year of 2017.

The “2018 State of Mozilla” report noted that overall Mozilla employed over 1,000 people. (It’s unclear whether this is as of the end of 2018 or as of November 2019, when the report was released.) Almost all of these people work at the Mozilla Corporation, so we can take 1,000 people as an estimate of the size of Mozilla Corporation staff at that time. (This same report noted the Mozilla Foundation as employing 80 people, while Part VII, line 2 of the Mozilla Foundation 2018 Form 990 listed the number of Foundation employees with compensation greater than $100,000 as 43.)

The Mozilla Corporate laid off 70 people early in 2020. Combined with the more recent reduction in 250 “roles” announced in the Mozilla Corporation internal memo [PDF], that amounts to a decrease in staff of up to a third from its highest point. Presuming that the Mozilla Corporation had staffed in 2018 in anticipation of 2017-like revenues continuing, and that 2019 search engine revenues were comparable to those in 2018 (which we’ll find out this November), that explains why such a large fraction of the Mozilla Corporation workforce was laid off: That’s the reduction in staff that was needed to bring costs in line with anticipated revenues for the years ahead.

Another topic I wanted to expand on was the issue of individuals paying to support Firefox development. Anyone perusing Reddit, Hacker News, or other online forums will encounter comments urging people to contribute to Mozilla to help keep Firefox alive. As I previously noted, there is in fact no direct way for individuals to pay directly for Firefox itself, since donations to the Mozilla Foundation go to support Mozilla Foundation activities, and do not find their way to the Mozilla Corporation where the actual Firefox developers are employed. Even with new products like Mozilla VPN the money users pay will support Firefox development only indirectly.

However, even if users were able to directly fund Firefox development the resulting revenue would most likely fall far short of the funds needed. To get a sense for this, let’s look at the current level of Mozilla donations:

The 2018 Mozilla Foundation IRS Form 990 [PDF] reports just under $14 million in revenue from contributions and grants (Form 990, Part I, line 8). But of this almost $9 million was received in the form of large contributions (Schedule B, Part I). So the amount contributed to the Mozilla Foundation by ordinary Firefox users was no more than $5 million at most and may even be lower, since it’s not clear whether the list of large contributions includes the over $3 million from government sources noted on Part VIII, line 1e.

The disconnect is clear: in their generosity individual Firefox users collectively would likely directly contribute at most a few million dollars per year to Firefox development, while the actual cost of Firefox development is one or two orders of magnitude greater than that. I was probably guilty in my previous Mozilla post of being overly pessimistic about the long term prospects of the Mozilla project. But there’s no denying that this is a major gap to fill, and it’s understandable why Mozilla Corporation senior management have not yet found a way to fill it, beyond depending on search engine royalties from a third party that owns a “cash cow” business with a dominant market position.

This is a specific instance of a more general problem with software development today, which I may address in a future post. As for Mozilla, I’ll likely be back with an update after the 2019 Mozilla financials are released in November, unless there’s major news in the meantime.

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

Some thoughts on the Iowa caucuses vs. the Irish general election.

There’s a saying in the world of information technology: fast, cheap, or good, pick any two. It captures the inherent trade-off between the desire for a solution that provides high performance and/or rapid response times, that’s doesn’t cost a lot of money, and that actually does what it’s supposed to do without errors or failures.

This saying applies in other areas as well, including elections. We saw this most recently in the Iowa Democratic caucuses, which devolved into a spectacular mess of non-working apps, jammed phone lines, voter confusion, and misinformation (and sometimes disinformation) propagated by media pundits, politicians, and activists all vying to declare definitive results as soon as possible.

The Iowa caucuses involved a mere 176,000 people, comparable to the number of people who turn out in a local Howard County election. Across the Atlantic ocean this weekend there’s an election involving an order of magnitude more voters. If past experience is any indication, while the election results may be surprising (see my next post) the actual process of getting to those results will be reassuringly boring.

I’m referring to the general election in Ireland—or the Republic of Ireland, to distinguish it from Northern Ireland, which is part of the UK. The election is to select representatives to the Dáil (pronounced “doyle”), the lower house of the Irish parliament (comparable to the House of Commons in the UK).

To take the “good” aspect first: the Irish voting system does a good job of producing results that reflect what voters actually want, and is pretty simple for voters to use: voters fill out a single ballot sheet listing all the candidates for the (multiple) Dáil seats in their particular constituency. They then put numbers beside the names of each of their favored candidates, ranking them in order of preference: 1, 2, 3, and so on. Then they turn in their ballot. That’s it.

(Incidentally, the Iowa caucus rules are also designed to also reflect voter preferences among multiple candidates, but they don’t do as good a job, and are much more complicated for voters to understand and follow.)

As implied above, Irish elections do not use any sort of electronic voting equipment. This reduces capital expenses for elections and removes the possibility of voting machines or other computer-based systems being hacked. This checks both the “cheap” and “good” boxes.

So, what about the “fast” box? Recall the catch: “pick any two.” Voting in the 2020 Irish general election was yesterday, Saturday, February 8. However, the final results of that election are not likely to be known until Monday or Tuesday at the earliest.

Why so long? Because by making the voting system more likely to reflect voters’ preferences, simple for voters to use, not needing electronic equipment, and not vulnerable to hacking, the vote count itself becomes a very people- and time-intensive process. Greatly simplified, it goes something like this (see this “dummy’s guide to election count” for a more detailed explanation):

After collecting paper ballots and storing them securely overnight, the next morning election workers open the ballot boxes in the presence of political party representatives (as shown above). The vote counters sort the ballot papers by voters’ first preferences: one pile for ballots noting candidate A as “number 1,” a second pile for ballots noting candidate B as “number 1,” and so on. They then count the number of first preference votes for each candidate and the total number of votes.

Based on the number of seats, the total number of votes, and the vote shares for each candidate, it’s possible that one or more candidates receive more than enough first-preference votes to be elected without further ado (i.e., “they exceed the quota on the first count”). If so, their excess votes (more than were needed to elect them) are distributed to other candidates who were indicated as the second preferences of voters voting for the just-elected candidate(s). A second count then occurs in which one of the remaining candidates now has enough votes or (failing that) the candidate receiving the fewest votes is eliminated.

This process of electing or eliminating candidates, followed by transferring votes to the remaining candidates based on voter preferences, continues until candidates are elected to fill all open seats. This typically requires several counts, each of which (again) requires manually inspecting the ballot papers. If the counting isn’t finished by the end of the day then everyone goes home and resumes the process the next day.

Americans, or at least American media, are notoriously impatient and desirous of instant gratification. Americans are also addicted to technology almost for technology’s sake, so that each new shiny thing that comes along (touch screens! the Internet! blockchain!) is hailed as the solution for voting that will truly and finally deliver on “fast,” “cheap,” and “good” all at once.

There are also many people with a vested interest in the current winner-take-all US electoral system, starting with the politicians and activists of the two major political parties. Thus it’s difficult to see the US ever adopting the Irish voting system or anything similar to it, at least at the national level. And indeed if this ever does happen I suspect it will not be until a generation or two has passed, and resistance to the idea is overwhelmed by ongoing political conflicts that motivate peoples’ desire for fundamental changes to the American political and electoral system.

But whether it comes soon or not for a generation or two, I think the US would be well served to look to change its way of electing candidates to better match the way it’s done in Ireland. It’s possible that with appropriate electronic assistance it can be done in a way that’s quicker than typical Irish experience (though of course not without bringing some risk into the system).

Some US cities and states are now experimenting with a form of the Irish system, under the names “ranked-choice voting” or (when a single position is being filled) “instant-runoff voting.” For some thoughts on why this nascent movement is on net a good thing for the US, see my next post.

I don’t usually post articles commenting on what my fellow Howard County bloggers write. However I’m more likely to jump into the fray when important points need to be clarified—all the more so when I myself haven’t done a good job of being clear in my own blog posts.

The topic today is the oft-repeated claim that Howard County has a shortage of 5,000 (or 5,500) affordable housing units, and in particular the post “The High Cost of Affordable Housing” by the pseudonymous blogger “Shaykh.” In this post they write,

The current refrain of the far left is that Howard County “needs more affordable housing,” making claims such as “we have a shortage of 5,500 affordable housing units.” Why is there a shortage? Where are these 5,500 families that are without housing? Are they living in a tent city like we now see out in Seattle?

Shaykh then goes on to speculate:

If each of these 5,500 affordable housing units sees an average of 2.5 persons, that would equal 13,750 more residents, or a 4.3% population increase over the current 321,000 residents. This is more than 3.5 times higher than the current growth rate of 1.2%. . . . If each of these 5,500 affordable units has an average of 1.5 children (I feel this is a reasonable guess), that would mean 8,250 more children in HCPSS, a 14% increase of the roughly 58,000 students currently enrolled.

First, to nitpick: I think the number 5,500 should actually be 5,000, if (as I think) the source of this estimate is ultimately the the 2018 Howard County Rental Survey commissioned by the Howard County Housing Commission. If so, this is not Shaykh’s fault, as various politicians and activists have used the 5,500 number in their public statements.

More fundamentally, the estimate of 5,000 “missing” affordable units does not mean that there are 5,000 Howard County households who are homeless, or 5,000 households living elsewhere who would like to move to Howard County but cannot. Instead it is an estimate of the number of households already renting housing in Howard County who are paying more for rent than is considered “affordable.”

I already posted about how this estimate was derived, but looking back on that post I’ve concluded that I didn’t do a very good job of explaining things. So, I’ll try again:

The 5,000 estimate is based on a comparison to a theoretical ideal: If the housing market were a totally free market and if there were no factors artificially inflating the price of housing, then we would presumably see developers and landlords offering a range of housing options priced to be suitable for households in each and every income range, just as we do with other types of consumer goods. (Just as, for example, with clothing we see the free market offering options ranging from buying low-priced outfits at TJ Maxx to buying haute couture from Christian Dior.)

At the time the 2018 Howard County Rental Survey was produced, there were an estimated 9,545 households with household income of less than $50K who were living and renting in Howard County. To put this in context, at the time of the survey there were an estimated 116,711 households in Howard County, of which an estimated 32,358 households were renting. (See page 17 of the survey, Table 10, “Renter Household Characteristics.”) These lower-income households thus make up about 8% of all Howard County households, and about 29% of renting households.

So, if the housing market in Howard County were a totally free market and if there were no factors artificially inflating the price of housing in Howard County, then we would expect to see these 9,545 lower-income Howard County households renting units that were priced roughly in line with the incomes these households have.

However at the time of the survey there were only 4,486 rental units in Howard County that were considered to be “affordable” to those 9,545 lower-income households, with “affordability” based on the (somewhat arbitrary) guideline that a household should spend no more than 30% of its income on housing. The difference of 5,059 (9,545 minus 4,486), rounded off to 5,000, is then considered to be the amount by which Howard County falls short in terms of providing affordable housing.

So this has nothing whatsoever to do with enticing thirteen thousand poor people to move to Howard County, or adding eight thousand more FARM students to Howard County public schools. Those people are already living here, and their children are already attending school here. They’re just paying more than 30% of their household income for their rental housing.

As for how to make housing more affordable in Howard County, Shaykh’s position is that the best solution lies in making the housing market more of a free market, for example by loosening zoning restrictions. I agree with this, and said so in my previous post.

The fundamental problem with government policy on housing is summed up by the economist Arnold Kling as “restricting supply then subsidizing demand.” On the one hand, zoning regulations and other government-imposed restrictions (e.g., APFO regulations) result in less housing built than would otherwise be built in a totally free market, reducing the supply of housing and driving up its price.

Government then seeks to offset the impact of increased housing prices through various direct and indirect subsidies, including things like Section 8 vouchers (which attract the most attention and controversy) but also including providing tax credits for developers building affordable housing and allowing homeowners to take tax deductions for mortgage interest and real estate taxes.

These subsidies not only divert government spending from arguably more useful purposes (like improving transportation infrastructure), they also have the effect of increasing the demand for housing beyond what it would have been in the absence of the subsidies. This then further drives up the already high housing prices caused by the restricted supply of housing.

From this point of view the way to address the issue of housing affordability is not to create more “affordable housing,” it’s to build more housing, period.

After writing something I almost always think of things I wanted to say but didn’t have the time or space for. So it was with my last post touching on housing affordability issues in Howard County. Here are a couple of random follow-up comments,:

School redistricting and housing affordability

I have for the most part stayed out of directly commenting on the current controversy on school redistricting in Howard County. My family is not directly affected by it since we don’t have a child in the public school system, and I didn’t have the time or inclination to do the sort of in-depth research I normally feel is necessary before I make public comments on my blog or elsewhere.

However I did at least obliquely comment on the issue in my post “Moving to opportunity in Howard County.” To recap briefly, I suspect supporters of the superintendent’s proposed redistricting plan are likely over optimistic as to how much of a real difference the plan would actually make in the lives of students, especially students already in high school. (I think one could make a stronger case for redistricting at the elementary school level.) At the same time I definitely believe opponents of the plan downplay the extent to which the current situation is the result of deliberate government “social engineering” that has the effect of isolating lower-income households geographically and discouraging their living in Howard County.

Some people claim that rather than redistricting the school system we should be instead focusing on issues of housing affordability, making sure that lower-income households are able to live in any part of the county and take advantages of the schools in those areas. That’s a perfectly legitimate opinion to hold, and I look forward to seeing these people work to build political support for concrete proposals to do just that.

The “no public transit” argument against affordable housing

Still on the topic of affordable housing, Tom Coale (who is variously an attorney representing developers, a progressive activist, and co-host of the Elevate Maryland podcast), recently tweeted, “Time after time, inadequate public transportation is used as an excuse to reject affordable [housing] projects,” and claimed that “The median income in certain parts of Maryland is high enough that those eligible for affordable housing, based on percentage of median [income], make enough to own a car, but not enough to pay for housing.”

As it happened, that tied in nicely with the analysis I had been doing on new vehicle sales, and so I tweeted in response. To expand a bit on my comment:

The median new vehicle price is around $35,000, and according to the affordability criterion I used (the vehicle price should be no more than 50% of pre-tax income) such a median vehicle should be affordable to a household making $70,000 a year. Although such an income is above the US median income of approximately $62,000 a year, it is well below the Howard County median income of almost $120,000 a year, and so it’s quite possible such a household would have difficulty affording housing in the county.

If we turn to households making between $25,000 and $50,000, they should be able to afford a single vehicle worth $12,000 to $25,000, used or even new. They could even probably afford two high-mileage used vehicles. But, again, they would almost certainly have problems affording housing in Howard County.

The result is that, as Coales tweeted, “There is no record of people being ‘stranded’ because we built affordable housing without a bus route.” As he also pointed out, housing affordability is a problem across the state, even in areas where people absolutely need a vehicle to get around, and presumably are able to afford one one way or the other.

That’s it for follow-up for today, though I may have some additional comments later.

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

tl;dr: The story behind the claim that Howard County needs 5,000 more affordable units.

If you’ve been following Howard County matters you’ve probably heard someone claim that there’s a shortage of over 5,000 (sometimes 5,500) affordable housing units in the county. Have you ever wondered where that number came from? What it means? How it was calculated? If so, I am here to (I hope) shed some light on these questions.

To answer the first question, this estimate is from the 2018 Howard County Rental Survey commissioned by the Howard County Housing Commission. (Thanks go to Tom Coale of the Elevate Maryland podcast for indirectly providing me this clue.) More specifically, it’s from pages 81–82 of that document, which is worth quoting at length:

There are 1,250 multifamily subsidized rental units and another 1,886 multifamily rental units that are rent-restricted. Additionally, Howard County administers approximately 1,350 tenant-based Housing Choice Vouchers (HCV) of which 800 are county vouchers and 550 are Port-In vouchers from other jurisdictions. Assuming the unlikely scenario that no vouchers are used at tax credit communities, a combined 4,486 units are available to support the 9,545 low to moderate-income renter households in the county, leaving a gap of over 5,000 units, which provides a context for the county’s HCV waiting list of a comparable length.

So there you go. This appears to be the definitive number that’s bandied about (although I’ve also seen the number 5,500 used a couple of times): 1,250 plus 1,886 plus 1,350 adds up to 4,486 (affordable rental units), and 9,545 (households) minus 4,486 equals 5,059, the size of the claimed gap. (I’m not sure where the number 5,500 comes from. It may just be someone misremembering the report’s conclusion.)

Two additional points: First, the figure of 9,545 low to moderate-income renter households is from Table 11 on page 19, and comes from adding 2,524, 1,870, and 5,151, the numbers of rental households earning less than $15,000, between $15,000 and $25,000, and between $25,000 and $50,000 respectively. (In comparison, the number of rental households earning more than $50,000 is 22,813.)

Second, the 5,000 figure is buried in the conclusion of the housing survey analysis, not highlighted in the introductory summary. If you didn’t have the patience to read through the previous 80+ pages of very number-heavy text you would have missed it. That’s pretty casual treatment for a number that’s since received so much attention from politicians and activists.

You can stop reading here if you want to. But if you’re interested in exploring more of the story, including the “penetration rate” analysis discussed in the rental survey and whether or not the market for housing looks like the market for other goods and services, read on.

Let’s start with this graph, Figure 12 on page 80 of the survey, part of section D, “Penetration Rate Analysis.” (It’s also duplicated on page viii of the Executive Summary.) To quote from the beginning of that section,

By dividing the number of units in a specific affordability classification by the number of renter households that can afford or qualify for a unit at that price point, the penetration rate can tell us the extent to which renter households at particular income bands are adequately served by the existing supply.

For example, according to the graph, there are few “low rent” units targeting families earning 50-60% of the area median income, relative to the number of such families. In contrast, there are lots of “moderate rent” units targeting families earning 60-80% of the area median income, again relative to the number of such families. The former income range is said to have a low penetration rate (under-supplied), and the latter income range to have a high penetration rate (over-supplied).

(The survey divides Howard County into multiple sections and presents penetration rates for each one. Also, the “area” in “area median income” refers to the general Baltimore metropolitan area, not to Howard County specifically. Thus a household making less than 60% of the area median income actually makes significantly less than that compared to the Howard County median household income.)

There are several factors and assumptions associated with this analysis. Let’s explore this a bit more, moving from housing to other goods and services. As I see it, the penetration rate analysis used in this survey and elsewhere rests on one fact—that households vary in their income—and two assumptions: that in a free market profit-seeking businesses (but not necessarily the same businesses) will produce goods affordable to all households no matter their income range, and that in a free market the supply of goods targeted to households of a particular income range should at least roughly match the number of households in that income range.

Clearly households have a wide range of household incomes, as shown by the graph above, which displays the percentage of US households that have household incomes in certain ranges. (The definitions of the various income ranges are from the US Census Bureau, and reference household income before taxes.) As for the potential market for goods and services targeted to lower-income households, almost half of all US households have household income of less than $60,000 (the US median household income is about $62,000), and based on the graph it looks as if 20-25% of households have incomes of less than $30,000.

These are fairly large markets, and in a market economy it seems as if some businesses somewhere would find profitable ways to serve them. In the case of products like clothing that’s exactly what happens: we see retailers like Ross and TJ Maxx coexist with retailers targeting middle-income and upper-income households, like Macy’s, Nordstroms, Saks Fifth Avenue, and so on. All these are in turn supported by a global apparel industry capable of producing clothing at a wide range of price points, including relatively low ones.

The result is that households of all income ranges are reasonably adequately served by the products offered in the clothing market. Lower-income households might not be able to spend a lot on clothes, but there’s a reason we don’t see headlines about the “clothing affordability crisis.”

However we shouldn’t necessarily expect to see an exact matching of clothing of different prices to households of different incomes—that lower-income households buy only cheap clothes and upper-income households buy only expensive clothes. Some lower-income people want or need to present a more fashionable appearance, and will stretch their budget to do so. At the other end of the income scale, higher-income individuals do not typically spend every day dressed in haute couture.

The result is that a penetration rate analysis for clothing would likely not show 100% penetration rates across all income ranges, even in a completely free market. Instead we should expect to see higher penetration rates in the middle income ranges, and lower penetration rates at the extremes: relatively more mid-price clothing will be sold because there are lower-income households dressing up and higher-income households dressing down.

Let’s turn to a penetration rate analysis for another important class of goods, namely cars and light trucks. I was able to find the number of new vehicles sold for each model in 2017 along with price ranges for each model for the 2019 model year, and combined those to produce the graph above, which shows the percentages of vehicles sold that are targeted to each of the income ranges in the previous graph.

In creating this graph I had to make two key assumptions. The first was how to account for sales of vehicles where there was a large difference between the minimum price and maximum price. I tried to distribute sales for a given model across the price range in a reasonable manner, accounting for the fact that relatively few buyers purchase either the lowest price base model or the most expensive “loaded” model. See the full penetration rate analysis for details.

The other assumption, which also comes up in the penetration analysis in the rental survey, is how to decide what income range a particular vehicle is targeted at based on its price. Put another way, how do we measure vehicle affordability? After looking at various recommendations of the form “how much should I pay for a car”, I decided to use the following criterion: a new vehicle is affordable to a given household if its price does not exceed 50% of the household’s before tax income. Thus, for example, a household with an income of $60,000 should be able to afford a vehicle costing $30,000.

Put another way, given a vehicle offered for sale at a given price, a household’s income must be at least twice that price in order to afford the vehicle under this criterion. As it turns out, the typical (mean/median) price for new vehicles is around $35,000, so under this criterion we’d consider a typical vehicle to be targeted to households making around $70,000 a year before tax.

Combining the previous graph with the graph on the number of households in each income range produces the graph above, which shows the new vehicle penetration rates for each income range.

There are several points worth making about this graph. First, the details of the graph are obviously dependent on the assumptions made about vehicle affordability. If the criterion for affordability, i.e., a vehicle price no more than 50% of household income, were changed then the penetration rates for the various income groups would also change.

However the overall shape of the graph should remain the same, with under-penetration in the lowest and highest income ranges, and over-penetration in the middle income ranges. This general shape can be explained as follows:

In the higher-income ranges there are fewer vehicles sold than one might expect. For example, a household in the income range “$200,000 or more” should be able to afford a vehicle costing $100,000 or more, but relatively few such vehicles are sold (about a 25% penetration rate). This is probably due to two factors: 1) as with clothes, there’s an upper limit on how luxurious a vehicle a given household wants or needs, and 2) higher-income households almost certainly have multiple vehicles, for example buying two vehicles for $50,000 each instead of one vehicle for $100,000.

In the lower-income ranges there are also fewer vehicles sold than one might expect. Again this is likely due to two factors: 1) auto manufacturers find it difficult to sell truly low-cost vehicles in the US market, due both to government regulations and lack of market demand for small vehicles with limited or no amenities; and 2) some households buy vehicles that are more expensive than what they could supposedly afford, due to need (for example, a truck for work or an SUV to transport a large family) or simply the desire to have a nicer vehicle.

The flip-side of under-penetration in the lowest and highest household income ranges is then over-penetration in the middle income ranges: many more vehicles are sold into the middle part of the market than one might expect given the number of households in those income ranges.

What lessons, if any, does the market for cars and trucks have for the rental housing market?

At the top of the market truly expensive luxury apartments are as rare as exotic supercars, at least in Howard County. Just as wealthy families have the option of buying multiple lower-priced vehicles rather than one super-expensive vehicle, wealthy families in Howard have an alternative to renting expensive apartments, namely purchasing expensive single-family homes on large lots. For example, a household that could spend $5,000-6,000 or even more in monthly rent could just as easily purchase a house worth $1-2 million somewhere in west Howard.

(However, that’s not to say that there is no market at all for expensive apartments in Howard County: If listings at Apartments.com are any indication, there appears to be strong demand relative to supply for large 3-bedroom apartments at the Metropolitan development next to the Mall at Columbia. That demand will presumably be tested as more luxury apartment buildings open in downtown Columbia.)

The more interesting comparison in the context of affordable housing discussions is what happens at the lower ends of the vehicle and housing markets. There are at least three possible approaches to serve lower-income households looking for transportation services:

The first is for manufacturers to make and sell lower-priced new vehicles. For an example of what’s possible, a typical model of the manufactured-in-India Maruti Suzuki Alto hatchback sells for the equivalent of about $5,000 in that country, two to three times less than the Nissan Versa, apparently the lowest-price car currently available in the US market. Based on the affordability guideline above, such a vehicle would be affordable to a household with annual income of $10,000.

However, US and state government standards for vehicle safety, emissions, etc., make new vehicles more expensive than they otherwise might be. Government decisions also distort the new vehicle market in other ways: Relatively low taxes on gasoline in effect subsidize the sales of larger more expensive vehicles, while land use and highway construction policies that encourage sprawl also discourage the use of smaller under-powered vehicles like the Alto that are more suited to city driving.

The result is that lower-income households wanting a personal vehicle turn to the used vehicle market, where they can pay lower prices in exchange for accepting a vehicle with perhaps significant wear and tear. (As a comparison, a quick spot check of Howard County used vehicle dealers finds the lowest-price options are in the $6,000-$8,000 range, all vehicles with over 100,000 miles. A quick check of craigslist.org finds several vehicles in the $3,000-$6,000 range, typically with even higher mileage.)

Finally, lower-income households who can’t afford personal vehicles rely primarily on public transit, which in a Howard County context means buses.

The perceived housing equivalent of public transit is public housing for which renters are directly subsidized, for example, the relatively new Monarch Mills development in Columbia or the renovated Forest Ridge apartments (shown above, left), also in Columbia. Such housing is often stigmatized (e.g. as “the projects”) and in suburbs like Howard County new public housing construction is typically strenuously opposed by nearby homeowners.

The housing equivalent of used vehicles are older apartments (or, in some cases, single-family houses) that have deteriorated to the point where they can no longer command typical market rents—what in extreme cases is pejoratively called “slum housing.” However, unlike vehicles, which have no real value once they reach a certain point of decay, deteriorated apartment complexes have an intrinsic value due to land and location. In an economically vibrant area like Howard County they will likely be replaced with new construction or heavily renovated, at which point they will no longer be affordable to their former tenants.

Finally, the housing equivalent of low-priced new vehicles is low-priced new (or renovated) housing designed to be profitable for its developers with at most indirect subsidization (for example, housing for which the developer gets tax credits, analogous to the tax credits provided for electric vehicles). Profitability here depends on at least two factors:

First, developers may incur certain costs directly as a result of government laws and regulations. For example, requirements to go through an extended legal process for approvals (e.g., the 16-step process required for development in downtown Columbia) impose direct costs on developers as well as indirect costs, for example to fight legal action by development opponents or to gain their assent by paying for community amenities or reducing the scope of a project.

Second, government action (or inaction) distorts the market in various ways that make it more difficult to profitably offer lower-priced housing. For example, zoning regulations that allow only single-family homes, or government’s declining to offer water and sewer service in certain areas, reduce the ability of developers to offer lower-priced multi-unit housing in areas that otherwise might be a good fit for it.

So, what’s the take away from all this? First, the claims about Howard County needing 5,000 more affordable housing units need to be understood in the context of the assumptions behind this number. Change the assumptions and the number will change. (For example, it would be smaller if we relax the criteria for affordability.)

Second, the penetration rate analysis used in conjunction with calculating this number itself has a major problem: it implicitly assumes that all income ranges should have a 100% penetration rate for a given type of good, and that any deviation from that figure indicates a problem. But as we saw from the example of the market for clothing, such deviations might just be the result of people’s preferences, relatively undistorted by any other factors.

However, the fundamental point remains: It is possible for markets in some goods to do a better job in serving lower-income households than they do. The market for clothing does a better job in serving lower-income households than the market for vehicles, and the market for vehicles does a better job in serving lower-income household than the market for housing.

There’s therefore a lot of opportunity to improve how the housing market works for lower-income households. Since government is to a large degree responsible for the problems in the housing market, government is equally accountable for addressing those problems.

As discussed above, this need not involve directly subsidizing lower-income households or the housing developments that serve them. There’s also ample scope for reducing the market distortions caused by government action (or, in some case, inaction) in the areas of planning and zoning, provision of essential services, school redistricting, and the like. Anyone who claims to care about housing affordability in Howard County needs to look at all the possible options on the table to make it possible for lower-income income households to lead productive lives as our fellow residents of the county.

That’s all for now. My apologies for the long delay in getting this post out. In addition to my usual work and family duties, I also now want to spend more time revising posts from this newsletter for publication on my main Civility and Truth website, plus I have other unrelated projects I want to work on. The upshot is that I will be posting here much less frequently than my already infrequent schedule. In the meantime, thank you again for your interest and attention.

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

“What do you think we need to do to elevate Maryland?” That’s the last question that co-hosts Candace Dodson Reed and Tom Coale typically ask their guests on the Elevate Maryland podcast. I’m not important enough to be invited on the show, but I can tell you what my answer would be: To “elevate” Maryland will require “elevating” Baltimore, because Baltimore’s success (or failure) will ultimately drive Maryland’s success (or failure).

A few days ago I had to go into Washington, DC, for business. As I drove in through northeast DC, I was struck by the newly-constructed and under-construction apartment buildings, the new stores along H Street (including a Whole Foods Market), and the new streetcar line that begins at Union Station and runs through this section of the city.

These are the physical manifestations of a meteoric rise in DC’s economic fortunes, as shown in the above graph that I created for an earlier post. In less than 25 years DC has gone from having median household income well below the national median to having close to (and in the latest statistics, exceeding) the median household income of Maryland, itself the state with the highest median household income.

DC’s good fortune has in turn maintained and lifted the economic fortunes of the various jurisdictions of northern Virginia, with Loudoun County pulling away from Howard County in terms of median household income, Stafford County advancing into a tie with Howard County, and Fairfax County continuing to outpace Montgomery County. (See another earlier post for details.)

The difference between Washington, DC, and Baltimore is that DC, like New York City or Los Angeles, is an “indispensable city.” As the center of national political power it will continue to attract attention, migration, and (as a consequence) investment from the rest of the country. Baltimore, like Philadelphia or Pittsburgh (to give but two nearby examples), is in no way indispensable. It is a regional center whose economic and other status ultimately depends only on the region that it serves. Put bluntly, if Maryland doesn’t care about Baltimore then no one else will either.

I’d add that Howard County isn’t an “indispensable county.” If Baltimore declines in its fortunes and that decline drags down its suburbs economically and otherwise, there’s always Loudoun County for people and/or companies to move to, or Fairfax, or Stafford, or Prince William, or any number of places far enough away so that Baltimore is just a place you might occasionally hear about on the evening news.

If elevating Maryland requires elevating Baltimore, then the question becomes “What do you think we need to do to elevate Baltimore?” I make no claim to being an expert, but I’ll go back to my earlier argument that the primary focus should be on reducing crime and improving policing. To repeat what I wrote previously:

Reducing crime is fundamental because ensuring the security of its citizens is the first and foremost task of any government. Improving policing is fundamental because we expect our governments to go about the task of ensuring citizens’ security in a way that embodies the ideals of a liberal democracy, as opposed to resembling the practices of a police state, an occupation force, or a paramilitary group.

No one should have to live in an environment marked by constant criminal violence. In addition to its effects on individuals, it has a corrosive affect on the neighborhoods and cities in which it occurs. Efforts to build communities and promote their economic development cannot be expected to succeed in the absence of adequate security for their inhabitants.

Let’s go back to my earlier comparison between Baltimore and Washington, DC. Young people and people who’ve recently moved here may not realize it, but there was a time when DC was seen as a city overwhelmed by violent crime. As shown in the graph above, in the early 1990s DC had a murder rate about twice that of Baltimore and well over ten times that of the nation as a whole.

Since then the DC murder rate has dropped to a level comparable to that of New York, a city heralded for its reductions in crime, while Baltimore’s murder rate has (in years not shown in the graph above) grown to a level higher than that in the early 1990s. (According to FBI statistics, in 2018 Baltimore had 309 murders in a population of 605,436, for a murder rate of just over 51 per 100,000 people.)

The murder rate is often seen as a reliable marker of the amount of crime in a city, murder being the crime that is least likely to go unreported. By this measure Baltimore is seen as a lawless city, almost (but not quite) as lawless as DC was during its worst days.

To rephrase the “elevate” question once again, “What do you think we need to do to reduce crime in Baltimore?” As I wrote above, I’m not an expert in this area, so I don’t have any detailed recommendations. I think it’s fairly clear what won’t work though, namely trying to wage a literal “war on crime.”

I’d rather look to principles of policing that are almost two hundred years old, laid down when the first professional police force in England, the Metropolitan Police of London, was established under the supervision of Conservative Home Secretary Robert Peel. (More recently “Peel’s principles” were popularized by William Bratton, former police chief of New York and Los Angeles. Bratton also for a time served as a consultant to the Baltimore police force, but I don’t know what influence he had, if any.)

The very first principle is that the goal of a police force is “to prevent crime and disorder, as an alternative to their repression by military force and severity of legal punishment [emphasis added].” In other words, there’s no “war” called for here.

The very next principle is that the police should “recognise always that the power of the police to fulfil their functions and duties is dependent on public approval of their existence, actions and behaviour, and on their ability to secure and maintain public respect.” That gets into the topic of this section, and a focus of my earlier post:

How can police forces (and, by extension, those who oversee them), reduce the frequency with which police officer kill unarmed members of the general public, especially unarmed African Americans, and especially unarmed African American men? Just as the murder rate is a key indicator for the crime rate in general, the frequency of such killings serves as a indicator of general levels of police violence, brutality, and corruption.

Why is this reducing the number of people killed by police important? There are some people who point out that the chances of an unarmed African American man being killed by police are extremely low, and far less than being murdered by someone else. Based on my analysis, this claim is correct. But in my opinion that’s ultimately beside the point. The chances of an American being killed by radical Islamic terrorists are also extremely low (lower, in fact), yet since 9/11 the US has spent billions if not trillions of dollars in an effort to combat this threat.

The fact is that even a relatively low rate of police killing unarmed African Americans results in an almost weekly stream of such stories, and that in turn drives their perception of police and their willingness to cooperate in the attempts by police to reduce crime. To quote James Smith, the neighbor of Atatiana Jefferson whose concern for her safety resulted in a police officer shooting her in her home: “They tell you if you see something, say something . . . but if it causes someone to lose their life, it makes you not want to do that.”

Or as the third principle of policing puts it, the police should “recognise always that to secure and maintain the respect and approval of the public means also the securing of the willing co-operation of the public in the task of securing observance of laws.” The number of police officers is very small in comparison to the population at large. If the people affected by crime fear cooperating with police then “securing observance of laws” becomes a difficult if not impossible task.

So now we come to the next question, “What do you think we need to do to reduce killings of unarmed people, especially African American people, by police?” Again, being no expert I will defer to others to make their case. One of the more interesting ones I found was in a Twitter thread by Samuel Sinyangwe, one of the Black Lives Matter activists who founded Campaign Zero. Both the goal and the approach of Campaign Zero are summarized on their website: “We can live in a world where the police don’t kill people by limiting police interventions, improving community interactions, and ensuring accountability [emphasis in the original].”

What I found interesting about Sinyangwe’s thread is that he is apparently willing to go where research leads him, including against some common notions about what will work. For example, although “body cams” are number 6 on the list of items highlighted on the Campaign Zero website (see the above graphic), Sinyangwe begins his thread by claiming that “the most discussed ‘solutions’ to police violence have no evidence of effectiveness. For example, Body cams don’t reduce police violence.”

This makes sense to me. For one thing, police can always turn them off. For another, many if not most situations in which police use deadly force are going to be difficult to judge in an unambiguous manner, and even after watching body cam videos prosecutors, judges, juries, and the general public alike are likely to give police the benefit of the doubt even in situations when they arguably do not deserve it.

Sinyangwe continues to debunk by-now-conventional wisdom in his second point: “There is no evidence that better police training programs or ‘implicit bias’ training changes police behavior.” Note that this contradicts item 7 on the nominal Campaign Zero list. In particular, the claimed ineffectiveness of implicit bias training, and by extension the claimed ineffectiveness of the “implicit association test” or IAT popular among employers and in other contexts, echoes a point I’ve seen made by others in a more general context: what really matters is not what people think and believe, it’s what the incentives are that drive their behavior.

Or to put it another way, if one takes the concept of “structural racism” seriously, then the emphasis should be on the “structural” rather than the “racism,” at least in the sense of racist attitudes that might be held by individual people, or in this case by individual police officers. We need to look at the broader set of conditions under which police operate, and change those conditions in a way that will promote changes in behavior.

The name “Campaign Zero” implies that the goal is to totally eliminate police killings of unarmed people. It’s certainly a laudable goal. Is it a realistic one? Or should we simply accept the current rate of killings as an inevitable accompaniment to police doing their job?

Rather than look to the policing situation in other countries, I thought it would be interesting to compare this to another challenge, namely reducing as much as possible commercial airplane crashes and the attendant loss of life. At one time flying commercial airlines was relatively dangerous—likely not as dangerous as driving a car (I haven’t looked up the statistics on this), but certainly air crashes were relatively frequently in the news.

However, as can be seen in the graph above from the Aviation Safety Network, the last forty years have seen a dramatic reduction in the rate of airline crashes, by about an order of magnitude. (The graph expresses this in terms of accidents per million flights. Since airplane crashes tend to happen on takeoff or landing, with accidents during the rest of the flight very uncommon, this is probably a better indicator than crashes per passenger-miles.)

This improvement was the result of a concerted and continuing focus on determining why air crashes occurred and then seeking to eliminate as much as possible the underlying root causes, whatever they might be. The campaign was carried out under the auspices of government agencies (including FAA, NASA, and the NTSB) with the cooperation of airlines, airplane manufacturers, and pilots and the pilots’ union.

We can imagine a similar officially-sponsored and supported campaign focused on driving to zero the number of police killings. However the two situations are different in the ways incentives are aligned (or not aligned, as the case may be).

For example, air crashes typically result in the deaths of both pilots and passengers, so pilots have just as much incentive to avoid crashes as passengers. However in police killings there is a conflict between officers’ desires to protect themselves from harm and the interests of those whom they encounter in preserving their own lives. Incenting officers to be less ready to resort to deadly force will be easier if other measures are seen to make policing less risky.

Airline pilots and police officers are also both unionized, and like other unions pilot and police unions are incented to put the private interests of their own members before the interests of others, including the general public. In the case of airline crashes the pilots’ union was and is concerned that pilots not be unfairly blamed for crashes in which pilot error, even if present, was but one factor among many in causing a crash. Similarly, police unions may be more or less resistant to measures to reduce police killings depending on the extent to which those measure focus on the immediate behavior of officers as opposed to larger organizational issues.

Finally, and perhaps most important, social, economic, and political elites had a major interest in improving air safety, since especially prior to cheap airfares they were much more likely to travel by air than typical people. However, from the perspectives of elites the killing by police of unarmed people is a problem that happens to someone else, and rarely if ever to them, and thus they have no direct interest in addressing it. Crime levels in general can affect them, but again they have ways of removing themselves from the problem, ways that are not available to ordinary people. Even political elites may not have a direct interest in solving the issue, if their electoral success doesn’t depend on addressing it.

So, to recap: I think the fate of Maryland is tied to the fate of Baltimore, the fate of Baltimore is tied to the success (or lack thereof) in reducing crime to levels comparable to those of cities like DC and New York, reducing crime in Baltimore will depend on policing that promotes cooperation in that portion of the public that has the most interactions with police, and that in turn requires addressing the problem of police killings and police violence in general.

While activists can promote new visions of how policing can be improved, as the Campaign Zero activists have done, in the end only elites, and in particular political elites—including those who populate the guest lists for the Elevate Maryland podcasts—are in a position to do what needs to be done to create an effective and responsible police force in Baltimore. Time will tell whether they are up to the task.

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

I’m back again with more population density maps, this time for Ellicott City—or, more precisely, the Ellicott City Census Designated Place or CDP.

The map above shows population density variations as of the 2010 census for the various census block groups that are wholly or mostly within the Ellicott City CDP. There are 34 such census block groups, compared to 54 census block groups for Columbia or 154 census block groups within the county, with Ellicott City thus accounting for about a fifth of Howard County. (The 2010 population of 64,245 for these census block groups is also almost a fifth of the county’s total 2010 population.)

The main take-away from the map above is that most of Ellicott City has pretty much the same population density: it has less multi-unit housing than Columbia, and fewer large lots with single-family homes than western Howard County. (The major exception is the area east of US 29 between US 40 and I-70, which contains a number of apartment complexes.)

This is confirmed by the histogram above, which shows the distribution of density among the various Ellicott City census block groups in 2010. Most of the CDP was between 1,000 to 4,000 people per square mile, with only two census block groups out of 34 having higher density. The overall population density for Ellicott City in 2010 was 2,234 people per square mile, about a third less than the overall population density for Columbia.

The next map shows changes in population density in Ellicott City between the 2010 census and the 2017 American Community Survey 5-year estimates (which reflect surveys done in 2013 through 2017). Even over this relatively short period we can see a significant decrease in density in the area between US 40 and I-70 west of Rogers Avenue (which includes a good chunk of Patapsco State Park).

There was also almost a doubling of population density in the area north of old Ellicott City and south of US 40. I don’t know if the latter is due to new development or to changes in household size. However it’s worth noting that based on the 2010 population density map compared to the map for 2013-2017 immediately above, that area was originally less densely populated than most of Ellicott City, and whatever changes occurred brought it up to a “typical” density for the area.

As I did for Columbia, I also mapped density variations at the level of census blocks. This uses population data available for the 2010 census but not for the American Community Survey.

This map is most notable for showing areas of Ellicott City that have very low population densities. These include retail areas like the Long Gate shopping center north of MD 100 and east of US 29, park areas including Patapsco Valley State Park, Meadowbrook Park, and Centennial Park, and golf courses like the one at Turf Valley.

Note that as with Columbia some areas appear on this map that were not on the prior maps. These reflect census blocks that are in the Ellicott City CDP, but that are in census block groups that are mostly not in Ellicott City. The most notable example of this is the Turf Valley resort.

There is also a set of Ellicott City census blocks west of Centennial Lane that appear to be almost disconnected from the rest of the CDP. These appear to be associated with the soccer fields and church at Covenant Park, with Centennial High School and Burleigh Manor Middle School another “low-density” area just north of there. (Recall that “low-density” in this context refers to the size of the residential population in a given area, not how built-up the area is.)

You can find the code and data behind this post, as well as more Ellicott City population statistics, in the document “Ellicott City, Maryland, population density.”

That’s all for this week. A reminder: if you find these posts interesting and useful please tell other people about them and encourage them to subscribe to the Civility and Truth mailing list. Having readers who care enough to subscribe helps motivate me to send these posts out on a regular basis, and the more readers I have the more motivated I’ll be. In the meantime, thanks for reading this post!

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

Hey! I finally managed to figure out to get a list of census block groups or census blocks for the Columbia CDP. (As a reminder, “CDP” or “Census Designated Place” is US Census Bureau jargon for a population center that’s unincorporated.) So now I can bring you some density maps that zoom in to focus on just Columbia as opposed to all of Howard County.

The map above shows population density variations as of the 2010 census for the various census block groups that are wholly or mostly within the Columbia CDP. (Some census block groups contain only a small portion of the Columbia CDP. I omitted them from the map.) There are 54 such census block groups, as compared to 154 census block groups within the county. Thus from this point of view Columbia accounts for about a third of Howard County.

As I mentioned previously, census block groups are a nice “not too large, not too small” subdivision. In 2010 the least populated census block group in Columbia contained 645 people, while the most populated census block group contained 3,632 people. The smallest Columbia census block group covered an area of 0.12 square miles (about 79 acres), while the largest block group covered an area of 3.11 square miles.

A typical census block group is considerably smaller than a Columbia village: since there are nine Columbia villages, each village would contain about 6 census block groups on average if they were equally distributed.

In the map above I’ve shown more roads than in my previous maps, to help orient you vis-a-vis the various parts of Columbia. I thought about also superimposing the boundaries for the various Columbia villages (data that’s available on the Howard County GIS site), but ran out of time to make this work.

The main take-away from the map above is the areas of Columbia that are relatively high-density vs. relatively low-density. Relatively high-density areas include portions of Harpers Choice and Owen Brown villages, presumably due to apartment complexes there. In contrast the Columbia Gateway area is primarily office space and thus low-density in terms of residential population.

The histogram shows the distribution of density among the various Columbia census block groups in 2010. Density varied from a low of 761 people per square mile to a high of 13,285 people per square mile, a difference of over an order of magnitude. Overall population density for Columbia in 2010 was 3,187 people per square mile.

As a comparison, the lowest density Howard County census block group in 2010 had 151 people per square mile, and the highest density block group had 15,181 people per square mile. Overall population density for the county in 2010 was 1,144 people per square mile. Thus the least-dense Columbia census block group was over five times as dense as the lowest-density Howard County block group, and Columbia as a whole was about three times more densely populated than the county as a whole.

The next map shows changes in population density in Columbia between the 2010 census and the 2017 American Community Survey 5-year estimates (which reflect surveys done in 2013 through 2017). Even over this relatively short period we can see significant decreases in density in the area of Harpers Choice village off Harpers Farm Road in northwest Columbia, and a increase in density in the area just south of there, north of Little Patuxent Parkway and east of Cedar Lane. I don’t know if there’s been any change in the total number of housing units in those areas, so my initial guess is that these changes are due to changes in household size: older children moving out on the one hand, and more young children on the other.

For my final map I decided to map density variations at the level of census blocks. This uses population data available for the 2010 census but not for the American Community Survey.

Census blocks are very small: there are 1605 census blocks in the Columbia CDP, with an average area of 0.02 square miles (about 13 acres). In 2010 over half of the census blocks contained no people at all, and the average population of a block was only 62 people; the largest block contained 2,696 people.

Because of the small size of census blocks the corresponding population density can be very high if the block mainly contains apartment complexes. In 2010 there were several census blocks in Columbia with population densities over 50,000 people per square mile, and a few over 100,000 people per square mile. At the other end of the spectrum the majority of census blocks in Columbia contain no people and thus have a population density of zero.

Because of this wide distribution of population densities I don’t think the block-level map is that useful for looking at population densities in residential areas. However on this map it’s very easy to pick out the parts of Columbia that are devoted to office, retail, or industrial uses (in dark blue). This includes the Columbia Gateway and Dobbin Road areas on either side of Snowden River Parkway, the areas along Broken Land Parkway, the Mall at Columbia and the future Merriweather District, and the area around the hospital and Howard Community College.

UPDATE: I was able to resolve the problems that were preventing me from publishing the underlying code and data for this post. See the document “Columbia, Maryland, population density.”

That’s all for this week. A reminder: if you find these posts interesting and useful please tell other people about them and encourage them to subscribe to the Civility and Truth mailing list. Having readers who care enough to subscribe helps motivate me to send these posts out on a regular basis, and the more readers I have the more motivated I’ll be. In the meantime, thanks for reading this post!

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

Should we in Howard County be encouraging more people to move here, especially people with lower incomes who hail from Baltimore city and elsewhere? Should we instead be focusing instead on the people left behind in areas with high rates of poverty, helping them where they live now? And the most basic question: Who is Howard County really for?

Recently I attended, read, and watched three things bearing on those questions: the initial meeting of the local Howard County for All advocacy group, the book The Lines Between Us: Two Families and a Quest to Cross Baltimore’s Racial Divide by local journalist Lawrence Lanahan, and an event on “Moving to opportunity: What’s next?” put on by the Brookings Institution, a DC think tank.

Howard County for All was recently founded and/or promoted by a number of local activists, including former local blogger T.J. Mayotte and Elevate Maryland co-hosts Candace Dodson Reed (who moderated the September 12 launch event) and Tom Coale. Its statement of purpose claims that “Howard County should be open to all,” that the county is “strengthened through inclusivity,” and in support of those values should “create and maintain a variety of housing options alongside a commitment to long-term affordability.”

In The Lines Between Us, Lanahan (who spoke at the Howard County for All launch event) weaves together the stories of three people: Mark Lange, who with his wife Betty was called by his Christian faith to move from their Harford County suburb to live in Baltimore’s Sandtown neighborhood; Nicole Smith, who moved herself and her young son from Baltimore City to Columbia with the help of a special housing voucher; and Barbara Samuels, a lawyer and activist who participated in the legal battles to open up affordable housing across the entire Baltimore metro area.

Finally, the Brookings event, broadcast over the Internet, featured presentations and discussions of research on the benefits to families moving from high-poverty areas to low-poverty areas, and whether and how government policies should promote this.

Since my writing tends to be data-focused I’ll spend most of my time discussing the Brookings event, and in particular the presentation on the Moving to Opportunity initiative, with occasional asides regarding the human impacts documented in Lanahan’s book and the ongoing controversies involving affordable housing and school redistricting in Howard County. Since the Brookings event, Lanahan’s book, and the Howard County for All group all are coming at these issues from a progressive/liberal point of view, I’ll also try to acknowledge conservative and libertarian perspectives as well.

As noted in the slide above (taken from the Brookings webcast, as are the other slides below), Moving to Opportunity was a Federal initiative to determine what would be the benefits (if any) of providing government assistance to help families with children move from high-poverty neighborhoods in central cities to low-poverty areas in the surrounding suburbs.

Stated this way the MTO project could be portrayed as a classic attempt at social engineering undertaken by central planners blind to social and economic realities. However, it could also be seen as an attempt to overcome the ill effects of previous social engineering by central planners with different priorities. Lawrence Lanahan devotes much space in his book to documenting government policies that had the effect of advantaging (mostly white) suburban dwellers at the expense of (mostly black) city dwellers, including radically changing the face of cities by destroying traditional neighborhoods in the name of “urban renewal.”

Some of these activities are in the past, but others are live issues today, most notably with respect to zoning. In effect zoning laws employ government coercion to restrict the freedom of property owners to make what might be more productive uses of their land. For example, a landowner might be allowed to build a large single-family home, but be prohibited from building a duplex, an accessory apartment, or a small apartment complex, even though these alternatives might be more lucrative for them and better address market demand.

In the past many Maryland counties, including Baltimore County in particular, went even further and specifically denied approvals for any attempts to build “affordable housing” projects, even when they otherwise conformed to zoning requirements. They were supported in this by politicians and homeowners who agitated against poor inner-city residents moving to the suburbs. As but one example, when 285 families from Baltimore public housing were chosen to receive special housing vouchers as part of the Moving to Opportunity initiative, opponents portrayed it as the first step in moving thousands of poor Baltimore city residents into eastern Baltimore County, displacing existing residents.

Given the role of past government policy in isolating poor people and their housing from the rest of society, it’s understandable that some government officials would look for a way to try to remedy the effects of those policies by providing lower-income families opportunities to move to more mixed-income environments. But leaving aside the political controversies, did this actually result in any benefits for such families?

The Moving to Opportunity initiative came in the wake of other small-scale experiments that suggested that there were major benefits to lower-income families moving to the suburbs. The results from the MTO initiative itself were more measured, and the initial presentation at the Brookings event was essentially a recap of why the MTO results were worse than those from prior initiatives, and what the best case results might be.

Leading with something positive, there were apparently some benefits related to the physical and mental health of those lower-income adults who moved with their families to the suburbs, including in particular lower incidences of depression and diabetes. Even though these benefits appear to be somewhat modest, they could both improve quality of life for the people involved and help lessen government health care spending on these chronic diseases.

Exactly why did the adult MTO participants have health improvements? The speaker at the Brookings event was silent on that point, but I’ll speculate that it may have been associated with reduced stress. (More on that point below.)

However, even though moving to the suburbs may have improved the health of lower-income adults, it ultimately didn’t make any real difference to their economic prospects. The graph above tells the tale: the percentage of lower-income adults employed ended up being the same no matter whether they moved to the suburbs as part of the MTO initiative (“experimental group mean”), moved elsewhere in the city through the normal Section 8 housing voucher program (“Section 8 group mean”), or stayed where they were (“control group mean”).

Even for those employed it’s not clear that moving to the suburbs led to better-paying jobs. For example, as described in The Lines Between Us, upon moving to Columbia Nicole Smith was able to complete an associate degree at Howard Community College and work as a child care provider. While she was able to support herself and her son, in the end she did the same type of work her mother did, just in a different setting.

What’s the explanation for the failure of the MTO initiative to change these life outcomes? My own personal theory goes somewhat as follows: We can view people as having both natural talents and certain personality attributes (e.g., being conscientious or extroverted) that suit them for certain types of work. These talents and these personality attributes are in turn partly inborn and partly developed through people’s unique life experiences in their early years.

For example, the child who is naturally good with their hands can develop related skills (for example, as an carpenter or auto mechanic) if given the opportunity to do so. A child who is naturally extroverted may find that they can stand out from their peers by emphasizing their ease and skill in public settings, and perhaps end up in sales or politics. And so on.

But once they have become adults people have mostly developed whatever natural talents they possess, and their personalities are pretty much fully-formed. Based on their talents and personalities there are some jobs for which they are well-suited, and some jobs in which they would likely never succeed. (For example, though I’ve worked in sales environments for most of my career I know that my personality is ill-suited for life as a salesperson.) Where one might move might affect this somewhat at the margins, but is unlikely to make a major difference.

A conservative could here point to Thomas Sowell’s “constrained vision”: that human nature puts limits on what society and especially government can accomplish in alleviating human suffering and promoting human flourishing. Of course one could also go on to point out that everyone is equal in the sight of God, and that a well-ordered society can and should provide opportunities for every person to live a life of dignity and worth within the scope of whatever talents and other attributes they possess.

For us to do otherwise, to seek to close ourselves off from those less fortunate by birth and circumstance, and to countenance their isolation out of sight and mind, would be like attending a church whose doors are open only to the conspicuously righteous. According to The Lines Between Us it was the struggle to reconcile their comfortable suburban lives with their felt duties as followers of Christ that led Mark and Betty Lange to leave the suburbs for life in Baltimore city. For those of us who choose to stay in the suburbs it may also serve as motivation to welcome those from the city who wish to move here, and to assist them in doing so.

It’s also worth noting here what the MTO results do and not mean: They imply that if a random person from a low-income neighborhood moved to a higher-income neighborhood it would make little or no difference to their employment prospects. But in real life the people who move from lower- to higher-income neighborhoods do not do so at random. Given the financial and other barriers to moving, the lower-income people who do move are those who are most motivated to make the move, and most motivated to make for themselves a successful life in their new neighborhoods.

(Incidentally, this is exactly why small-scale programs prior to MTO showed such good results, because they had a “selection bias” of this type. The random assignments in the MTO program removed the effort of this bias.)

The discussion above was with reference to lower-income adults moving to the suburbs. What about their children? Here the picture is more complicated: all things equal, those children who moved to the suburbs as part of the MTO initiative as young children had increased incomes later in life compared to those who stayed where they were, while those who moved as teenagers did not. (The differences in income shown in the graph may seem small, but the cumulative differences in lifetime earnings could be quite large.)

What could account for these different outcomes? To go back to my hypothesis above, I presume that natural talents in children can be enhanced by opportunities they have to discover what they like doing and get better at it, and that personality development and socialization in children are strongly influenced by the environment outside the home, including both their peers at school and play and the overall influence of neighborhood adults. (See for example my post on the ideas of Judith Rich Harris.)

Children who moved as youngsters would then have more opportunities to develop their inborn talents and a longer period of socialization to prevailing norms. They would be analogous to the young children of immigrants, who may have begun life speaking a different language but then learn to speak a new one fluently. On the other hand, children who moved as teenagers would already be well on the way to full development of their talents and personalities, and the new environment would likely have less effect on them. They would be like immigrants who arrive as older teenagers or young adults, who may learn a new language but still speak with an accent.

The speaker at the Brookings event then went on, as speakers at such events typically do, to discuss implications for policy, in this case various options for improving the effectiveness of programs that encourage lower-income families to move to higher-income neighborhoods. The most obvious of these is to focus on families with young children, for reasons discussed above. The others you can read on the slide, but I’ll recap them briefly: provide additional counseling, allow for higher rents, and work with tenants, landlords, and housing authorities to improve the experiences of these families and make it more beneficial to rent to them.

(You can find more information about these recommendations in the article “Neighborhoods matter for opportunity: Time for more place-conscious policy” on the Brookings web site, the “Collinson and Ludwig, forthcoming” article referenced on the slide.)

Of course, this assumes that moving families is the best way to approach this problem. The alternative approach, again as outlined by the speaker, is to try to improve high-poverty neighborhoods directly. Here the key insight was that families participating in the MTO experiment wanted to move in large part because they felt unsafe living where they were, and wanted to escape those neighborhoods for those in which they could be more secure.

This harks back to my discussion in an earlier post about the importance of reducing crime in the context of promoting racial equality, and to do so in a way that (as I wrote later) “both protects and respects” the people living in these neighborhoods.

Let’s bring this back to Howard County. The county has replicated in miniature the income and housing patterns of the larger Baltimore metro area, with a relatively low-income high-density core almost encircled by relatively high-income low-density areas. Here Columbia plays the role of Baltimore city and western Howard the role of Baltimore County and other suburban jurisdictions. What, if anything, should we try to do about the resulting patterns of segregation?

This question is at the heart of the ongoing controversies in Howard County over development, affordable housing, and school redistricting. I can’t offer a definitive answer to the question, but I see at least three possible paths for the county:

The first is simply to maintain the status quo: Let existing patterns of segregation remain roughly the same, but at the same time make some moves toward maintaining affordability of housing in select areas of the county, typically in and around Columbia.

The second takes us down a path similar to that traveled by Silicon Valley and other high-tech meccas we often wish to emulate: In this scenario the problem of school overcrowding would be addressed by limiting development as much as possible (with possible exceptions for senior living projects). With development brought to a near halt, prices for existing housing would soar as demand for Howard County’s “good schools” and quality of life remained high.

Political pressure would then rise for existing homeowners to be protected against tax increases, and for the burden of paying for new schools and other new county services to be put on new residents as much as possible. The result would be to price out low-income or even moderate-income families. The adults in such families would still be needed to fill a wide range of jobs in Howard County, but they would have to commute in from other places and then leave the county at the end of the day.

The third scenario would see more aggressive efforts to reduce residential segregation and make Howard County more affordable to anyone wishing to live here. This might include school redistricting along the lines of the proposed plan, an expanded program of school construction to get ahead of demand for housing, and placing affordable housing projects in areas outside of Columbia. It might also include encouraging more development in selected areas in west Howard, for example along commuter routes like I-70 and MD 32, and extending county services into those areas as necessary to support higher-density housing.

Which path we take depends on the values we each think are most important, and the political power we can bring to bear in coalition with others. My own inclination is for us to pursue the path of making Howard County a county that as much as possible is open to all who aspire to live here, not one populated by a monoculture of highly-educated high earners singularly obsessed with home values and test scores. That reflects the values I hold, and is why I support the efforts of groups like Howard County for All.

As always, if you find these posts interesting and useful please tell other people about them and encourage them to subscribe to the Civility and Truth mailing list. Having readers who care enough to subscribe helps motivate me to send these posts out on a regular basis, and the more readers I have the more motivated I’ll be. In the meantime, thanks for reading this post!

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

This is a brief follow-up to last week’s post. I had hoped to be able to take a closer look at density variations in Columbia and Ellicott City. However I haven’t yet found a good way (at least via the Census API) to get a list of census block groups or census blocks for the Columbia and Ellicott City CDPs. (“CDP” or “Census Designated Place” is US Census Bureau jargon for a population center that’s unincorporated.)

I also happened to think about whether the census block group boundaries had changed from 2010 to 2017. After looking at this I concluded that they probably had not, but this took a while to nail down.

The one new thing I did was to produce a map of density variations based on the quintiles the various census block groups fell into. In other words, in the map above the census block groups in dark blue (quintile 1) are the 20% of all census block groups with the lowest population density in 2010, while the census block groups in yellow (quintile 5) are the 20% of all census block groups with the highest population density in 2010. The other areas in light blue, green, and orange (quintiles 2, 3, and 4 respectively) are intermediate between those two groups.

I think this map does a slightly better job of letting you tell at a glance which are the most dense and least dense parts of the county, as well as which areas are roughly in the middle in terms of density.

Here’s a similar map using the 2017 ACS 5-year estimates, which cover the 2013-2017 timeframe. At first glance I can’t see any differences between this map and the prior map. This means that even though some areas of the county may have experienced changes in population density between 2010 and 2013-2017, the changes weren’t large enough to make any real difference in the overall density picture.

I’m going to try again to do density maps for Columbia and Ellicott City. In the meantime, see the revised version of “Howard County density trends by census block groups” for the code behind the maps above.

That’s all for this week. A reminder: if you find these posts interesting and useful please tell other people about them and encourage them to subscribe to the Civility and Truth mailing list. Having readers who care enough to subscribe helps motivate me to send these posts out on a regular basis, and the more readers I have the more motivated I’ll be. In the meantime, thanks for reading this post!

tl;dr: Some areas of Howard County are over a hundred times more densely populated than others.

A long-time theme in writings about Howard County is the distinction between the more densely populated suburban and semi-urban areas like Columbia and the less densely populated rural areas in the western part of the county. This has implications for issues from political affiliations to school redistricting, and of course for affordable housing as well.

In this post I’m going to ignore those issues though, and just look at the simple facts about density variation across the county. The map above shows density variations as of the 2010 census—a data source I chose because it contains accurate population counts at a fairly fine-grained level.¹ The map shows population densities (people per square mile) for each of the census block groups within the county.

A census block group is a geography defined by the US Census Bureau that is one level below a census tract. There are currently 154 census block groups in Howard County, compared to 55 census tracts. The smallest geography in US Census data is the census block, one level below the census block group. There are currently 4,845 census blocks defined for Howard County.

Census tracts are relatively large, about 6,000 people or so on average in Howard County. On the other hand, census blocks are too small: more than half of all census blocks in Howard County contained no people at all in the 2010 census.

Census block groups are a nice “not too large, not too small” subdivision of the county’s overall area. In 2010 the least populated census block group contained 645 people, while the most populated census block group contained 3,632 people. The smallest block group covered an area of 0.1 square miles (about 64 acres), while the largest block group covered an area of 13.4 square miles.

A typical census block group is thus comparable in both size and population to a traditional village or small town. It is large enough to be a recognizable “place,” but small enough to have its own identity distinct from that of other places in the county.

What about density? One of the most surprising things to me in doing this analysis was the wide variation in population density across the county. Population density in 2010 varied from a low of 151 people per square mile to a high of 15,181 people per square mile, a difference of two orders of magnitude. In comparison, overall population density for the county in 2010 was 1,144 people per square mile (287,085 people divided by the county land area of 251 square miles).

The map above is a variation on the first map. It is based on the same 2010 census data, but divides the census block groups into five groups (or “quintiles”) of 31 block groups each (or 30, for the highest quintile). This map shows much more clearly that almost all of the census block groups with the highest population density are in Columbia and eastern Howard County, and almost all of the census block groups with the lowest population density are in western Howard County.

(Some of the major exceptions are areas like Columbia Gateway and the light industrial districts east of I-95 that have little or no residential construction. Here “low population density” is not the same as “not built up.”)

As shown in the above histogram, and also in the first map above, there are many census block groups in Howard County that had a population density of less than 500 people per square mile in 2010. At the other end of the spectrum a few census block groups had population densities of well over 10,000 people per square mile.

The largest number of census block groups fell into the range of 2,000–2,500 people per square mile; the typical (median) block group had a population density of about 2,400 people per square mile. (This is different from the overall population density of Howard County of 1,441 people per square mile quoted above.)

To help think about what these numbers mean, consider a square mile, about 640 acres. Suppose we have a few neighborhoods of single-family houses on 3- to 4-acre lots, with three or four people per house. With 50–100 total houses we have a total of 150–400 acres and 150–400 people. Throw in two or three 100–150 acre farms plus road surfaces and open spaces and you’d have a typical Howard County rural census block group with a population density of 200–500 people per square mile.

Now suppose instead we have 0.1 square miles, about 64 acres, occupied by five or six apartment buildings with 50–100 units each, with two to three people per unit. Now we have 500–1,800 people total in that 0.1 square mile area, for a total of 5,000 to 18,000 people per square mile—in other words, a typical semi-urban setting in Columbia or eastern Howard County.

How is population density in the various parts of Howard County changing over time? It’s hard to get a good picture of this in between censuses, because the available population estimates are from surveys taken over multiple years (from 2013 through 2017 for the latest available data) and have fairly high margins of error at the level of census block groups (up to 30% or more above or below the estimates themselves).

The map above is an attempt to show density changes from the 2010 census forward, using the American Community Survey 2017 5-year estimates. As with population density itself, there is wide variation in population density changes.

A few areas stand out as having significant increases in population density, from 50 to 100%. These appear to be include the Maple Lawn Farms development in Fulton as well as adjacent neighborhoods south of MD 216, areas along US 1 south and north of MD 175, and areas near downtown Ellicott City.

Other areas apparently experienced decreases in population density. Assuming that the number of housing units did not decrease in those areas, this likely was caused by the number of people per household decreasing, for example due to children leaving families and “empty nesters” remaining. (Additional census data, for example on household size and the ages of household members, should be able to confirm or refute this idea.)

To sum up: we may argue about how the density divide in Howard County came about and what it all means, but I don’t think there’s any dispute that it exists. It is especially clear in western Columbia, where the drop off in density west and north of MD 108 and (to a lesser extent) south of MD 32 is particularly dramatic. It’s also apparent that new development and demographic changes in family size are having disparate impacts across the county.

Combining this density data with information on socioeconomic and political variables could uncover some interesting patterns. Hopefully I’ll have time in the future to look at this.

Further exploration

For more on how I created the maps and histogram above, see the following:

• “Howard County density trends by census block groups” shows the R code used to produce these and other graphs.
• My hocodata code repository includes copies of the R Markdown files for this and another analyses. (Look in the “affordability” subdirectory.)
• If you sign up for a free account on the Rstudio.cloud service you can open and make a copy of my hocodata project for this and other analyses, and try your hand at it yourself. (Again, look in the “affordability” subdirectory, and check out the RStudio primers to learn how to use the system.)

I also did two other articles focusing specifically on population density in Columbia and Ellicott City for my (now deprecated) Civility and Truth newsletter.

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

I’m back this week with some more data analysis and maps, this time regarding the relative cost of housing in Howard County for both renters and homeowners. But before we get to the maps, a brief aside on some ways in which the impact of housing costs can be measured:

The first way is to look at the median percentage of household income taken up by housing costs. That is, for each household, first compute the percentage of its income taken up by housing costs. Then compute the median of all such values, such that half of all households spend more than that percentage and half less.

An alternative measure of housing costs is the percentage of households whose housing costs exceed a certain amount, typically taken to be 30% of household income. Such households are said to be “housing cost burdened.” (Households spending more than 50% of their income on housing are often referred to as “severely burdened.”)

Finally, others have suggested looking at residual income, that is, how much income households have left over after paying for housing.

For this analysis I use the median percentage of housing costs relative to income, mainly because these estimates are directly available at the census tract level in the American Community Survey 5-year estimates. (I may do an analysis based on housing cost burden in my next post, and am looking further at how one might do at least a simple form of residual income analysis.)

Anyway, back to the map above: It shows for each census tract in Howard County the median percentage of household income spent on “gross rent”—rent plus utilities and fuel—in the 2013–2017 timeframe. (There are lots of gray areas on this map and one of the maps below because for some census tracts there were too few renters or homeowners to calculate an estimate, or the estimates were too unreliable.)

It looks if there are a number of places in Howard County where people spend fairly high percentages of their household income on rent and related expenses. Most of these appear to be in Columbia and eastern Howard County, but there’s also an area in western Howard County where the typical renter looks to be housing cost burdened.

The histogram above shows a different view of the same data. It shows the number of Howard County census tracts falling into different categories: median percentage of rent relative to income between 15% and 20%, between 20% and 25%, and so on. (Again I’ve omitted census tracts with no estimates or unreliable estimates.) The dashed vertical line shows the median percentage of rent to income for all of Howard County (28%), and the solid vertical line shows the 30% figure above which a household is considered to be housing cost burdened.

As noted above, there are a lot of census tracts in Howard County where the median household spends more that 30% of household income on rent and related expanses.

The next map has the same color scale, but shows the median selected costs of housing as a percentage of household income for homeowners who have a mortgage. (“Selected costs of housing” includes not only the mortgage payments but also utilities, insurances, taxes, and condo fees—which would appear to include Columbia Association annual charges.)

Based on this map it appears that typical homeowners in Howard County with mortgages generally spend relatively less on housing than renters.

The histogram above confirms this initial impression: In all Howard County tracts with valid estimates median homeowners with mortgages pay less than 30% of household income in housing costs. The median percentage of housing costs to income for the county as a whole was 22% for homeowners with mortgages in the 2013-2017 timeframe, significantly lower than the 28% figure for renters.

The data as presented don’t really explain this difference: it could be due to relatively higher rents and related expenses for renters, or relatively lower household incomes for renters, or both. (And relatively lower household incomes in turn could be due to lower incomes in general, or fewer numbers of workers per household.) This is a possible subject for future posts.

The final map shows the median selected costs of housing as a percentage of household income for homeowners who do not have a mortgage. These households presumably consist of older people who’ve paid off their mortgages, people wealthy enough to pay cash for their homes, or people who’ve inherited homes from their parents. There are relatively few of these people, so most census tracts do not have reliable estimates for their housing costs.

These households appear to be in even better shape than homeowners with mortgages: their median percentage of income spent on housing is fairly low across the entire county.

The final histogram confirms this: In almost all census tracts the median percentage of household income spent on housing by homeowners without mortgages is less than 15%. The median percentage for the county overall is 10%.

So, what does this all mean?

People in all three categories receive government subsidies of one sort or another: renter can be eligible for housing choice vouchers (also known as Section 8 vouchers), homeowners can deduct mortgage interest and real estate taxes (although real estate taxes are no longer deductible on Federal tax returns), and older homeowners can get additional help with housing costs (for example, the Howard County senior tax credit) even if they don’t have a mortgage.

After taking all those subsidies into account, it appears that relatively speaking renters are worse off in Howard County than homeowners. One way to address this imbalance would be to provide further subsidies to renters, whether at the county, state, or national level. Another way would to increase the stock of affordable rental housing, whether through a specific focus on building below-market-rate units, or by building more apartments and other rentable housing in general.

How best to provide affordable housing is a topic of great debate in Howard County and elsewhere. I don’t have a definitive opinion on this matter yet, so I’ll sign off this week with the promise that I’ll try to look at this issue in more detail later when I have time to do more research. For now you can check out the code and data behind this post in the article “Howard County, Maryland, relative housing costs by census tract.”

As always, if you find these posts interesting and useful please tell other people about them and encourage them to subscribe to the Civility and Truth mailing list. Having readers who care enough to subscribe helps motivate me to send these posts out on a regular basis, and the more readers I have the more motivated I’ll be. In the meantime, thanks for reading this post!

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

This week I’m again taking off from doing data analysis, but my topic is not actually that removed from my previous posts on population density in Howard County: I take a look at the long history of political conflict between urbanized areas and rural areas, with a focus on how that played out in the United Kingdom in the 19th century.

From their beginnings in Mesopotamia six thousand years ago cities have attracted those in search of opportunity and attractions. But cities have also been seen as alien intrusions upon a landscape of farms, forests, and grasslands perceived to be the true heart of a nation.

In a predominantly agricultural economy those who possessed land thereby possessed political power. As cities grew larger and wealthier their inhabitants, and especially their elites—merchants and traders—sought a share of such power, but faced an uphill battle in a political system that was rigged in favor of a land-owning aristocracy.

One such battle occurred in the UK in the late 18th and early 19th centuries, as it was making the transition from an agricultural to an industrial economy. At that time members of Parliament were elected by a voting population that was not only small but unevenly distributed between rural and urban “boroughs” (electoral districts).

The most infamous example of this was the borough of Old Sarum (shown above), site of a hilltop fortress and associated settlement that had been abandoned in the 14th century but was still allotted two members of Parliament through the 18th and early 19th century. In 1831 those two members were elected by a total of only eleven voters, all absentee landowners.

Old Sarum was the most extreme of what were termed “pocket boroughs” or “rotten boroughs”, but it was far from the only one. According to Wikipedia, “By the time of the 1831 general election, out of 406 elected members, 152 were chosen by fewer than 100 voters each, and 88 by fewer than fifty voters.” The overall electorate at that time was about 400,000 people (where “people” here really means “men with substantial holdings of property”), and the population of the UK around 25 million or so.

Thus in the early 19th century a majority of the Parliament (240 members out of 406) were elected by less than 20,000 voters, less than 5% of the electorate and less than 0.1% of the total UK population.

The painting above shows a different kind of landscape, dominated by the factory chimneys of the city of Manchester. At the center of Britain’s industrial revolution, Manchester grew from a small market town of 10,000 people in the early 18th century to a city thirty times that large by the mid-19th century. It ended the 19th century encompassing a metropolitan area of over 2 million people.

This rapid growth left Manchester both disenfranchised as a city (it shared one member of Parliament with the surrounding county) and populated by a new working class that grew restive at its treatment and hungered for representation. However this movement was strongly resisted by the political powers that be: the French revolution and the subsequent Napoleonic wars had rendered the landowning British aristocracy paranoid about the possibility of revolution closer to home.

The local militias raised against an invasion of England that never came were thus repurposed into a force to guard against the possibility of internal rebellion. The villain George Wickham in Jane Austen’s Pride and Prejudice was a member of such a militia, and at the end of the novel was deployed to the industrial town of Newcastle upon Tyne in the north of England.

Opposition to reform was of course not couched as a way to retain power and privilege, but rather as advocacy of government by an enlightened and educated property-owning aristocracy, instead of “mob rule.” Jane Austen implicitly made this argument in novels like Pride and Prejudice, which features a young landed aristocrat (Mr. Darcy) growing into his responsibilities as a member of the ruling class, and Mansfield Park, which contrasts the sedate and staid morality of the eponymous country estate with the laxity of London—where “everything is to be got with money,” in the words of antagonist Mary Crawford.

However Austen was no fool. Through the words of her heroine Fanny Price she implied that the tranquility of Mansfield Park was established on the backs of slaves, and she also knew that many local gentry fell far short of the ideal of Mr. Darcy—see for example her portrait of the idle and vain baronet Sir Walter Eliot in Persuasion. She was a member of a nascent middle class, many of whom would make common cause with their working class brethren in support of political reform.

In August of 1819, almost exactly two hundred years ago (and two years after the death of Jane Austen), the growing movement for Parliamentary reform met violent resistance, in what became known as the Peterloo Massacre: A crowd of tens of thousands of people in Manchester (a substantial fraction of the local population) was gathered to hear a speech by the radical reformer Henry Hunt, when they were set upon by a local militia ordered by city magistrates to disperse the crowd. Eighteen people were killed in the ensuing chaos (depicted in the cartoon above), and several hundred wounded.

(Incidentally, the British director Mike Leigh—himself from Manchester—did a film about this event, titled simply Peterloo; it’s available on Amazon Prime Video. It’s less of a conventional film and more of a dramatized documentary, and its portrait of the aristocracy and the local magistrates shades over into caricature more than once, but it does give a good feel for the rhetoric and passions of the time.)

This act of violent suppression of a peaceful political gathering raised a wide public outcry, but in practical terms its result was a doubling-down of efforts to suppress movements for electoral and other reforms.

The movement for Parliamentary reform won its first victory in the form of the 1832 Reform Act, which eliminated the worst of the rotten boroughs, created new boroughs for cities like Manchester, and expanded the franchise significantly—although it was still restricted to male property owners. (The painting above shows the first Parliament elected after passage of the Act.)

Even this partial victory did not come easily: the bill went down to defeat twice, the second time due to opposition in the House of Lords, and only passed the third time after widespread urban riots and threats by the king to appoint new reform-minded peers.

The movement for Parliamentary reform continued, embodied in the movement known as Chartism. The actual demands of the Chartists seem rather mild and conventional by present-day lights: they included universal male suffrage, secret ballots, constituencies of equal size, and salaries for Members of Parliament. These were presented in the form of various petitions to Parliament over the years, each signed by over a million people. (The engraving above shows a procession conveying the petition of 1842.)

Nonetheless the Chartists, like the Parliamentary reformers before them, were portrayed as radical revolutionaries bent on overturning the established order. The petitions were ignored, and eventually the movement withered away at mid-century.

However in the longer term the reformers’ demands were met: the secret ballot was introduced in 1872, constituencies were reorganized in 1885, and the franchise was steadily expanded, culminating in the extensions of the franchise to women in 1918 (for women 30 and older) and 1928 (for all adult women).

So, enough history. What’s the point? The point is that we still live with this legacy of city vs. country, in both our political institutions and our political rhetoric. The rhetoric harks back to the conflicts between Thomas Jefferson, who envisioned America as a land of independent small farmers, and Federalists like Alexander Hamilton, who sought to make America’s cities powerhouses of commerce and industry. It echoes down to the present-day, when areas like the deep South or the Great Plains are lauded as “the real America” while multicultural metropolises like New York City and Los Angeles are not, even though the latters’ metropolitan areas are home to 1 in 14 and 1 in 25 Americans respectively.

More importantly though, our political institutions favor country over city, most notably in the US Senate, an upper legislative chamber in which predominantly rural states with small populations have equal representation with heavily-urbanized states with populations an order of magnitude larger. The more rural states can thus exercise disproportionate political power in everything from blocking legislation to confirming Supreme Court justices.

The US Census bureau cartogram above helps illustrate this; it retains the shape of states but shows their sizes relative to their population. For example, California by itself appears to equal at least six of its neighboring western states in population, with Alaska thrown in as well, but has only two Senators compared to their twelve or fourteen.

A common response to this observation is “we’re a republic, not a democracy.” This harks back to the Founders’ fear of direct (as opposed to representative) democracy, but in the present-day context is just a fancier way of saying “them’s the rules, go pound sand.” Thus I don’t think this is an answer that should be taken seriously.

So what are the “take aways” here? If the history of Parliamentary reform in the UK is any guide, in the political war between city and country the continued growth of cities ensure that they will win in the end. In the UK it proved unsustainable for a relatively small minority to exercise power over a large majority, and so it will prove in the US. The only questions are: how long will it take? how exactly will it happen? and how severe will be the conflicts that drive the change?

The answer to the first question I think may be, surprisingly long. Depending on exactly when you consider it started and ended, the drive for Parliamentary reform took anywhere from fifty to a hundred and fifty years. (And even today the British political system, like that in the US, uses a system of “first past the post” elections in single-member districts that blunts the political power of popular movements: a party that can command 49.9% of the vote in all districts—but no more—finds itself with no representatives at all.)

In the US it took about eighty years from the creation of the US Constitution to the Reconstruction Amendments that formally gave former (male) slaves the vote, and almost as long again to extend the franchise to women and provide for direct election of US Senators. Given this history, it could easily take half or more of the 21st century to change the institutional imbalances inherent in the present system.

As for how it might happen, the US has a problem in that its political framework is considerably more rigid than that of the UK. The key is the US Senate: it (along with the House of Representatives) must approve proposed amendments to the Constitution, must (again with the House) approve applications for statehood and thus the new Senate seats those would bring, and has the sole power to confirm Supreme Court justices who might be relied upon to uphold the existing order as much as possible. The Senate is likely therefore the last redoubt of an otherwise minority party: if it falls, and falls for good, defeat is almost certain.

Amendments to the Constitution require a two-thirds majority in the Senate as well as the House (or approval of two-thirds of state legislatures to hold a constitutional convention), along with ratification by three-fourths of the states (or their respective ratifying conventions). I therefore doubt any fundamental constitutional change will occur in the foreseeable future relating to the Senate (or the electoral college, for that matter). More plausible is the admission of new states to the union, either jurisdictions like the District of Columbia and Puerto Rico already proposed for statehood, or new states formed by splitting up large and populous states like California.

Assuming the state legislatures were in favor (which is not necessarily a good assumption in splitting states, since many might object to losing territory), this would require only simple majorities in the Senate and House along with a President willing to sign the bill. (I am assuming that by this time the Senate filibuster would have been abolished in an effort to retain Senate control with only a slim majority.) Of course, this works both ways: assuming the opposing party were to subsequently gain control of Senate, House, and Presidency, it could work with cooperative small states to split them into even smaller states and thus regain ground in the Senate.

Finally, how will change happen, especially if control of the Senate remains contested? Those with a bloodthirsty bent may be hankering after a second US civil war, but based on past history in other countries it’s more likely that widespread nonviolent civil disobedience will be the key to future US institutional change. The “magic number” is hypothesized to be about 3.5% of the population or more. In US terms this would correspond to at least 12 million people actively participating in protests and generally mucking up the workings of society, business, and government in a nonviolent way in pursuit of a coherent set of political goals.

The bottom line: I think change will come, but I anticipate that the (US) war between city and country will not end in my lifetime.

That’s all for this week, I hope to write to you again in two weeks. A reminder: if you find these posts interesting and useful please tell other people about them and encourage them to subscribe to the Civility and Truth mailing list. Having readers who care enough to subscribe helps motivate me to send these posts out on a regular basis, and the more readers I have the more motivated I’ll be. In the meantime, thanks for reading this post!

My apologies, there is no Civility and Truth newsletter this week. As I’ve mentioned previously, my typical posts take a fair amount of time to prepare, and I just didn’t have the hours this week to do so.

It looks as if this is going to be an ongoing problem given my work and family responsibilities. Rather than missing deadlines again and again, I’ve decided to just move to a new schedule where I aim to send out a post every two weeks. Expect the next one next Wednesday, August 28, at the usual time.

I’m out of town this week on a business trip, so I haven’t had time to prepare a proper post. However, I wanted to keep to my weekly schedule, so what better than a quick post on why I’m so slow in putting my opinions out on the Internet?

The simple fact is that my personality is not suited for hot takes. Before I write something about a topic I feel compelled to have as deep an understanding of it as I possibly can. Unfortunately, given that blogging is a spare-time hobby of mine, that means it can be a long time between the time I conceive of an idea for a post and the time I actually publish it.

For example, one of the posts I did for my “Creating the Chrysalis” series took about a year to write, because I just didn’t feel I knew enough about the technologies and concepts that went into how the Chrysalis amphitheater was designed. In the end it took watching two hour-plus videos featuring Marc Fornes (the Chrysalis designer), transcribing all of one video plus a good portion of the other, reading two books on Computer-Aided Design, and trying out for myself the CAD software that Fornes used.

Another example is my “Seven Answers” series: I was inspired by a post that Jason Booms did in January of 2018 prior to the primary election filing deadline, laying out “seven questions” for the local Democratic candidates to address. I didn’t start the series until late March, and didn’t finish it until just before the general election. If I’d been running for office it would have been a dismal performance.

But the simple fact is that I’m not running for office. I’m also not writing for a living, or looking to become a social media “influencer.” If I need to take the time to understand something before commenting on it, nothing is stopping me from doing it.

Well, nothing except the obligation I feel to those of you who’ve signaled your willingness to read what I write by signing up for this newsletter. The compromise I’ve made is to try to put something out every week, even if it advances by only a bit the end goal of saying something interesting about whatever topic I’m currently focusing on. (Lest we forget, right now I’m trying to get my head around housing affordability, density and zoning, and related issues.) And along the way you’ll get the benefit of seeing how I fumble my way to a better understanding, with links and code and data if you want to explore things further on your own.

So thank you for your patience, which I’m afraid I will continue to try.

P.S. If you’re interested, I also was finally able to publish the code and data for my previous post.

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

This is a follow-up to my post about home value declines across Maryland, and a companion to my post about median household income within Howard County. For reasons discussed in that post, this analysis covers only the period from 2010 forward, and in particular compares median home values in the 2006-2010 time frame with median home values in the 2013-2017 time frame.

As with median household income, the margins of error for median home value estimates at the census tract level are high, up to 10% or more above and below the estimates themselves. So again I decided to divide the 55 Howard County census tracts into 5 quintiles of 11 tracts each, with quintile 1 containing the 11 census tracts with the lowest median home values, quintile 5 containing the 11 census tracts with the highest median home values, and the other quintiles in between.

The result is the map above, showing which census tracts are in which quintiles according to the 2010 American Community Survey 5-year estimates of median home values. The areas with high and low median home values should not surprise anyone familiar with Howard County real estate: with one exception the census tracts with the highest home values are west of US 29, and the tracts with the lowest median home values are east of US 29 or in Columbia.

How did this situation change over time? In the graph above I show how the average (inflation-adjusted) median home value for each 2010 quintile changed from 2006–2010 to 2013–2017.

(A reminder here: I write “2010 quintiles” because the census tracts are grouped according to their rank in 2010. If a tract was assigned to, say, quintile 2, in 2010 then it will still be considered in quintile 2 when comparing 2013-2017 to 2006-2010. Also, as noted in the last post averaging median values is not strictly speaking correct, but is the best that can be done in the circumstances.)

From the graph above we see that all quintiles experienced a decline in average inflation-adjusted median home values over the time frame in question. The quintile with the largest decline in percentage terms was quintile 1, for which average inflation-adjusted median home values dropped 19% from 2006–2010 to 2013–2017. The quintile with the smallest decline in percentage terms was quintile 3, with a 13% drop.

This map is an alternate approach to showing the changes in median home value between the 2006-2010 and 2013-2017 time frames. It shows the percentage declines in median home value for each census tract in Howard County. No census tract had an increase in median home value between the two time frames, and only a couple of census tracts experienced small declines. At the other end of the spectrum a few tracts experienced declines of between 30–40% in median home value.

The final map shows the distribution of census tracts among median home value quintiles based on the 2017 ACS 5-year estimates. (In other words, I recalculated which tracts were in the top 20%, etc., based on the newer figures.) The overall picture is similar to that for the 2010 map (high median home values west of US 29, low median home values east of US 29), but some tracts have moved up to a higher quintile or down to a lower quintile. This may be due to how home values fared in those tracts, or simply to random errors in the estimates. (Recall my note above about relatively high margins of error for tract-level estimates.)

For the code and data used to produce the graphs above, see “Howard County median home value trends by census tracts.”

That’s all for this week. A reminder: if you find these posts interesting and useful please tell other people about them and encourage them to subscribe to the Civility and Truth mailing list. Having readers who care enough to subscribe helps motivate me to send these posts out on a regular basis, and the more readers I have the more motivated I’ll be. In the meantime, thanks for reading this post!

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

After looking at median household income for Maryland counties, I’m now turning my attention to median home values for the same geographies. But before I start, a brief apology for getting this out two days later than my (self-imposed) schedule; I’m switching to a new method for retrieving census data and map geometry, and it took longer than I thought to get the kinks out of my code.

As you can see in the map above, median home values in Maryland have not yet recovered from the Great Recession. As the following bar chart shows, all Maryland counties experienced declines in inflation-adjusted median home values from the 2006–2010 timeframe (used for the 2010 American Community Survey 5-year estimates) to the 2013–2017 timeframe (used for the 2017 ACS 5-year estimates). Allegany and Garrett counties fared the best, followed by Montgomery and Howard counties and Baltimore city, with declines ranging from 8–15%. The worst counties for median home value declines were Prince George’s and Charles counties, with almost 25% declines.

(For the code and data used to generate these and other graphs in this post, see “Maryland median home value trends per county.”)

Now, the declines across Maryland are not always directly comparable. For example, as you can see in the two bar charts below, the inflation-adjusted 2010 5-year estimates for median home values in Allegany and Garrett counties and Baltimore city range from just above $100,000 to almost $200,000, so those home values were relatively low to begin with.

On the other hand, the inflation-adjusted 2010 5-year estimates for median home values in Montgomery and Howard counties were over $500,000. A decline of almost 15% from the 2010 estimates to the 2017 estimates therefore represents a decline of about $70,000 to $80,000 in real terms.

The following maps correspond to the bar charts above, and show median home values across Maryland per the 2010 and 2017 ACS 5-year estimates. Again Montgomery and Howard counties stand out for having relatively high median home values, even after the declines between the 2006-2010 and 2013-2017 time frames.

Some concluding remarks: In general we see stable or declining prices as a good thing—no one ever wishes for (say) cars to be more expensive than they are, and the decline in prices for computers and other electronic devices has been universally welcomed.

However homes are simultaneously something you use and something you invest in. Before you buy a home the price is an obstacle. After you buy a home its price is an indicator of your future returns. This complicates the issue of housing affordability in many ways, most notably by pitting the interests of existing homeowners against those of would-be homeowners. I’ll probably have more to say about this in future.

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

After looking at the median household income of Howard County relative to other local jurisdictions, I now look at median home values. As I noted previously, the earliest county-level data I can find dates from 2005 and the beginning of the American Community Survey.

The graph above shows Howard County median home values over time compared to a select set of other jurisdictions. (I chose the same jurisdictions as in my previous analysis of median household income, for the same reasons.)

All values are in current dollars as of the year of the survey. Thus, for example, the values for 2005 are expressed in 2005 dollars, the values for 2006 are in 2006 dollars, and so on. (I did not use inflation-adjusted values because housing costs are themselves a major component of the Consumer Price Index. Multiplying by a CPI deflator would therefore understate any actual rises in home values.)

Here’s some immediate takeaways from the graph above:

• Unlike other counties (not to mention the US as a whole), Howard County home values have not yet fully recovered from the Great Recession, and have been essentially stagnant since 2014. The same is true for Montgomery County and Baltimore city, as well as possibly for Anne Arundel County.
• In contrast, home values in Virginia counties are rising, and in the case of Fairfax County have fully recovered (at least in nominal terms).
• The District of Columbia has experienced a major rise in home values, over 50% since the start of the time period covered in the graph.
• Stafford County, Virginia, has significantly lower home values than Howard County, but a median household income that is essentially equivalent to that of Howard County. (See my previous article.)

This second graph shows the ranking of Howard County over the years versus the most affluent local jurisdictions. (Note that these rankings do not reflect home values in counties or county-equivalents with populations less than 65,000. These jurisdictions are too small to be included in the American Community Survey 1-year estimates.)

The story is similar to that from the previous graph:

• Howard County has gone from being in the top 25 counties (or county-equivalents) based on median home value to barely being in the top 50. (Its actual rank for 2017 was at number 45.)
• As it did in the median household income rankings, the District of Columbia has come from being out of the top 50 to being in the top 20 in terms of median home value. (Its actual rank for 2017 was at number 16.)

This is consistent with the narrative of major “showcase” cities (DC, New York, San Francisco, etc.) experiencing economic success and an accompanying rise in home prices, while other cities (like Baltimore city) and some suburban areas (like Howard County) languish.

I’ll continue looking at median home values, but may post about something else next week for a change of pace. As always, if you find these posts interesting and useful please tell other people about them (feel free to forward them these emails if you’d like) and encourage them to subscribe to the Civility and Truth mailing list themselves. In the meantime, thanks for being a subscriber and reading this post!

UPDATE 2024/01/04: This post originally appeared on my Civility and Truth Substack newsletter. I’ve moved it to my main site in an effort to collect all of my writing in one place.

After my data analysis of Virginia counties’ household income fell apart so spectacularly (see my last post) I felt I needed to revisit the general issue of looking at county-level median household income statistics. Because I live in Maryland (not Virginia) I decided to do it with my own state instead of worrying about someone else’s.

Almost every data analysis has a “money graph” that jumps out at you as being particularly insightful, or at least particularly interesting. The map above is my candidate for the money graph for this analysis, following closely by the bar chart below.

They show the same information in two different ways: I took the US Census Bureau American Community Survey 5-year estimates for median household income for all 24 Maryland counties (including Baltimore city), and adjusted them for inflation to express everything in 2017 dollars. (2017 is the last year for which we currently have estimates.) Then I looked at the differences for each county between the 2010 5-year estimates (which actually cover the years 2006-2010) and the 2017 5-year estimates (which actually cover the years 2013-2017). (See “Maryland median household income trends by county” for the full analysis with code.)

As you can see from the bar chart above, between these two timeframes every Maryland county experienced stagnant or declining median household income except for Baltimore city, for which median household income grew about 5% in real terms. Now, admittedly Baltimore city still is not a very affluent place, as you can see in the next bar chart, showing 2017 5-year median household income estimates.

Nonetheless, given the public perception of Baltimore city it’s surprising that its median household income was growing at all over this period, much less that it outpaced every county in Maryland. Contrast Baltimore city with Somerset County, also a relatively poor jurisdiction, which saw its median household income drop about 20% between the two timeframes in this analysis.

As I think I may have mentioned previously, median household income in a city or county can increase because the people already living there get more affluent, because more affluent people move there, or because less affluent people move away—or it could be a combination of all three factors. I don’t know exactly what’s happening in Baltimore city with respect to its increase in median household income, and I’m not sure exactly what data would be needed to know more.

In any case the situation of Baltimore city brings to mind the conventional wisdom that large cities and their metropolitan areas are thriving at the expense of rural areas and smaller cities. (See for example these articles from the Washington Post and New York Times that I turned up in a quick Internet search.) That story is certainly true for DC, which I previously highlighted as having a median household income that has skyrocketed in the past thirty years, and now matches or exceeds that of Maryland.

That in turn makes me wonder: if over the next thirty years Baltimore city could see even a significant fraction of the income growth that DC has seen over the last thirty, what might that mean for the economy of central Maryland? We often think of the troubles of Baltimore city, economic and otherwise, as being irrelevant to us out in the suburbs. However, it’s possible that even the relatively modest income gains Baltimore city has seen are showing us something different: that the city is an economic engine for central Maryland, and if it could revved up even more it might pull Howard and other counties out of the relative income stagnation they’ve experienced over the past few years.

That’s all for this week. As always, if you find these posts interesting and useful please tell other people about them (feel free to forward them these emails if you’d like) and encourage them to subscribe to the Civility and Truth mailing list] themselves. In the meantime, thanks for being a subscriber and reading this post!

I’ve finished up my data analyses on median household income for now (although I have yet to write up the census tract analysis for civilityandtruth.com), and will next be looking at home values. However, before I started that I was curious about an issue I touched on in my last post on civilityandtruth.com:

I will note that median household income for the Northern Virginia suburbs is increasing even as median household income for Virginia as a whole is stagnant or decreasing. This is presumably due to the rest of Virginia suffering economic problems to which Northern Virginia is immune.

I was about to write something about how Virginia’s problem is poor rural areas vs. rich urban areas, following the conventional wisdom that a lot of political commentator are talking about. I then decided not to be so dogmatic about it and commit myself to any particular interpretation before I looked at any data. Since I already had data on median household income by county for the entire US, I did an off-the-cuff data exploration, including doing graphs of median household income for all counties and cities in Virginia.

To some degree this went against my initial impression, and showed me that “urban = rich, rural = poor” is an oversimplification. However it also showed me the dangers of being too quick to take data at face value: there’s a twist at the end that threw my entire analysis into question.

Since Virginia is not my main focus I decided not to do a full write-up on this, but I thought it might be interesting and informative to take you through how RStudio and the Tidyverse package make these sorts of informal analyses fairly easy to do, and also how you need to watch out for potential problems along the way.

A quick and dirty analysis

I started with my RStudio.cloud hocodata project open in my browser, with the R Markdown source file open that I used to create my full analysis of Howard County median household income vs. other counties and cities.

R Markdown is a document format in which you can interweave prose material with “chunks” of computer code, in this case code in the R language. The file already contained the code I needed to read in the census data files and recreate the county-level median household income table I wanted to use, so I just stepped through the file and had RStudio execute the chunks of code I needed to run.

Once I had done that I had a copy of the data table mhi_tbl that I had used as the basis for my original analysis. This table contains median household income data on all counties for the years 2005 through 2017, as well as equivalent data on the state and national level. I used RStudio to display the table and refresh my memory on the column names (geo_id, geo_id2, etc.) and the format of the data values.

The state-level rows have a geo_id value starting with “04”, and the county-level rows have a geo_id value starting with “05”. Since I was interested in county-level data only my first step was to filter the data table and see if I could get just the county-level data returned. In Tidyverse syntax this is done using the command mhi_tbl %>% filter(substr(geo_id, 1, 2)) == "05"): take the table mhi_tbl and use it as input to the filter operation.

This looks like it worked. The next thing I needed to do was to just look at counties in Virginia, so I went back to look at the table mhi_tbl and searched for the word “Virginia.”

This told me that I could find counties specifically in Virginia by looking for values of geo_id2 that began with “51”. I already had a command that filtered for counties, so I took advantage of the Tidyverse ability to use the output of one command as the input to the next, and added an additional filter step filter(substr(geo_id2, 1, 2) == "51") at the end:

This looked like it was just Virginia counties, but why were there 379 rows in the result? I had forgotten that the table mhi_tbl contained data for multiple years. I was mainly interested in comparing Virginia counties against each other, so I decided to cut the data down to look at just 2017. This was easily done by adding another filter step: filter(year == 2017). (I should note at this point that in RStudio I didn’t have to retype the entire line. I just used the up-arrow cursor key to go to the last command executed, and then added the new bit of code to the end.)

This returned only 30 rows, which sounded like a more reasonable number.

Since I was using multiple filter steps in a row, I could have combined them into a single filter step with three criteria: mhi_tbl %>% filter(substr(geo_id, 1, 2) == "05", substr(geo_id2, 1, 2) == "51", year == 2017). However, since this was a quick-and-dirty analysis and I was being lazy I didn’t bother. Instead I just wanted to get to graphing something.

(Also note that I was using the table mhi_tbl containing income values in current dollars, i.e., not inflation-adjusted, as opposed to the table mhi_adj_tbl, which expresses income all values in 2017 dollars. Since I was specifically looking at median household income for 2017 the values in the two tables were the same and there was no reason to use one vs. the other.)

Graphing the data

I wanted to compare Virginia counties against each other, so I decided to do a bar chart where each bar represented a county and the height of each bar corresponded to the median household income for that county. Using the ggplot2 package (part of the Tidyverse) this is done using the syntax ggplot(aes(x = geography, y = mhi)) + geom_col().

The ggplot part specifies that the x-axis for the graph will have the names of the counties (geography) and the y-axis will show the median household income values (mhi). (The “aes” is from “aesthetic”, i.e., how we want the graph to look.) The geom_col part then tells the ggplot2 package that we want a bar chart specifically, as opposed to some other type of chart. Finally, note that when we get to the ggplot2 functions we use + to separate the various operations, not %>%.

Unfortunately this is pretty much unreadable, as all the county names have been displayed on top of each other.

Fortunately I had seen this problem before, and after doing some Internet searching I had discovered how to tell ggplot2 to display the x-axis text at a 45-degree angle. I had adopted this solution in the analysis from which I was stealing code, so I just cut-and-pasted it in: add theme(axis.text.x=element_text(angle=45, hjust=1)) to the end of the plotting statements. Unfortunately the particular value of 45 degrees didn’t work too well, as the leftmost county name got cut off a bit. I tried using a 90-degree angle, which produced vertical text, before settling on a 60-degree angle.

This is more readable, but I still wasn’t satisfied. What I really wanted was a graph of counties and cities arranged left to right from highest to lowest median household income.

I did another Internet search to look for ways to do this. It was a bit more complicated, as R and ggplot2 are not intuitive when it comes to handling so-called categorical variables, i.e., variables like county names that are not numeric values. I tried a couple of suggested approaches that didn’t work before finding one that did: insert mutate(geography = fct_reorder(geography, -mhi)) into the “data pipeline” before passing the data to ggplot.

This takes the geography variable containing the county names, which was previously just a simple character string, and converts it into a new variable of the same name that is a factor variable (“factor” being R-speak for a categorical variable). Factor variables have an inherent ordering imposed on them, and in this case we use the fct_reorder function to order the geography values based on the values of the mhi variable. (The - before mhi means we are ordering the geography values in decreasing order of median household income.)

This was almost what I wanted, but I didn’t like the scientific notation (“1e+05” for 100,000) on the y-axis. Fortunately I had already investigated this issue and found a solution that I used in my previous analysis. I just cut-and-pasted it from my existing code and tweaked it slightly, using scale_y_continuous(labels = scales::dollar, breaks = seq(0, 140000, 20000)) to display the y-axis values as properly-formatted dollar amounts, and display y-axis labels from 0 to $140,000 in intervals of $20,000.

This is much better in terms of readability. And then I tried to interpret it…

Interpreting the data

The US median household income in 2017 was about $60,000, while the median household income for Virginia is above $70,000. When I eyeballed the graph above it looked as if there were three Virginia counties with median household incomes below $60,000: Rockingham, Augusta, and Montgomery, all in the Shenandoah Valley/I-77 corridor.

However, the bottom seven “counties” in median household income in the graph above are actually cities, which in Virginia are often independent jurisdictions (in the same way that Baltimore city is an independent jurisdiction in Maryland). This complicated the “urban = rich, rural = poor” story: was it possible that rural areas in Virginia are actually doing relatively better than cities—or at least cities that are not in Northern Virginia?

Hold that thought. In the meantime I also wanted to look at Maryland counties for comparison. Having built this long R command line to produce a graph for Virginia, I could re-use it to produce the same graph for Maryland, simply by filtering on a geo_id2 value of “24” rather than ‘“51.”

The resulting graph looked familiar, with Howard County at the top and Baltimore city (almost) at the bottom. The counties at or below the US median household income are Wicomico on the Eastern Shore and Washington and Allegany in western Maryland, all relatively rural. So far so good according to the conventional wisdom.

But then I looked closer at the lower-income counties and thought, “where is Garrett County, shouldn’t it be near the bottom as well?” I looked even closer and couldn’t find Garrett county at all in the graph. Then I counted just to be sure and found there were only 16 jurisdictions listed. I knew from prior analyses that Maryland had 24 counties (including Baltimore city). What happened to the other eight?

Misinterpreting the data

After some thought I figured out the problem. (I’ll pause here to let you think about it as well.) The problem is that . . .

The median household income data I was using from the American Community Survey 1-year estimates. I used these because I wanted to be able to do year-to-year comparisons for changes in median household income by county. But as the US Census Bureau explains in “When to Use 1-year, 3-year, or 5-year Estimates,” the 1-year estimates include only areas with population of 65,000 or greater.

The population of Garrett County is only about 30,000 people, so there are no 1-year estimates for it. Ditto for the other Maryland counties not on the list above; the most populous of the group is Worcester County at just above 50,000 people. The situation for Virginia is even worse: there are 95 counties and independent cities in Virginia, so the 30 jurisdictions in the graph above are less then one-third of the total.

The bottom line is that my off-the-cuff analysis was fatally flawed, and any conclusions I might have drawn from it were suspect to say the least.

Fixing the problem

How could I have corrected my mistakes? The obvious fix is to use the ACS 5-year estimates, which do include all areas no matter how small. As it happens I never bothered to download the ACS 5-year data for all US counties, so I wasn’t able to quickly implement this fix. Instead I just abandoned this particular analysis, at least for now.

But even if I had ready access to the 5-year data I couldn’t just repeat the above analysis verbatim. In particular, showing 30 Virginia counties and cities in a bar graph produced a busy enough graph, but showing 95 cities and counties in the same kind of graph would make it almost unreadable.

What would be the best alternatives? One approach would be to graph only the top ten counties and cities by income, or only the bottom ten. Another approach would be to use a list format rather than a graphical format, and again list only the top and bottom ten.

Lessons for today

So what might we learn from my experience?

First, I hope you come away from this with the feeling that these types of basic analyses and visualizations are not that difficult to do. I may have written this elsewhere already, but in my opinion anyone who’s comfortable with working with Excel spreadsheets and their associated formulas and macros should be able to do useful and interesting things with RStudio and the Tidyverse package.

(I’ll also add that I think these sorts of general skills with manipulating and visualizing data are exactly what we should be teaching children in school, to make them more “numerate,” prepare them for future jobs, and enable them to be more educated citizens.)

Second, at the same time the classic computer programming axiom applies here: “garbage in, garbage out.” If the data is bad to start with, or if we’re using the wrong data, then skill in analyzing it won’t help. The mistake I made above would have been immediately obvious to someone with more knowledge of Virginia than I have, but other data errors can be more subtle and hard to catch.

For example, when I did the graph and list ranking counties by median household income, those were based on the ACS 1-year estimates, and thus omitted counties falling below the population threshold of 65,000 people. For all I know there may be more affluent counties that might have ranked in the top ten but are only represented in the 5-year estimates.

Third, following on from the above point, in order for people to determine how accurately data visualizations reflect reality, and how much credence they should give data-based arguments, they really need to have access to the underlying data and code used to generate the visualizations and arguments. It’s not that everyone needs to be able to replicate the graphs and analyses, but there should be at least some people who can provide independent opinions on the matter.

Traditionally data visualizations and data-based arguments have been generated by governments, the media, and academia. In many if not most cases these have been presented as final products, with no indication as to what went into making them. Instead we were asked to rely on the authority and prestige of their creators.

Generating sophisticated data analyses can now be done by independent individuals, and for various reasons people no longer trust “the experts” as much as they use to. That’s why I “show my work” by making all the code and data for my analyses available to anyone who wants to inspect them for errors and attempt to replicate what I’ve done.

Thankfully there are more and more organizations that are doing the same. One great local example is the “data desk” at the Baltimore Sun. They have an active presence on Twitter (@baltsundata) and make a lot of their code and data available on GitHub, a service used by many open source developers. (For example, see the code and data repository for a recent story about Baltimore population losses.)

That’s all for this week.

I’ve set an informal goal for myself of sending out something to this mailing list each and every week, and Wednesday by default seems to have become the day I do that. This week work and home responsibilities have prevented me from doing a full post.

However, I did want to recommend a Niskanen Center blog post from last year by Joshua McCabe: “How the SALT Deduction Subsidizes Opportunity Hoarding” (“SALT” = “state and local taxes”). When you combine McCabe’s argument with an argument I previously made in my post about Judith Rich Harris, “How do schools and parents matter?,” I think it provides what (as an ex-physics major) I’d call a “unified theory” regarding the problem of housing affordability in Howard County. Hopefully I’ll have time to sketch that overall argument out in next week’s post.

UPDATE 2021/07/04: I have discontinued this newsletter and am in the process of moving its articles to my personal blog. To keep track of what I write in future, follow me on Twitter (@hecker) or point your RSS news reader at frankhecker.com.

For many years now I’ve maintained a personal blog, now published as Civility and Truth, on which I’ve written about a variety of topics of interest to me, including the history of county council redistricting in Howard County, Maryland, the evolution of the Inner Arbor plan for Symphony Woods in downtown Columbia, Maryland, and the creation of the Chrysalis amphitheater, my data-influenced takes on political, economic, and social issues, and whatever else comes to mind, from Maryland’s restrictions on direct-to-consumer genetic testing to reviews of signs for local political campaigns.

I don’t update the blog very frequently, and it’s too much to expect people to check out the web site on a regular basis. Hence this newsletter: I’ll use it to send new Civility and Truth posts direct to your inbox at the same time I post them to the site—or even before I post them to the site, as a reward to subscribers. I’ll also send out other material that for whatever reason doesn’t warrant a full Civility and Truth post, including links to recommended articles, mini book reviews, and commentary on issues where I haven’t fully formed an opinion and/or haven’t yet done a lot of research. But I’ll respect your time: I plan on sending something out no more than once a week.

As with my web site, I will not include advertisements or other promotions for goods or services the sale of which would benefit me—although of course I’ll link to sources for books, movies, etc., that I review. However since I’m publishing this for free through the Substack service it’s possible that Substack itself may insert ads for itself or others, either now or in the future.

Finally, this is a free newsletter and will remain so for the foreseeable future. (In the unlikely event that I ever do get to the point where it makes sense to provide paid content, I’ll have some sort of promotion to reward early subscribers.) Please feel free to forward any of my posts to anyone you think may be interested.