Whether its residents relish the role or not, the District of Columbia is the nation's capital city. As such, it symbolizes what America stands for. What it does, and how it does it, is broadcast around the world. It helps shape global opinion of our down-to-earth success in achieving our lofty goals and ideals. Who lives here, votes here, pays taxes here, and governs here has the unique burden of being on display as America's urban showcase.
The city is now struggling to prepare a new 20-year Comprehensive Plan that appropriately recognizes virtually every American special interest. A combination of idealism and pragmatism, it will provide a public guide book on how the city should grow. How it grows will depend on who it attracts to live here and work here. This in turn depends on how the city's limited land is zoned for continued development. Housing, and all that goes with it, will help define the desired evolution of the city's "residential mix", the lifeblood of the capital city's 'body politic'.
This analysis is intended to provide some (updated) quantitative data to inform that ongoing debate.
[It is also the subject of NARPAC's Editorial for February, '06]
Due to the amount of quantitative graphic data presented in this new analysis (much of it from the Census Bureau's 2004 American Community Survey), this work has been broken into seven bite-sized pieces for easier loading, printing and greater reader selectivity. You can read the brief summary immediately below, and then decide to continue to the end of this chapter, or shift to another one by clicking on the chapter titles listed below.
There are a number of instances in the CHSTFR where quantitative material is presented that seems either too vague, or in some instances inconsistent. For instance, the "normal" share of a householder's income that can conveniently be devoted to housing appears at odds with available statistics. In this and other areas, NARPAC presents additional statistical data which may improve the credibility of the planning effort. Some of these include:
Issues addressed here:
No indication of how/why the housing allocation was made
If this does not hold your interest, please click ahead to:
Chapter 4: LOOKING POVERTY SQUARELY
IN THE FACE
No indication of how/why the housing allocation was made
Informal comments at the briefing quite candidly mentioned that these numbers of houses was a "compromise" between task force members. There is, of course, no harm in that, but some additional rational is surely warranted. For instance, the largest number of units (7600) is recommended for the poorest residents. These would surely be the hardest to scatter around the city, and would not contribute to the theme of "economic mobility", at least not "earned economic mobility".
There also needs to be some very clear and convincing reason why a twenty year comprehensive plan (as opposed to a ten-year stop-gap effort) should have an implicit goal of supporting the a larger number of poon than the city is struggling with now. Under this housing plan, very low income units would increase from 18.2% to 21.4% of the city's housing stock. The chart below shows the housing add-ons proposed for DC over the next 20 years, compared to the current baseline. It is difficult to see what has really been gained by spreading these housing units so evenly across the entire housing spectrum. More diverse? More inclusive? More vibrant? More of the same?
Where is the analysis of supply and demand?
Both plans seem to deal in excessive detail with manipulating the "supply" side of the housing equation, but completely ignore the "demand side". However there seems to be little or no indication in the CHSTFR, or in the Comprehensive Plan for that matter, as to just who wants to leave the District, and who is trying to get in. It is somewhat surprising (once one thinks of it) that there seem to be no readily available polls of buyers or sellers. Thousands of householders come and go every year, and NARPAC has found nothing but occasional anecdotal information from loquacious real estate agents about the motivations of either the leavers or the arrivers.
A recent Post article (12/15/05) provides an inkling from Census Data (though NARPAC has been unable to confirm it or embellish on it!) indicating the population change between 1995 and 2000. The chart below is drawn from that Post data and reconfigures it to more easily show the trends. Bars to the right show in-migration and to the left, out-migration for six different (clearly overlapping) groups. The white stubs indicate the net loss for a total net reduction of some 45,000 people. Oddly enough, according to these data, as many poor people came into the city as left; a significant net number of blacks left town; even more Hispanics left than arrived, almost all of those leaving took kids with them; and there was a significant net reduction in homeowners. Of those who came in, about a third was moving from "within the region", the rest from further away, although this seems to be based on a very fuzzy distinction. In any event, DC apparently lost out in share of homeowners, and gained in share of the poor. Is this a trend the city wishes to continue?
Surely it should be possible to develop more complete and reliable data than shown above. Any of several DC agencies, and many commercial concerns should be interested in who wants to live in the nation's capital and who doesn't.
No indication of the composition of those households
There are three very distinct categories of housing demand at the lower end of the income spectrum. The largest is for single female heads of households with one or more kids. The second is for non-family (generally solo) householders. The third is for poor couples with kids. From a numerical standpoint, these three household types have very different sizes. For instance, 7600 single mom households would contain 28,900 persons, including 21,300 kids; 7600 non- family households would have only 10,600 people with less than 3000 kids; and 7600 poor married couples would include 15,200 adults and 15,200 kids. A mix of one-third each would have an average family size of 3.5, not 2.1 as assumed.
In addition, each type of household would require a very different housing unit with a generally not-interchangeable layout. Each also presents a different cost for city human support services and schooling. These costs are important because the other 34,000 units proposed in the CHSTFR will have to provide offsetting net revenues.
No indication of why there are 55,000 units for 100,000 people
Never afraid to quibble over arithmetic, NARPAC must note the seeming inconsistency of a fixed 2.12 persons-per-household and the need for 55,000 units to house 100,000. 47,200 should do the trick unless a 16.5% vacancy rate is assumed, considerably above DC's nominal 9.6%. If the 19,000 low affordability units is fixed, then the number of market rate units could be substantially lower than assumed, down, in fact, from perhaps 36,000 to 25,000 units.
As noted in the prior section, the 2.12 household size number also fails to acknowledge the tragic fact that the poorer households are generally larger than the wealthier ones. At the poverty end, family household size could easily reach 3.5, whereas it may be only 2.0 at the upper end of the spectrum. Again, the plan needs to be more specific about which part of the underprivileged spectrum the city plans to benefit.
No acknowledgment of shifting trends with time
Unlike a near-term solution, a longer range plan should take into account recent trends as possible portents of the future. From 1970 to 2000, the total number of households in DC was dropping slowly, reversing only slightly in the last four years. It should be noted that those off-year estimates between each major national decennial census often change significantly after the next full count is tallied. Clearly, the increase in households is to be a departure from the past 35 years.
More pronounced however, as shown on the left chart below, is the 30-year trend toward fewer married couples, more non-family householders and a remarkably constant number of single moms. The lack of shift between 2000 and 2004 is in part a technique of off-year projections.
The same is true to a lesser degree concerning the gradual drop in household size (above right) over the past 30 years. Again the 2004 bar may be an illusion, and the 20-year projection is also a departure from the past, albeit a moderate one.
Questions about the "typical" costs of housing as a share of total income
The direct linkage between the candidates for affordable housing and the value of the units they can afford is set by the factor dictating what share of income can reasonably be put into "housing costs". The task force study (and the draft Comprehensive Plan as well) appears to have accepted the value of 30% as a sort of well-known national norm, and NARPAC sought to confirm this number. According to Census reports on consumer expenditures, there is a category entitled "housing costs", and the national average is between 25% and 30%. The hooker is that this category includes not only the costs of "shelter" (i.e., the housing unit itself) which run close to 20%, but two other housing costs as well: utilities, and operations and furnishings.
The "stack charts" below left show the distribution of every consumer spending component, and also show how they vary between "all" (primarily white on the national scale), blacks and Hispanics. The upper chart to the right shows the trends in "shelter" vs "other housing costs" over the decade (shelters costs are rising slowly). The lower right chart shows that there appears to be some dip in shelter costs and a gradual decline in "other" as income bracket rises.
Possibly more interesting, at least to those unfamiliar with national consumer spending patterns, the yellow bands on the stack charts (right above the housing costs) represent typical American spending on "transportation", almost all of which is for owned vehicles (not public transportation). It does suggest, however, that for urban residents, there may be some realistic trade-offs between housing costs and (foregone) vehicle costs.
The next chart displays some other aspects of housing costs. To the left, there is a significant difference in renters vs owners between the relatively low income DC householders, and the unusually affluent Monty householders.
The chart to the right above indicates what percent of house owners and renters devote what share of their income to housing. Despite the large differences in economic status between DC and its neighboring counties, the proportions are remarkably similar. Nearly 20% of owners devote more than 35% of their income to housing, and almost 30% of renters do likewise. Perhaps more surprising is that about 30% of all home owners spend less than 20% of their income on housing. A demonstrable explanation is not available, but it is possible that a) as income increases, householders do not feel the need to aggrandize their homes proportionately; b) a few home buyers may have chosen not to take out a mortgage; and c) a good many may have actually paid them off. In fact, more DC householders (22.6%) have no mortgage than in either PG County (15.4%) or Monty (20.1%). Clearly a larger share of DC's home owners are long- term residents than in the more rapidly growing, younger, suburbs.
No consistent definition of "middle class", vice "extremely low", "very low, "low", and "moderate" income
The Housing Element in the Comprehensive Plan draft removes a bit of the mystery surrounding the four low-end income definitions by relating them to both "areawide median income" and the availability of various major (and DC) housing assistance programs. The four categories equate to 30%, 50%, 80% and 120% ("moderate income") of AMI. Although there are doubtless good reasons for making poverty standards constant across metro areas, it also places limits on their usefulness in particular local situations. The difference between AMI and 'mean household income (MHI)' will be treated in the following section. Remaining unanswered is the definition of the term "middle class" with its supposedly "middle income". If moderate income is 20% above median income, and middle is above moderate, middle could well turn out to be higher than average. The confusion continues.
The dubious use of "area median income" and "median household income"
For analysts, even well-meaning analysts, the loose terminology concerning income and/or class categories is somewhat frustrating. It can also be somewhat deceptive due to the large disparities in household composition. Additional distortions can be added by using mean values vs median values as NARPAC has belabored elsewhere, But the "AMI", or Areawide Median Income, for a family of four, commonly used by the government as a metric for determining justification for various regional housing subsidies, has very little relevance for the nation's capital city.
In the first place, those median household income (MHI) values are starkly different for the entire gamut of household demographics. This is shown of the lefthand side of the graphic below, indicating the relative income for DC, PG, and Monty family and non-family households of various types. There are three categories in which median incomes across the three jurisdictions vary by less than about 10%: these include white families, married families, and non-family households. But the variation is far greater for non-whites (blacks and Asians shown); for families with kids; and for single householders Many of those differences are due to the shockingly different numbers of individuals below the poverty line in each and every age category:
But these differences are even further skewed by the differences in household types, which have a major impact on the demand for housing types, and doubtless on whether the householders will own or rent. This is shown below, where DC has a very different share of families to "loners", and among married vs single householders (an ever-recurrent theme). Is any one the more inclusive or the more diverse? When the various household types are multiplied by their differing income levels, the "household productivity" proportions are very different indeed. This is shown on the chart below.
DC's demography is characterized by fewer families, many fewer married families, and consequently a higher fraction of non-household units (over 50%, actually). As a consequence, DC's "household productivity" is generated as much by "loners" as by families. More important from the standpoint of this particular issue, married couples account for a significantly smaller share of the total and are therefore less representative of "the norm". This difference also results in a more exaggerated difference in "Mean Household Productivity", with DC at $47,600; PG at $60,700; and Monty at $76,900. The merits of using composite area-wide housing "norms" are cast into considerable doubt with this much spread so close together geographically.
Why be so vague about the low-end household income levels DC wants to accommodate?
Of all the quantitative data needed to make the case for affordable housing in DC, the average wages of its "workforce" should be paramount. It is generally defined to focus on government jobs essential to the city's quality of life, and those data are available each year in DC's Operating Budget. The chart below depicts the average pay for eight major district departments. Together they comprise 24,600 employees of the city's 33,400 (!) total. Pay, overtime, and fringe benefits are included in each bar, though the base pay is clearly the prime indicator. They are ranked in the order of declining average pay, with police and fire/rescue near or above the $50K mark, public schools over $40K (including all staff, not just teachers), and Parks and Recreation personnel down closer to $30K.
The background bands show the various levels of household income used to determine eligibility for certain types of housing assistance as percentages of AMI (discussed above). Park personnel pay rings in at just above "extremely low" household income, while the average pay of public school personnel just makes it over the "very low" household income upper limit. By these definitions, none of the major DC operating departments provide average worker pay above "low" household income. On the other hand, four of the eight DC departments make it above the DC median household income for all DC residents (shown by the red line). Furthermore, all but the lowest-paying department, rise above the "moderate" household income waterline for two similarly-paid earners. The beginning of truth is to understand the difference between individual pay and total household income.
Surely DC cannot afford to set a goal to individually house all lower pay, often entry-level, workers. The entire US has turned into a two-earner society. In the past, when moms didn't work, their older kids contributed to household income. Throughout the 20th Century, individuals just entering the job market either continued to live at home, or lived together in small group housing units, now redefined as "co-housing" arrangements. And since the demise of the "factory town"in the Nineteenth Century, those at the lower end of the pay scales have had to scratch to find affordable housing, often some distance from their places of work. The American Dream of a cottage with a white picket fence may seem dated to a new generation, but it was a powerful motivator to get a good education, a good job, and a good work ethic. Turning housing from a proud reward into a social/human right (like health care?) bears little relationship to capitalism or a free market economy.
Nevertheless, it does seem to NARPAC that a good case can be made for providing some sort of housing assistance for some of these dedicated public employees. But it also seems to us that there are better solutions than to: underwrite the housing itself; define eligibility itself; provide the funds for housing assistance, and last but not least, provide the space on which the affordable housing is located. Such alternatives could include:
o providing a cash voucher restricted to use for housing allowance;
NARPAC remains committed to its earlier suggestion tion to provide affordable housing units on available school property for school, police, or emergency workers in return for some household participation in school (and co-located adult education) activities, as well as some school participation in housing unit maintenance. Hundreds of acres could become available if such facilities were combined with underground relocation of current school surface parking lots.
Why set a 20-year goal to keep home ownership well below 50%?
Somewhat associated with the foregoing homily (diatribe?) is the capitalist concept of owning real estate as the means to prosperity. For years, kids have been taught in school that "90% of the fortunes in the US have been made in real estate". There is no greater facilitator for upward mobility than improving the value of one's home, and/or being able to "move to better neighborhoods". It is also a prime mover in raising property values by resisting and overcoming neighborhood blight and indifference. It is in direct contradiction to current notions of enforced "diversity, balance, and inclusiveness" which appear to devalue self-effort and self- improvement.
Both the CHSTFR and now the Comprehensive Plan set a twenty-year goal to maintain the ratio of homeowners and renters at roughly 45%, at least a third lower than the current national average, which is improving every year. There has always been a general relationship between mean household income (MHI) and home ownership, and NARPAC has dealt with the issue before. This time, we have looked at current data on home ownership in each of Maryland's counties, and compared them to DC (bright red marker) on the chart below:
While there is significant spread in the data, using either median or mean household income, the trend is nonetheless apparent, and DC lies well off the curve. To NARPAC, discouraging, or at least not encouraging, home ownership seems almost "unAmerican" (!), and for the nation's capital city to plan to stay below the American norms is simply inappropriate.
Go back to:
or Go forward to:
Any serious reader who wishes an (informal) hard copy of this text and/or charts can provide a mailing address by using the "feedback" feature immediately below. Please signify "all" or a specific chapter. Expect some modest delay, please.
Did you find this of interest?
Please give us your FEEDBACK
This page was updated on Feb 5, 2006
© copyright 2007 NARPAC, Inc. All rights reserved