Tuesday, April 30, 2013

Testimony: Tom Waters, CSS

Last week, the RGB heard testimony from advocates for apartment tenants, apartment owners, and SRO tenants. (SRO owners, for whatever reason, declined to testify.) Tom Waters, Housing Policy Analyst for the Community Service Society, spoke expertly about the increasing hardships facing tenants, the pressures rising rent burdens put on households, and the regressive nature of the "supplemental increases" that the board has passed in recent years. Below is a transcript of Mr. Waters' testimony, with accompanying charts. For more analysis, check out CSS' new housing report, "Good Place to Live Hard Place to Work", written by Tom Waters and Victor Bach.


Invited Testimony
Tom Waters and Victor Bach
Housing Policy Analysts
Community Service Society of New York
At Public Meeting
New York City Rent Guidelines Board

April 25, 2013

Thank you for the opportunity to present our concerns about the potential impact of this year’s RGB decisions on low-income New Yorkers. Rent-regulated apartments are still the primary source of housing for the city’s 1.1 million low-income[1] households—about two out of five of these families will be affected by the rent guideline increases set by this body. Only 18 percent of the city’s low-income rent stabilized tenants have a Section 8 voucher, leaving more than 336,000 poor and near-poor households without one. The slow recovery of the city’s economy is still not producing enough jobs, with dire consequences for these low-income renters. The unemployment rate remains at 8.9 percent,[2] almost double its level before the financial crisis of 2007.

Tenant Experience in the Private Rental Market, Pre- and Post-Recession

The 2011 New York City Housing and Vacancy Survey and the 2011 American Community Survey, both conducted by the U.S. Bureau of the Census, are still the most recent sources of statistical information on New York City’s rents and incomes. These surveys allow us to draw a picture of New York tenant experience before and after the recession. The city’s continued high unemployment rate suggests that this picture remains largely accurate in 2013.
Chart 2 describes rent and income shifts experienced by tenants in private unassisted rentals over the three most recent HVS studies conducted in 2005, 2008, and 2011. Net increases in median rents over the six-year period vastly outpaced net gains in median incomes for the typical renter. Overall, rents soared to a net gain of 31 percent over the six years, compared to an income gain of only 12 percent. The disparity between the two is a clear indication of rapidly rising rent burdens, the portion of household income that is paid for rent.
The disparity between rent and income trends is evident in both rent-regulated and unregulated apartments. Even under regulated rents, the median contract rent escalated by 26 percent over the six years, more than twice the 12 percent net increase in median renter income. In unregulated units, rents increased by 36 percent against an increase in tenant incomes of 25 percent. Belts tightened for all tenants, regulated and unregulated, as rent escalation took a larger and larger bite out of household income, leaving families with less residual income to cover other non-housing living costs.
The triennial HVS surveys also confirm the dramatic impact of the recession on renter incomes and their ability to keep up with rising rents. Median income increases roughly paralleled rent increases through 2008, a growth period in the local economy, after which there is post-recession fall-off in household incomes while median rents continue to escalate.

Median Contract Rent
Median Household Income

Median Contract Rent
Median Household Income

Median Contract Rent
Median Household Income

The ACS data provide a clearer annual picture of rent and renter income trends from 2005 through 2010, which we compare to city unemployment[3] in Chart 3. Again, the picture is one of persistently rising median rents through the six years against median renter incomes that rose through 2008 and fell off dramatically while the unemployment rate was high from 2009 to 2011.
Unemployment declined from 2005 to 2008 during the city’s upward economic cycle, then spiked in 2009 and 2010 following the recession.[4] Per capita household residual income—the income remaining per member once rent is paid—is arguably the best proxy for rent-income stresses, because it takes into account household size. As a whole New York tenants across the rental sectors experienced a net loss of over 2 percent in residual per capita income (constant 2010 dollars), largely occurring after 2008. Once the recession struck, New York renters, regardless of income, had to tighten their belts to make ends meet and keep up with market rents that persistently increased while rising unemployment and an unfavorable labor market were taking a toll on tenant incomes. 

It can be assumed that the picture for renters in unassisted private apartments, particularly lower-income renters, is bleaker than the ACS data indicate. ACS data do not distinguish among the range of housing types that renters occupy, from public housing to private government-subsidized housing, to private unassisted rentals, both regulated and unregulated. As a result the ACS renter population includes over 300,000 households who live in government-assisted housing, where rents are affordable and based on income. Even so, the ACS data point to a rapid rise in the incidence of high gross rent burdens (rent plus utilities at 50 percent or more of income) to new highs over the 5-year period, with a net 3.7 point post-recession increase between 2008 and 2011. (See Chart 4.)

Low-Income Tenants in the Private Rental Market, Pre- and Post-Recession

               The experience of low-income tenants in the private rental market mirrors that of renters throughout the city. But because their unemployment rates are roughly double those of renters as a whole[5], the impacts on rent-income stresses and residual incomes is much more severe. Chart 5 confirms the extent to which net rent increases over the 6-year period outpaced income gains. Surprisingly, based on the 2005 starting points, the net rent increases in regulated units (29 percent) exceeded those in unregulated apartments (25 percent). By comparison, household incomes increased by from 14 percent.
               Residual per capita incomes declined sharply by 11 percent for low-income renters, but the impact was far more severe in the unregulated market (21 percent decrease) than in the regulated market (7 percent decrease). In short, the dynamics of an escalating local rental market, combined with the post-recession effects on employment and income, have left low-income renters in far worse economic circumstances than they were in before the recession.

Med. Contract Rent
Med. Household Income
Med. PC Res Inc Mo ($2011)[6]

Med. Contract Rent
Med. Household Income
Med. PC Res Inc Mo ($2011)

Med. Contract Rent
Med. Household Income
Med. PC Res Inc Mo ($2011)

               As a result, median rent burdens and the incidence of high rent burdens (rent at 50 percent or more – this time not including utilities) among low-income renters reached a 6-year high as of 2011. (See Chart 6.) Estimates based on the CSS subsample[7] indicate a 4-point increase in median rent burdens in both regulated and unregulated apartments—a rise from 45 to 49 percent in regulated units by 2011, and from 48 to 51 percent in unregulated units. The growing incidence of high rent burdens was even more dramatic. By 2011, a majority (51 percent) of low-income renters in the market were carrying burdens of at least half their incomes, up from 43 percent in 2005. In the regulated stock, high rent burdens rose from 45 to 49 percent of low-income families.

Supplemental rent increases

The New York City Rent Guidelines Board sometimes considers adding extra rent increases for apartments renting for under $1,000. They added such a supplemental increase last year.
This supplemental increase would seem to be based on the assumption that a $1,000 rent is excessively low. But from the point of view of affordability, this is simply not true. In 2010, the median income for rent-stabilized tenants was $38,000. That means that for half of the tenants affected by the RGB’s actions, an affordable rent (30 percent of income) is no more than $950.
The tenants living in the city’s under-$1,000 rent-stabilized apartments are primarily people who cannot afford more. Their median income is $28,000, and half of them have incomes below twice the poverty line (23 percent poor and 29 percent near-poor). More than half of these tenants (54 percent) live in lower-rent areas of the Bronx, Brooklyn, Queens, and Staten Island. Just 10 percent live in Manhattan below Harlem, 13 percent in Upper Manhattan, 8 percent in the gentrifying areas of Brooklyn and Queens adjacent to Manhattan, and 15 percent in higher-rent outer-ring neighborhoods such as Flushing or Bay Ridge. Three quarters of these tenants are people of color: 30 percent black, 38 percent Latino, and 6 percent Asian. Almost half (47 percent) are in households headed by an immigrant.


In light of rent escalation trends that persisted before and since the recession struck the city, it would appear that the private rental industry has not suffered a decline as a result of the economic crisis, certainly not a decline comparable to the losses in income and employment that continue to beset New York renters, particularly low-income tenants.
Certainly the impacts of a global recession and financial crisis on the income and employment of New Yorkers, and their ability to pay for their apartments, are hardly within the control of the city or of the RGB. However, RGB decisions over recent years have contributed significantly to the growing rental affordability crisis at this stage of the presumed recovery. At a time when the RGB should have exercised restraint to ease the impacts of the recession on its struggling resident constituency, it has tended to err on the side of owners by granting larger guideline increases than were necessary.
In recent years, the RGB has overestimated projected operating costs compared to the actual cost increases later reported in owner surveys, resulting in excessively high guidelines. The 2009 Price Index of Operating Costs projected a 4 percent increase, and the RGB adopted a guideline of 3 percent – a little more than enough to cover the projected increase in expenses, given that operating costs consume less than 75 percent of rent.  But the actual increase in costs as measured by Real Property Income and expense statements was only 0.1 percent. In 2010 and 2011, the Price Index again greatly overestimated cost increases. Chart 7 shows the evolution of the Price Index, rent guidelines, and RPIE costs since 1990, clearly showing the great divergence in recent years.  This divergence has resulted in higher, unaffordable rent increases for tenants.
This is an important time for the RGB to reflect on and revise the methodology it uses to determine what it considers reasonable rent guideline increases. Before and since the recession struck the city, it has tended to over-compensate owners to the expense of rent-stabilized tenants. This is certainly a time for the RGB to exercise greater restraint by keeping guideline increases to a minimum, for an industry that appears to be prospering in the wake of recession. RGB should consider its unique potential contribution to stabilizing the rental market and easing the rent-income stresses that New Yorkers continue to experience.


1)    CSS urges the New York City Rent Guidelines Board to take into account the disastrous effects of the recession on many New Yorkers, regardless of income, and exercise overdue restraint in setting new, minimal rent guideline increases. In past years CSS has recommended a rent freeze; we continue to do so in light of the persistent rent escalation that has benefited the industry in the face of a major economic set-back for many tenants.
2)    CSS urges the RGB to refrain from further supplemental increases at lower rent levels—what tenant advocates refer to as the “poor tax.” RGB should do its best to minimize increases at the low-rent levels where the poorest tenants tend to live.

APPENDIX: CSS Renter Sub-Sample

Because of unavoidable inconsistencies and inaccuracies, in respondent reporting of household income and contract rent, this analysis of rent burdens is based on a sub-sample of renter households within each of the HVS samples used. The CSS renter sub-sample for each HVS year was selected on the following basis:
               1) Rent-paying households only (exclude rent-free and owned housing)
               2) Head of household age at least 25 and less than 65.
               3) Households with a positive HVS contract rent burden
               4) Households within the middle 90 percent of the income distribution for renters
(excludes 5-percent outliers at either extreme) . The resulting household income intervals used for each HVS year are as follows:
               2011    $7,896 to $175,000
               2008    $6,912 to $160,000
               2005    $6,006 to $133,000
               2002    $6,000 to $130,000
               1999    $5,700 to $131,000
               1996    $5,000 to $119,950
5) Households within the middle 90 percent of the contract rent distribution for renters (excludes 5-percent outliers at either extreme.) The resulting contract rent distributions used for each HVS year are as follows:
               2011    $342 to $2,800 monthly
               2008    $252 to $2,500 monthly
               2005    $208 to $2,100 monthly
               2002    $200 to $1,900 monthly
               1999    $177 to $1,550 monthly
               1996    $163 to $1,300 monthly
6) Residual (after-rent) household income of at least $100 monthly, in 2002 dollars. For each HVS year, the residual income threshold, in 2002 dollars, was:
               2011    $129
               2008      $123
               2005    $111
               2002    $100
               1999    $93
               1996    $87

The resulting CSS sub-sample can be considered a more "mainstream" sample of New York City renters than the HVS renter sample as a whole. The comparison below of some of the key parameters for each of the two samples suggests that the CSS results are more likely to underestimate rent burdens and related measures of rent-income pressures for the city as a whole.

Comparison: HVS and CSS Renter Samples

                              1996                   1999                   2002                   2005                   2008                   2011

Median Income         
HVS:                    $24,680             $27,600             $32,000             $33,904             $40,000             $40,000
CSS:                     $31,000             $35,000             $39,000             $40,050             $46,400             $50,000

Median Contract Rent                         
HVS:                    $ 600                  $ 648                  $ 706                  $ 850                  $ 950                  $1,100
CSS:                     $ 600                  $ 650                  $ 730                  $ 850                  $ 996                  $1,100

Median Contract Rent Burden       
HVS:                   28 %                   27 %                  27 %                  28 %                  29 %                  31 %
CSS:                    24 %                   23 %                  23 %                  25 %                  25 %                  27 %

Percent Households with High Burdens (50% or more)
HVS:                   26 %                   26 %                  23 %                  26 %                  26 %                  29 %
CSS:                     12 %                   12 %                  12 %                  14 %                  15 %                  18 %


1 Preliminary estimate from the U.S. Department of Labor’s Bureau of Labor Statistics.

We use the term “low-income” to refer to households with incomes no greater than twice the federal poverty threshold, in 2010 about $34,114 for a family of three persons. Household incomes recorded in each HVS are for the previous calendar year. Low-income households include the poor, as well as the “near-poor” with incomes above poverty but no greater than twice the poverty threshold.

Unemployment rates are calculated by the Bureau of Labor Statistics from the Census Bureau’s Current Population Survey.

4 Unemployment figures are for individuals at least 18 years old and below 65.

5 ACS 2005-2010 data estimate renter unemployment rates of 9, 8, and 12% for 2005, 2008, and 2010 respectively, against unemployment rates of 19, 15, and 22 % for low-income renters.

6 Median monthly per capita residual income figures were calculated only for the CSS renter subsample (See Appendix)

7 For description of the CSS subsample, see Appendix.

No comments:

Post a Comment