Full text of DOI:10.1080/10511482.2000.9521381

Housing Policy Debate
ISSN: 1051-1482 (Print) 2152-050X (Online) Journal homepage: http://www.tandfonline.com/loi/rhpd20
A preliminary assessment of the new home seller
capital gains law
Thomas Bier , Ivan Maric & Winifred Weizer
To cite this article: Thomas Bier , Ivan Maric & Winifred Weizer (2000) A preliminary
assessment of the new home seller capital gains law, Housing Policy Debate, 11:3, 645-673,
DOI: 10.1080/10511482.2000.9521381
To link to this article: http://dx.doi.org/10.1080/10511482.2000.9521381
Published online: 31 Mar 2010.
Submit your article to this journal
Article views: 16
View related articles
Citing articles: 4 View citing articles
Full Terms & Conditions of access and use can be found at
http://www.tandfonline.com/action/journalInformation?journalCode=rhpd20
Download by: [University of Lethbridge]
Date: 16 June 2016, At: 21:06

Housing Policy Debate
·
Volume 11, Issue 3
645
645
©
Fannie Mae Foundation 2000. All Rights Reserved.
A Preliminary Assessment of the New Home
Seller Capital Gains Law
Thomas Bier, Ivan Maric, and Winifred Weizer
Cleveland State University
Abstract
In August 1997, the capital gains law was changed to enable most sellers to move down
in price without incurring a tax liability. The previous law was critiqued as detrimental
to cities by promoting out-migration to higher-priced homes; it was asserted that the
requirement that sellers buy a home equal to or greater than the value of the one they
sold to defer tax liability obstructed movement down in price. This study asked whether
movement down in price increased in four Ohio cities after the law was changed and
whether movement out of the central cities decreased and movement in from the sub-
urbs increased.
Statistically significant change in movement down was found in only one city. Move-
ment up in price and outward dominated all four areas. The study produced no direct
evidence of the factors that influenced seller-buyer move decisions. The process of
life-course change—younger households moving up to larger, more expensive homes
and older households moving down to smaller, less expensive properties—was proba-
bly the driving force.
Keywords: Housing; Mobility; Tax policy
Introduction
The Taxpayer Relief Act of 1997 contained a provision that altered the
treatment of capital gains on home sales. Before the change, the Inter-
nal Revenue Code Section 1034 specified that taxes on capital gains
resulting from the sale of a homeowner’s principal residence would be
deferred (“rolled over”) if the seller purchased another home of equal
or greater value within two years. At age 55, however, the seller became
eligible for a one-time exemption of up to $125,000 of gains. Any gains
over that amount were taxable.
The new law, Section 121, allows a married couple filing jointly to ex-
clude up to $500,000 of gains and a single person up to $250,000. For
those who exceed those figures and pay a tax, the marginal rate was
reduced from 28 percent to 20 percent. Further, the exclusion is avail-
able not just once, but with unlimited frequency, the only restriction
being that the property must be the seller’s principal residence for at
least two years out of the preceding five. For all practical purposes,
this change eliminated capital gains tax liability for most homeowners.

6
46
Thomas Bier, Ivan Maric, and Winifred Weizer
President Clinton proposed the change during his 1996 reelection
campaign and was criticized in some quarters for simply seeking the
votes of middle-class homeowners. But underlying that political con-
sideration was a matter of possibly serious importance for the core of
urban America: The key feature of the old law was the stipulation that
sellers could defer tax liability only by purchasing another home cost-
ing as much as or more than the one they sold. In effect, the law en-
couraged sellers younger than 55 to move up in price. Moving down
meant a likely tax bill. Theoretically, that stipulation defined the sellers-
buyers who would move into a jurisdiction. For example, if all the homes
in a city were priced lower than all the homes in its suburbs, suburban
sellers could not move into the city and purchase a home unless they
were at least 55 or were willing (or compelled by circumstances) to pay
a capital gains tax on the appreciated value of the home being sold.
Only sellers who met those conditions could be expected to move from
suburbs to city.
Again, theoretically, the old law dictated that movement would be in
accordance with the spatial location of home values. Sellers would move
where homes of equal or greater value were located. Homes priced
less than the home being sold were not an option. If values were ran-
domly distributed across a metropolitan area, movement up in price
would be randomly distributed. If values were distributed such that
they were lowest at or near the center of the area and progressively
higher with distance from the center, then movement up, at all price
levels, would be spatially outward.
To the extent that a central city lacked higher-valued housing in rela-
tion to its suburbs, the old capital gains law disadvantaged the city by
obstructing movement in from the suburbs. To the extent that a metro-
politan area’s spatial distribution of home values was one of increasing
value with distance from the center, the old law promoted movement
out and away from the urban core.
These are all theoretical considerations. It is well established that
people who sell and purchase another home do it for various reasons,
such as schools and concern for safety. Capital gains may have been
inconsequential for most sellers during the 46 years that the law was
in effect, or it may have been consequential enough during the 1950s
and 1960s to weaken cities more than they would otherwise have been
weakened, resulting in urban conditions in the 1970s and 1980s that
confirmed the good sense of moving out or not moving in.
In any case, the consequences of the more than 46 years of out-
migration by middle- and upper-income households in many metropol-
itan areas—consequences ranging from contemporary urban sprawl to
massive concentrations of urban poverty—now seem to be shaping the

The New Home Seller Capital Gains Law
647
perspective that healthy cities require a sizable proportion of residents
who are at least moderately affluent. A growing number of state and
local governments, under the rubrics of “smart growth,” “sustainabil-
ity,” and “livability,” are adopting policies designed, in part, to attract
into cities people who can afford to live in many suburbs. It would be
helpful to know whether the new capital gains law is likely to have
effects consistent with that objective. The study that is the subject of
this article represents a preliminary investigation of possible effects.
Literature review
Little is known about effects of the old law, including its spatial impact,
possibly because it was considered to have had little significance since
most sellers either bought a more expensive home or did not report
their sale/purchase to the Internal Revenue Service (IRS) on Form 2119.
In 1995, there were 3.8 million sales of single-family homes, while 1.5
million taxpayers filed Form 2119. And of those who filed, 13 percent
(
5 percent of all sellers) reported a taxable gain (Auten and Reschovsky
1
998). To the extent that the law did draw attention, it was from econ-
omists interested in its potential to influence homeowners’ decisions
(
for example, by causing an owner to wait until age 55 before selling)
Hoyt and Rosenthal 1990; Newman and Reschovsky 1987).
(
Burman, Wallace, and Weiner (1997) reported that “while several
authors have noted that capital gains taxes distort housing decisions,
there is very limited empirical evidence available to support the theo-
retical claims” (384). Their empirical model, based on data provided
by sellers who filed a Form 2119 in 1993 to report a home sale, used
probit regression equations to quantify elements of a sequential-
choice problem faced by homeowners: move or stay; given move, buy
or rent; given buy, move up in price or down; given move down, how
much housing to purchase. These choices were modeled as a function
of the capital gains tax rate, the tax rate on ordinary income, other
housing costs, and demographic variables. The conclusion was that
the tax distorted the decisions of 2 to 5 percent of sellers.
By contrast, Hazel Morrow-Jones (1997, personal communication) found
in her survey of Columbus, OH, 1995 sellers who purchased another
home (all sellers, not just those who filed Form 2119) that 17.2 percent
of the 80.7 percent who moved up in price would have considered buying
a less expensive home if there were no capital gains tax.
The Ohio Housing Research Network (1994) did an empirical study of
seven Ohio cities to determine whether the old law was related to less
movement down in price and inward toward the city center than might

6
48
Thomas Bier, Ivan Maric, and Winifred Weizer
1
otherwise be expected. It differed substantially from Burman, Wallace,
and Weiner (1997) in data and method. The network analysis used data
on (1) actual moves of sellers-buyers and (2) the supply of housing in
a metropolitan area. The methodology employed in that study will be
fully described because it conveys the theoretical influence that the old
law was having in the housing market and because the same method-
ology was used in the present study.
The data source was computerized public records of changes in prop-
erty ownership (deed transfers). Each record contained the sale price,
the name of the buyer, the name of the seller, the date the deed was
recorded, the address of the property, the census tract, and various
characteristics such as year of construction, size, and number of rooms.2
Subject properties were single-family homes and condominiums. Moves
of sellers-buyers were determined by matching sellers’ names with
buyers’ names across the entire metropolitan area, which included the
central county and all adjacent counties. Although moves within and
between counties were documented, the study focused on moves made
by sellers in the central county.
An estimated 52 percent of all possible matches were made. The basis
for calculating the match rate was the American Housing Survey (AHS)
(
tables 3–10 and 4–10), which gives previous and current tenure of
recent movers. The 1992 survey for the Cleveland area indicated that
4
5.3 percent of those who owned at their previous residence were
again owners (U.S. Bureau of the Census 1994); the 1990 survey for
the Cincinnati area indicated 42.3 percent (U.S. Bureau of the Census
1
992); and the 1991 survey for the Columbus area indicated 34.5 per-
cent (U.S. Bureau of the Census 1993a), but that figure changed to
4
3.5 percent in the 1995 survey (U.S. Bureau of the Census 1997). By
contrast, the 1991 national figure was 42.6 percent (U.S. Bureau of
3
the Census 1993b). Because (1) those figures vary somewhat from
1
The Ohio Housing Research Network was formed in 1989 to conduct research on
housing dynamics in Ohio’s seven major metropolitan areas (Akron, Cincinnati, Cleve-
land, Columbus, Dayton, Toledo, and Youngstown) to inform local and state policy
makers of linkages between public policy and urban conditions. The network consists
of a dozen faculty and staff (including the authors) at eight state universities and
has been supported with state funding through the Ohio Urban University Program.
For a network report, see the special issue of Urban Geography (1998, Volume 19,
issue 8) titled Metropolitan Change: Elasticity, Housing, and Policy in Ohio Cities.
2
The record did not indicate capital gains, if any. But some records did contain the
previous purchase price. However, because under the old law the costs of major home
improvements and of selling the home could be used to offset gains, the simple differ-
ence between purchase and sale prices was probably not the taxable gains.
3
Chicago Title and Trust (1996) in its annual survey of Cleveland home buyers found
that in 1993, 55.7 percent previously owned a home; in 1994, the figure was 58.6 per-
cent; in 1995, 57.1 percent. Those figures include buyers who did not move directly from
ownership to ownership. Clark and Dieleman (1996) report that “only about half of all
owners in the United States who move do so within the homeownership sector” (94).

The New Home Seller Capital Gains Law
649
year to year, (2) previous residence was not necessarily in the metro-
politan area where the purchase occurred, and (3) deed transfers did
not identify all new construction (particularly custom-built homes), a
lower figure, 40 percent, was assumed. Owners selling in the central
county and buying again in the metropolitan area were assumed to
account for approximately 40 percent of all home and condominium
purchases. That percentage times the number of sales in the area was
the possible number of matches.
A match was considered acceptable when a seller’s name on one record
matched a buyer’s name on another and when the sale and purchase
occurred within six months of each other. Six months was used rather
than the two years permitted by the old law to increase the likelihood
of accurate matches. As the months between sale and purchase increase,
common names, such as John Smith, are less likely to match correctly.
Because most sellers were found to purchase within six months, it was
chosen as the cutoff point. Many common names did match because of
the presence of two names (husband and wife) on the records.
The deed transfer records used in the study were for the year 1991. A
question was posed as to whether a single year of data was sufficient
for the analysis, since moves at the beginning and end of the year
could have extended from 1990 and to 1992, respectively. It was pos-
sible to resolve that question, because the researchers had several years
of records for the Cleveland area. By comparing matches for a single
year with those identified by including the six months before and the
six months after that same year, it was found that about 8 percent of
the matches were lost when the before and after periods were exclud-
ed. But the results of the two analyses were the same (e.g., proportion
who moved down in price).
After the sell-buy matches were made, the next step was to determine
whether the sellers in the central county moved farther from or closer
to the center of the city. The census tract location of each property in
the match was determined. Each tract in the central county was as-
signed (on the basis of centroid) to a one-mile-wide ring whose center
of radius was the center of downtown. Moves were then identified in
terms of the rings in which the sale and purchase were located. Sellers
who moved to an adjacent county were considered as having moved
outward. Movers who bought in the same ring (approximately 15 per-
cent of all movers) were designated as having moved outward or inward
in proportion to the sellers in that ring who did change rings (in or out).
Sellers who had been living in the central city were distinguished from
those who had been living in suburbs. In Cincinnati and Columbus,
some suburbs are completely surrounded by the city. In those instances,
suburban sellers were identified as such, and their movement was

6
50
Thomas Bier, Ivan Maric, and Winifred Weizer
located in or out by the same ring structure that encompassed the
city and its more distant suburbs.
It should be noted that the most accurate way to do this analysis is to
locate each property by its coordinates (either latitude/longitude or its
position on a digital parcel map) and by precisely calculating the direc-
tion of the move. At the time, the researchers lacked the capacity to do
that. But because most moves involved distances of more than a mile,
the somewhat rough method of mile-wide rings was effective in docu-
menting the scale of movement in and out. In/out accuracy was also re-
duced slightly by designating moves to adjacent counties simply as
being outward (one objective of the study was to determine the balance
of movement in and out of the central counties). In each metropolitan
area, it is possible to move out of the central county to an adjacent
county and be closer to the center of downtown as a result. In most
areas, only a small proportion of all movers can do that. The possibility
is most apparent in the Cincinnati area. A move from northern Hamil-
ton County, OH, to Covington, KY, just across the river, was categorized
as “outward” in the analysis when in fact it positioned the mover closer
to downtown Cincinnati. Such distortion, however, was minor. Only 2.4
percent of the moves made by Hamilton County sellers were to the three
counties across the river. Nonetheless, the most accurate way to do the
analysis is by means of digital coordinates.
The next step involved determining the theoretical effect, if any, of the
old capital gains law on purchase choices. The analysis addressed the
question of whether the IRS provision that sellers could avoid a tax
liability by purchasing a home of equal or greater value altered the pur-
4
chase choices available to them —choices that were closer in or farther
out. The analysis was based strictly on the location and price of homes
in the region and did not consider other factors, such as schools, that
can influence a decision to move.
The analysis had to take into account the actual geographic location
of homes in a metropolitan area because the closer sellers lived to the
downtown center of the city, the less likely they would move inward,
5
irrespective of the capital gains provision. That is, the closer sellers
lived to downtown, the smaller the percentage of homes located closer
in—thus the smaller the likelihood that they would move inward, all
other things being equal.
4
Purchase of a home costing as much as or more than the home sold (the condition
for the capital gains rollover provision) was defined in this study as “moving up”;
thus “moving down” in price was truly down. Very few sellers bought a home costing
exactly as much as the home they sold.
5
The methodology was devised by Bier and Maric (1994). The Ohio Housing
Research Network modified the method slightly in doing its analysis of the seven
cities.

The New Home Seller Capital Gains Law
651
The ring location of homes was the reference against which the appli-
cation of the capital gains provisions was compared. For each seller,
the analysis determined (1) the percentage of all homes in the multi-
county area that were closer in (example, 8 percent); (2) the percent-
age of all homes costing as much as or more than the seller’s that were
closer in (example, 3 percent); (3) the difference between those two
numbers, which was the theoretical effect of the capital gains provision
(
which, in the example, meant a reduction in the seller’s choice to move
inward by 62.5 percent, irrespective of all other factors that can influence
decisions to move).
In addition to using 1991 deed transfers as the data representing the
area’s supply of homes, the same analysis was done using census data
6
(
U.S. Bureau of the Census 1991), which provided a means of assess-
ing the representativeness of the deed transfers. If the 1991 deed trans-
fers were a good sample of all homes, the results of the two analyses
should have been fairly close. And they were: Using census data, the
proportion of all central-county sellers (the sum of all rings) that
could theoretically move inward was 36.7 percent (i.e., for those sellers,
3
6.7 percent of the area’s supply of homes was closer in); using deed
transfers, it was 37.8 percent. (Those are median figures, the middle
figure of seven percentages. All findings for the seven areas are report-
ed that way.) Because the deed transfers represented the choices that
happened to be available to sellers when they were looking to buy, the
deeds were considered to be more appropriate for the analysis than
census data. Also, census data were limited to owner-occupied, single-
family homes, while the deed transfers included condominiums.
The results were consistent across the seven areas. The median fig-
ures were as follows:
1
. On the basis of the number and location (not price) of homes, 37.8
percent of all sellers in the central county could move inward (if
price were not a consideration and disregarding any other factor
that can influence the purchase decision).
2
. Some 18.4 percent of all sellers actually moved inward. Moves in-
ward occurred at one-half the theoretical rate associated with the
location of homes (18.4 percent versus 37.8 percent).
3
. Of all movers, 80.5 percent complied with the capital gains provi-
sion and purchased a home costing as much as or more than the
home they sold. The uniformity of the proportion moving up in
price (six of the areas ranged from 79.5 to 82.6 percent, while the
6
This material is drawn from Summary Tape File 1, table H–23.

6
52
Thomas Bier, Ivan Maric, and Winifred Weizer
7
seventh was 75.5 percent) suggests the possibility that those who
moved down were mostly ages 55 or older, taking advantage of the
one-time $125,000 exemption. In their analysis of IRS Form 2119
filers for 1993 to 1995, Auten and Reschovsky (1998) found a sub-
stantial shift at age 55 in the proportion of sellers who did not
purchase another home. Thirty-five percent of sellers ages 45 to
5
4 did not purchase, compared with 61 percent of those ages 55 to
9. Crowe and Dubin (1998), however, using AHS data, found no
5
such shift. They also asserted that “any change in housing demand
[
as a result of the new capital gains law] is likely to be small and
take place over a fairly long time period” (Crowe and Dubin 1998,
8
). Further, Auten and Reschovsky (1998) found that compared
with younger sellers, sellers 55 or older who did buy another
home showed little difference in moving down.
4
. Theoretically, if all sellers complied with the capital gains provision
and did not move down in price, 23.1 percent could move inward,
on the basis of location and price. (Using 1990 census data for home
value and location, the figure was 23.3 percent.) Thus, the capital
gains provision had the effect of reducing the choice for sellers to
move inward from 37.8 to 23.1 percent, a reduction of 38.9 percent.
5
. Of those who complied, 15.8 percent moved inward; of those who
did not comply, 36.1 percent moved inward. The rate of movement
in by those who did not comply was 2.3 times greater than the rate
of movement in by those who did comply.
The network concluded that the capital gains provision had the effect
of discouraging sellers from moving down in price and inward toward
the center of the city and recommended in its report that the tax on
moving down be eliminated.
President Clinton’s 1996 proposal to change the law may have precip-
itated thought and investigation with a spatial, urban/suburban per-
spective. The Environmental Law Institute (1998) assessed the influ-
ences of the old law and concluded that it contributed to urban decline
by obstructing movement down in price, which undermined home buy-
ing in central cities where prices were lower than in the suburbs. Using
1
990 census data on 38 metropolitan areas, the institute study com-
pared median housing values for the central city and its metropolitan
statistical area (MSA) with net migration into and out of the central
city (movers who stayed within the MSA). The finding was that the
greater the difference in housing value between the city and the balance
7
Hoyt and Rosenthal (1990), using 1981 AHS data for 15 standard metropolitan sta-
tistical areas across the country, found that of homeowners who moved during the 12
months before the interview and repurchased, 85 percent moved up in price and 15
percent moved down.

The New Home Seller Capital Gains Law
653
of the MSA, the greater the ratio of city out-migrants to in-migrants.
These results suggest that, in metropolitan areas where suburban
“
housing is on average considerably more expensive than city housing,
a greater percentage of movers within the MSA will move out of the
central city than in MSAs where urban and suburban housing are
closer in price” (13).
The institute study recognizes that factors other than tax law, such as
schools and concern with crime, influence decisions to move and that
disparity in home values is a function of demand. Nonetheless,
the perverse nature of the old law [by requiring virtually all home
sellers to keep buying up with each home purchase simply to
avoid a tax] necessarily made the trend toward disparity in hous-
ing prices [between city and suburbs] self-reinforcing, regardless
of how it had been initiated. This is because once demand shifted
to suburbs (whether for tax reasons or other preference reasons),
central city housing prices would fall (or at least rise less quickly
than suburban prices) because of reduced demand. A significant
number of potential buyers (those with accrued gains) would then
be excluded from the market for lower priced homes by the opera-
tion of the old law. (14)
The new law, however, was judged as having the potential to curb the
rate of sprawl by enabling movement down in price (to homes pre-
sumably located in central cities and older inner suburbs) and to pro-
mote urban revitalization. Since under the new law homeowners can
secure their capital gains after two years of occupancy, tax-free profit
can be made by purchasing a property in need of repair, fixing it up,
selling it after two years, and then repeating the process any number
of times, all tax-free (up to the new, higher limits).
The Ohio Housing Research Network analyzed the movement of sell-
ers (using the same ring format as above) in relation to city and sub-
urban prices in the central counties of the seven Ohio metropolitan
areas (Bier and Howe 1998). All seven evidenced the same patterns:
Prices generally increased in an outward direction, 91 percent of city
sellers moved outward, 72 percent of suburban sellers moved outward,
8
0 percent of all movers went up in price, and of those who did, the
median increase was 57 percent.
The network study also examined the relationship between the extent
to which central-city sellers moved out of the city and the extent to
which the city had housing priced above the county average; the hypoth-
esis is that the less higher-priced housing a city has, the more its resi-
dents (who are moving up) will leave. That hypothesis was confirmed;
there was a strong negative correlation between a city’s share of higher-
priced housing and the proportion of sellers who moved out of the city.
Toledo, Columbus, and Akron, OH, which had the largest share of their

6
54
Thomas Bier, Ivan Maric, and Winifred Weizer
county’s higher-priced homes, had the smallest loss of city movers, while
Cleveland and Youngstown, OH, with the smallest share of higher-
priced homes, had the largest loss. The findings were consistent with
those of the Environmental Law Institute (1998).
The Environmental Law Institute and the Ohio Housing Research
Network focused on intraregional movement; Klein (1998) focused on
interregional movement and posited that the old law encouraged over-
investment in housing, inflation in housing prices, and conversion of
farmland into suburban housing, particularly in locales where movers-
in from high-housing-cost regions had large capital gains to protect,
but where the local market did not contain appropriately expensive
properties. Klein’s empirical basis was a case study of movers from
Santa Clara County, CA (a high-cost location) to Boulder County, CO
(
a lower-cost location) from 1990 to 1996. She also compared home
prices in both counties and documented construction activity in
Boulder County.
Through a survey of buyers in a large-lot, luxury development in Boul-
der County, Klein found that virtually all respondents had moved up
in price and that 47.8 percent stated that tax considerations had influ-
enced their purchase. It was a modest survey of 73 households, of which
2
3 responded, but the findings supported her hypotheses. With respect
to home price inflation, Klein reported that the average price in Boul-
der County increased from 49 percent of Santa Clara County’s in 1990
to 80 percent in 1996, indicating that tax-induced move-ups triggered
overinvestment in housing.
With respect to converting farmland into suburban housing,
Throughout the peak of the California migration period, only new
homes in unincorporated Boulder County were of equal or greater
value than Santa Clara homes. Therefore, as predicated, compli-
ance with the rollover rule by new residents from Santa Clara and
other expensive areas would lead to conversion of farmland and
open space into new housing developments. Construction data
support this prediction. (451)
Klein concluded, “the phenomenon of expensive suburbs surrounding
a less expensive city center was not confined to aging, industrial cities.
Rather, the rollover rule acted as a magnet to deter investment from
vibrant cities as well as those in decline” (452).
The above studies that specifically addressed home seller capital gains
suggest that the influence of the law on purchase decisions could have
ranged from slight (Burman, Wallace, and Weiner 1997), to modest
(
Hazel Morrow-Jones 1997, personal communication), to substantial
Klein 1998), based, respectively, on national data, a central county of
(
a major metropolitan area, and a small county that was being affected

The New Home Seller Capital Gains Law
655
by extensive in-migration from higher-priced markets. The Ohio Hous-
ing Research Network (1994) found that 80 percent of repeat buyers in
the central counties of Ohio’s seven largest metropolitan areas moved
up in price and that, theoretically, the law obstructed movement inward
toward the center of the city because of the spatial location of home
values in relation to the location and value of sellers’ homes.
Thus far only the capital gains law has been considered as a factor
that might affect the decisions of sellers-buyers. But a number of fac-
tors typically influence move and purchase decisions. In the context of
other factors, such as personal circumstances of the household, commu-
nity characteristics, economic climate, and other laws and codes, capi-
tal gains may have been inconsequential for most sellers. For example,
the deductibility of mortgage interest on income tax filings, which was
not changed by the Taxpayer Relief Act of 1997, can be a major consid-
eration for many buyers. The amount of mortgage interest can enable
a taxpayer to itemize deductions and claim other deductions (such as
property taxes and charitable contributions) resulting in a total that
is larger than the standard deduction. Without mortgage interest, many
homeowners cannot itemize. That provision encourages buyers to max-
imize the size of their mortgage and the price of their house (Jackson
1985; Vaughan and Vogel 1979).
Income growth enables people to purchase more of what they would
like to have, and in the case of housing and its location (e.g., a new or
newer house, better schools, safer neighborhood), migration is the result
(
Knapp and Graves 1989). Movers usually have a range of types of com-
munities or neighborhoods from which to choose. For example, Varady
1991) defined the spectrum of community choices available to home
(
buyers in Hamilton County (Cincinnati), OH. Through surveys of buy-
ers over a three-year period, he identified clusters of communities that
were distinctive on the basis of buyers’ demographic, housing, and at-
titudinal characteristics. A factor analysis of 20 survey items yielded
seven broad search criteria influencing residential choice: a quest for
suburban attributes, a preference for downtown or near-downtown
living, an attempt to preserve neighborhood ties to friends and relatives,
a quest for efficient local government services, a concern for proximity
to schools and other children’s services, a quest for accessibility to jobs
and entertainment, and an interest in good housing prices.
Life-cycle changes (such as marriage), income and employment changes,
housing attributes (need for four bedrooms), neighborhood needs (near-
er to relatives), and accessibility needs (nearer to services for the elder-
ly) can be reasons for changing residence (Bourne 1981). Indeed, “the
principal reasons people change houses are associated with their hous-
ing needs at different stages in the life cycle” (135). Clark and Dieleman
(
1996) use the term “life course” and relate the sequence of changes
that households experience over time (e.g., marriage, income, and

6
56
Thomas Bier, Ivan Maric, and Winifred Weizer
children) to housing market circumstances. These circumstances, “such
as composition of the local housing stock, mortgage rates, and prices of
various types of housing also influence residential mobility and housing
choices. Thus, life course involves not only individual life events but
also social forces and structures” (22). Essentially, Clark and Dieleman
define a process of movement up in size and price “as households seek
to improve their housing over the life course,” beginning with “simple,
small and inexpensive” units, followed by moves to “larger and more
expensive dwellings” (70) until the latter stage of life when downsiz-
ing occurs. They consider the spatial distribution of housing stock
mainly in terms of aging dwellings in the central city and new units
at the expanding outer edge—implying that moving up means mostly
moving out.
Thus, the reasons that underlie a household’s move from one house in
a particular place to another house in a different place can involve a
range of factors, with the most significant of them associated with the
household’s position in its life course. In that context, the old capital
gains law was probably not a factor in the sell, buy, or rent decisions
of most homeowners during the past 20 or 30 years. And if it was a
factor, it was probably associated with circumstances such as the owner’s
age (the 55 or older exemption) or an interregional move that involved
selling a home priced higher than those available where the seller was
moving (thus “forcing” new construction).
Home seller movement under the new capital gains law
The only way to know the reasons that shaped the decisions of home
sellers and buyers is to ask them directly. But metropolitan-scale sur-
veys are expensive, which means that when a survey is done it is usu-
ally limited in scale (e.g., Klein 1998). We lacked the resources to con-
duct a multicity survey after the capital gains law was changed, but
we had deed transfer records for four Ohio metropolitan areas (Cincin-
nati, Cleveland, Columbus, and Dayton) with which moves before and
after the change could be documented. The factors that influenced each
move could not be studied, but the physical evidence of the move could
be. The deed transfer data were for the nine years preceding August
1
997, when the new law went into effect, and the 17 months afterward,
through 1998.
Hypotheses
The data enabled the rate of seller-buyer movement down in price to
be determined, but it provided no evidence as to why the moves were
made. However, during the 17-month postchange phase of the study,

The New Home Seller Capital Gains Law
657
the only macro change that was known to occur and might have af-
fected decisions to move was capital gains. All other influences were
assumed to be essentially random and steady. If the analysis were to
show significant change in the rate of movement down in price, the
capital gains law might have been the precipitating factor.
Hypothesis 1. In the 17 months following the change in the capital
gains law, there was no significant change in the extent to which seller-
buyers moved down in price. This hypothesis was based on the follow-
ing assumptions and considerations:
1
. Not enough time had passed for most homeowners to know about
the change and its possible relevance to them. Newspapers car-
ried articles before and after the change (the Cincinnati Enquirer
had 10, the Columbus Dispatch 4, the Cleveland Plain Dealer 4,
8
and the Dayton Daily News 1 ) and presumably television and
radio stations reported on the Taxpayer Relief Act when it was
passed, possibly mentioning that homeowners were getting a tax
break. But the amount of time the public had to focus on the mat-
ter was not great. Given the amount of publicity, homeowners in
the Cincinnati area should have been the most knowledgeable.
2
. The analysis was limited to intrametropolitan moves. Movers from
higher-priced housing markets (e.g., California) who would be more
prone to moving down in price (as in Klein’s 1998 study) could not
be included in the analysis.
3
. Out-migration dominated movement in three of the cities and was
moderately strong in the fourth. In 1991, 86 percent of Cleveland’s
seller-buyers left the city; Cincinnati lost 74 percent, as did Dayton,
while Columbus lost 54 percent (Bier and Howe 1998). Cleveland
and Dayton in particular have histories of urban discord, racial
tension, undesirable public schools, and lack of middle-income hous-
ing and neighborhoods that could be attractive to people who oth-
erwise would live in the suburbs. Cincinnati was a fairly stable
city into the 1980s but then began to evince decline (Howe et al.
1998). What had been old, middle-class enclaves on hilltops around
the city began to weaken as access to new suburbs in adjacent
counties was provided by new interstate highways and drew away
market demand; problems associated with the city’s unusually
large renter population (two-thirds of all households) grew as the
employment market changed; and the public schools, which had
remained attractive partly because of voluntary busing, were per-
8
These figures come from a Lexis-Nexis search of the four newspapers during the
period between August 1996, when President Clinton first proposed the change, and
December 1998, the end of the study period. Example articles are Harney (1997),
Gray (1998), Budish and Karakul (1998), and Wells (1998).

6
58
Thomas Bier, Ivan Maric, and Winifred Weizer
ceived as declining. In all three cities, the momentum of movement
was strongly outward (and up in price). We did not expect change
in the capital gains law to make a measurable difference in any of
the three.
Columbus, however, is different. As a result of large amounts of
annexed land on which new middle- and upper-income housing
has been built in recent decades, it has neighborhoods that would
be new suburbs in many other metropolitan areas (Aryeetey-Attoh
et al. 1998). Its supply of middle-income housing dwarfs that of
Cleveland and Dayton and is much greater than that of Cincinnati.
In 1991, Columbus had 47 percent of the homes in the county that
were priced above the county’s median value; Cleveland’s share of
higher-valued homes was 4 percent, Dayton’s was 12 percent, and
Cincinnati’s was 24 percent (Bier and Howe 1998). Columbus also
has several revitalized Victorian and turn-of-the-century upscale
neighborhoods that exhibit strong housing markets. And, most sig-
nificantly, the public schools in the newer developments are sepa-
rate from the city system, an arrangement that was established at
the time of the annexation agreements (Jacobs 1998). In effect, the
city “bought” annexation by agreeing to separate schools. Columbus,
therefore, is a place where moving down and possibly closer in can
be seriously considered. Of the four cities, Columbus was the most
likely to show an increase in movement down in price after the
change, although we did not expect to find it.
4
. Life-course dynamics of moving up to larger and more expensive
properties (Clark and Dieleman 1996) dominate the repeat buyer
market. The fact that 80 percent of all sellers-buyers originating
in the central counties of Ohio’s seven major metropolitan areas
were found to move up in price an average of 50 percent (Ohio
Housing Research Network 1994) suggests that as the purchasing
power of owners increases, they act on their preference for a dif-
ferent home and another living environment, which leads to our
second hypothesis.
Hypothesis 2. In the 17 months following the change in the capital
gains law, the rate at which sellers moved farther out from the center
did not decrease and the rate at which suburban sellers moved into
the city did not increase. With the repeat buyer market dominated by
movement up in price and with home values generally increasing with
distance from the city center (Bier and Howe 1998), movement up is
inescapably linked to moving farther out. Because of the particular
conditions there, Columbus was the one city of the four where move-
ment inward might have increased after the law was changed.
Hypothesis 3. The higher the value of the home or condominium sold,
the greater the rate of movement down in price. If the Clark and Diele-

The New Home Seller Capital Gains Law
659
man (1996) concept is accurate (i.e., households move up in size and
price until the latter stage of life when downsizing occurs), then move-
ment down should be related to price since older people (those in or
near the latter stage of life) are more likely to own higher-priced
properties.
Research design
The same deed transfer name-matching procedure described in the
literature review and the same mile-wide ring designation were used
in our analysis (Bier and Maric 1994; Ohio Housing Research Network
1
994). Match rates averaged approximately 42 percent of the estimated
9
number of possible matches. The annual number of matches averaged
1
0,100 for Cleveland, 6,100 for Columbus, 6,900 for Cincinnati, and
,500 for Dayton. (These figures are metropolitan area totals, i.e., all
6
10
matches within and between all counties. ) The number of matches is
not necessarily proportional to the size of the central city (Cincinnati
is larger than Dayton) because of the counties included in each area.
The Cincinnati and Dayton central counties have a common adjacent
county between them, which was included in both areas’ matches.
To address hypothesis 1. Did a statistically significant proportion of
sellers move down in price after the law was changed compared with
before? The study period after the law changed was from August 1997
through December 1998; data on moves before the change were for the
preceding nine years. For purposes of the analysis, the preceding years
were divided, moving backward, into one 19-month period and five 18-
month periods. In that way, each period included, as the postchange
period did, one calendar year, which was important because home sales
in Ohio are heavily seasonal. Thus, the six prechange periods approxi-
mated the 17-month postchange period. Moves originating within each
period were grouped, resulting in six before-change data points, and
one after-change data point. Because economic conditions were fairly
constant and favorable for home buying across the entire pre- and post-
change periods, except for the relatively mild recession of 1990 to 1991
9
The match rate was lower than in the previous study (Ohio Housing Research Net-
work 1994) because the most conservative comparison of names was used. For example,
in the previous study the names “Sara B. Jenkins” and “Sara Jenkins” were an accept-
able match. But in this study, only exact matches were used. Also, it should be noted
that the shift from owner to renter may have increased as a result of the capital gains
change; if it did, it would have reduced the universe of potential matches. In each of
the four central counties, matches as a percentage of sales did decline slightly in the
postchange period, possibly because more owners became renters.
10
Counties included in each metropolitan area, in addition to the central county,
were those that touched the central county. In the case of Cincinnati, three counties
across the Ohio River in Kentucky were included.

6
60
Thomas Bier, Ivan Maric, and Winifred Weizer
(
which fell mainly within one of the six periods), no other basis for
dividing the nine years was considered necessary.
For each of the four metropolitan areas, the following were determined:
the percentage of sellers in each of the seven time periods who moved
down in price and who had been living in (a) the central city, (b) the
suburbs of the central county (in which the city is located), and (c) the
multicounty metropolitan area (i.e., the total of all those who sold some-
where in the multicounty area and purchased somewhere in that area).
An interrupted time-series technique was used for the analysis (Posavac
and Carey 1997). Time-series designs are used in program evaluation
when a definite intervention (change in the law) occurred at a specific
time (August 1997). The technique determines whether the interven-
tion coincided with a change in the underlying process that was occur-
ring (rate of movement down in price), which is defined by a trend line
through the six prechange data points. The postchange data points are
analyzed for significant variation from the trend. Our single postchange
data point was analyzed for variation at the 0.05 level. This technique
was appropriate, given that the analysis was limited to one variable
and time-series data (the physical evidence of seller-buyer moves) and
that all influences on move decisions, other than the change in the cap-
ital gains law, were assumed to be random in their influence on move
direction (up or down in price) and constant over the study period.
To address hypothesis 2. Did a statistically significant change occur in
the proportion of sellers who moved farther out from the city center
and in the proportion of suburban sellers who moved into the city?
The following were determined for both sellers who moved down in
price and sellers who moved up: first, the proportion of city sellers
who (a) moved farther out and (b) moved out of the city; second, the
proportion of suburban sellers (those who sold in the suburbs of the
central county) who (a) moved farther out and (b) moved into the city.
The same statistical test as for hypothesis 1 was performed on all
seven time periods.
To address hypothesis 3. Was the rate of movement down in price re-
lated to the value of the property sold? The concern here was not with
change before and after capital gains, but simply with the relation-
ship between sale price and movement down. For each metropolitan
area, all moves during the entire study period (July 1988 through
December 1998) were grouped according to the price of the home or
condominium sold. Five price ranges were designated: less than $50,000,
between $50,000 and $99,999, between $100,000 and $149,999, be-
tween $150,000 and $199,999, and $200,000 or more. The ranges were
inflation adjusted for each year. The proportion of sellers in each
range who moved down in price was determined. The median percent
change in price (up and down) for all movers was also calculated by

The New Home Seller Capital Gains Law
661
range (for those who sold for less than $50,000, the median percent
price increase for those who moved up and the median percent decrease
for those who moved down were calculated). And finally, because the
deed transfer record contained the year in which the property was
constructed and its size, the median change of all moves in terms of
living area and the age of the home was calculated. According to Clark
and Dieleman (1996), those who moved up would move to larger and
newer properties, while those who moved down would move to smaller,
possibly older properties.
Results
Hypothesis 1 was largely confirmed. At the metropolitan level, three
of the areas—Cincinnati, Cleveland, and Dayton—showed no change
in the extent to which sellers moved down in price (table 1). In all
three, 20 to 22 percent moved down before and after the law changed.
In the Columbus area, however, the proportion that moved down in-
creased from approximately 23 to nearly 26 percent (significant at
the 0.05 level).
Sellers in all of the central cities moved down in price much less than
sellers who had been living in the suburbs of the central county (ap-
proximately 40 percent less), but only in Columbus was there statisti-
cally significant change. In Columbus, 17 percent moved down after
the change compared with 14 to 15 percent before the change. Colum-
bus’s suburban sellers who moved down increased from 28 percent to
nearly 30 percent (statistically significant). City of Columbus movers-
down outnumbered suburban movers-down by 4.7 to 1.
City of Cleveland sellers who moved down increased from less than
1
0 percent to nearly 12 percent, which was almost significant at the
.05 level. Sellers in the central suburbs of Cleveland, Cincinnati, and
0
Dayton showed no change. Thus, in three of the four areas, our hypoth-
esis was confirmed. Columbus was the deviant, a possibility we had
anticipated.
Hypothesis 2 was also confirmed. Virtually no significant change was
found in the rates at which sellers and buyers moved farther out, out
of the cities, or into the cities (table 2, a summary table, contains the
results of all moves during the seven time periods combined; see the
appendix for detailed results). Significant exceptions were city of Cleve-
land sellers who moved up and farther out (the rate increased), and
Columbus suburban sellers who moved up and farther out (the rate
decreased). Generally, across the four areas, there were substantial
differences between those who moved up and those who moved down.
Those who moved up (approximately 78 to 80 percent of all movers, per
table 1) mostly went outward, including 82 to 96 percent of city sellers,

Table 1. Percentage of Sellers-Buyers Who Moved Down in Price
before and after the Capital Gains Law Changed
Before Change
After Change
July 1988 → (Six 18-month Time Periods) → July 1997
August 1997–
December 1998
Area
6
5
4
3
2
1
Columbus
Central city
15.6
25.7
22.8
15.9
24.4
22.5
14.0
24.3
21.2
12.9
24.1
20.5
14.9
25.2
23.0
14.0
27.6
22.7
17.14*
29.94*
25.70*
Central county suburbs
Total metropolitan area
Cincinnati
Central city
17.3
22.2
22.7
17.2
21.1
22.6
17.3
20.9
21.4
16.2
20.1
20.2
15.7
23.3
22.6
14.9
19.7
20.2
13.86
22.33
21.12
Central county suburbs
Total metropolitan area
Dayton
Central city
14.0
25.2
27.3
15.5
23.8
27.0
11.4
24.5
24.5
14.0
23.5
23.7
11.7
21.3
23.4
12.7
19.6
22.4
12.73
20.78
22.76
Central county suburbs
Total metropolitan area
Cleveland
Central city
9.5
20.1
22.0
10.6
18.3
21.5
9.6
19.8
21.5
7.3
18.2
19.7
10.7
19.7
20.9
9.2
20.1
20.8
11.93
19.17
20.44
Central county suburbs
Total metropolitan area
Note: Each column represents the percentage of sellers in each area who moved down in price.
p < 0.05.
*

The New Home Seller Capital Gains Law
663
Table 2. Move Direction of Sellers
Sellers Who Moved Down in Price (%)
Sellers Who Moved Up in Price (%)
a
a
City Sellers
Suburban Sellers
City Sellers
Suburban Sellers
Farther
Out
Out of Farther
Into
the City
Farther Out of
Farther
Out
Into
the City
Area
the City
Out
Out
the City
Columbus
Cincinnati
Dayton
66.8
81.7
78.3
75.6
41.4
58.4
50.9
53.0
59.2
63.2
62.4
68.9
35.5
10.5
7.7
82.0
90.0
96.4
94.2
59.9
74.3
80.7
87.5
74.8
77.1
74.3
84.0
17.9
4.5
2.2
0.9
Cleveland
5.8
Note: This table includes all moves during the study period (July 1988 through December 1998).
a
Suburbs of the central county.
and 75 to 84 percent of suburban sellers. The rate of movement out-
ward by suburban sellers is striking, almost as large as city movers.
Those who moved down moved outward less, moved out of the central
city less, and moved into the city from central county suburbs more.
Movement into the city by suburban sellers did not increase in any of
the areas, including Columbus, the one place where movement down
increased significantly after the law changed. Indeed, suburban move-
down sellers went to Columbus less after the change. However, move-
ment out of Columbus by downward movers declined. Cleveland sellers
moving down showed greater propensity to stay in the city, but Cincin-
nati and Dayton sellers did not. Table 2 indicates the greater capacity
of the city of Columbus, relative to the other three, to retain existing
homeowners and to attract suburbanites, although the rate of subur-
ban sellers moving up and into Columbus shows a decline (see the
appendix). As expected, Cincinnati is in second place, a distant sec-
ond, with respect to retention and attraction. Cincinnati sellers mov-
ing up to the suburbs left the city at increasing rates, and suburban
sellers moved farther out at increasing rates (see the appendix).
Hypothesis 3 was confirmed as well. The higher the price of the home
11
sold, the more sellers moved down (table 3). That relationship extend-
ed from approximately 11 to 14 percent who moved down in the lowest
price range to approximately 55 to 65 percent who moved down in the
highest. The pattern, uniform and distinct across the four metropolitan
areas, is consistent with Clark and Dieleman’s (1996) life course, where-
in older owners move down (assuming that the higher-priced sellers in
our study were generally among the older ones). However, the highest-
11
Not all purchases of new construction could be included in the analysis because the
transfer record did not always contain a clear indication of new construction. Also,
custom-built homes typically begin with land purchase, and a deed recording at that
time, but when the house is built there is no transfer and thus no name match. Since
buyers of new construction are most likely to be sellers in the higher price ranges, the
rates of movement down shown in table 3 for the two highest ranges are probably
somewhat overstated.

6
64
Thomas Bier, Ivan Maric, and Winifred Weizer
Table 3. Percentage of Sellers Who Moved Down, by Range of Price Sold
Price Range
Area
< $50K
$50–$99K $100–$149K $150–$199K
$200K+
Columbus
Moved down
Relative frequency
13.7
3.9
14.6
28.1
22.3
32.8
39.0
17.8
58.1
17.4
Cincinnati
Moved down
11.7
2.8
13.3
29.7
23.0
32.7
38.2
18.2
54.8
16.6
Relative frequency
Dayton
Moved down
Relative frequency
14.2
5.4
16.8
36.1
29.9
33.9
48.2
15.4
64.5
9.2
Cleveland
Moved down
Relative frequency
10.9
3.5
12.8
27.1
22.0
32.7
35.1
19.2
55.1
17.6
Note: This table includes all moves during the study period (July 1988 through December 1998).
Price was adjusted for inflation (base year 1998).
priced move-downs (sale price of $150,000 or more) comprised only
about a third of all those who moved down. Older owners are of course
at all price levels. It is possible that movers-down who were being forced
by personal circumstances (job loss, divorce, and so on) were propor-
tionally overrepresented among the sellers of mid- and lower-priced
properties.
Figure 1 displays the relationship between the price of the home sold
and the price of the home purchased for movers up and movers down.
The data are for the Columbus area, and the other three areas are
similar. For those who moved up, the lower the selling price, the larger
the percentage increase, which was expected. The median move-up was
about 50 percent. Assuming that central-city residents were heavily
represented in the mid-to-lower price ranges, their purchasing power
dictated an outward move, and for 60 percent of them (table 2) their
move was out of the city.
Finally, sellers who moved up typically bought homes that were larger
(
in square feet of living area) and newer than what they sold; sellers
who moved down bought smaller and newer homes, but not as new as
the movers-up (table 4)—a finding that lends additional support to
life course as the primary driver of movement and housing choices.
The median size increase for those who moved up was approximately
4
0 percent (e.g., from 1,000 square feet of living area to 1,400 square
feet). Move-up sellers in the cities of Cincinnati, Cleveland, and Day-
ton bought homes that were on average almost 30 years newer, while
in Columbus, with its large supply of newer homes, move-up sellers

The New Home Seller Capital Gains Law
665
Figure 1. Price Change, by Range of Selling Price,
Columbus, OH, Area from 1988 through 1998
133
1
00
57
50
45
5
0
42
<
$50K
$50–$99K
20
$100–$149K $150–$199K
23
$200K+
Price Range
0
–
–
–27
–29
–
37
–
50
–
100
Note: Includes all moves during the study period (July 1988 through December 1998). Price was
adjusted for inflation (base year 1998).
gained just 8 years. Suburban sellers moved up to homes that were 9
to 17 years newer.
However, those who moved down in price bought homes that were gen-
erally about 85 percent of the size of the home sold. The purchased home
was newer, but not to the same extent as the homes of those who moved
up. City of Cincinnati sellers were the exception; those who moved down
bought homes that were newer and almost comparable to those of
movers-up (25 years versus 30 years).
Table 4. Change in Unit Size and Age
Sellers Who Moved Down
City Sellers Suburban Sellers
Sellers Who Moved Up
City Sellers Suburban Sellers
a
b
a
b
a
b
a
b
Area
AB/AS
AGE
AB/AS
AGE
AB/AS
AGE
AB/AS
AGE
Columbus
Cincinnati
Dayton
0.87
0.82
0.94
0.92
3
25
9
0.81
0.82
0.85
0.84
6
5
1
7
1.44
1.39
1.39
1.30
8
30
27
28
1.43
1.43
1.45
1.47
11
13
9
Cleveland
13
17
Note: This table includes all moves originating in the central county during the study period
(
July 1988 through December 1998).
a
b
AB/AS: Median ratio: living area bought/living area sold.
AGE: Median difference in the age of the home sold and the age of the home purchased:
number of years newer.

6
66
Thomas Bier, Ivan Maric, and Winifred Weizer
Discussion of results
During the decade leading up to the change in the capital gains law,
movement of sellers and buyers was characterized as follows: (1) Ap-
proximately 80 percent moved up in price, and most moved farther out
from the center of the city; (2) 20 percent moved down in price, but
outward less, out of the city less, and into the city from suburbs more
than those who moved up. The study produced no direct evidence as
to why that was the case, but we surmise that life-course change (Clark
and Dieleman 1996) was the driving force. The finding that central-
city homeowners moved up at higher rates than suburban owners is
consistent with that dynamic. During the first 17 months after the
law changed, a significant increase in movement down was found in
Columbus, but not in the Cincinnati, Cleveland, and Dayton areas.
We did not expect to find any change because the move-up portion of
the repeat buyer market is driven by factors contrary to moving down.
What is the likelihood, for example, that a household that wants four
bedrooms instead of its current two will move down in price? The poten-
tial for movement down was greater among the segment of the market
that was inclined to move down in any case: those approaching or in
the latter stages of life. The age of movers in the study was not known,
but if the higher-priced properties were owned disproportionately by
older people, then the finding of a positive relationship between the
price of the home sold and the rate of movement down suggests that
older owners moved down. But if it was the older segment of the mar-
ket that produced the significant change in the Columbus area, why
did the same thing not happen in the other three areas? Indeed, the
homeowner population in the Cleveland area is a bit older than in Co-
lumbus, yet Cleveland’s rate of movement down was lower than Colum-
bus’s, both before and after the change. The Columbus increase may
have been the result of older movers, but more likely not.
Possibly Columbus homeowners knew more about the tax code change
and thus more of them took advantage of the move-down option. But if
newspaper articles on the change were an indication of awareness,
owners in Cincinnati (where 10 newspaper articles on the topic appeared
compared with 4 in Columbus) should have been most active. There is
no reason to think that knowledge of the change in the law was more
widespread in the Columbus area than in the other three.
We posit as a possible explanation of the Columbus results that a com-
plex set of conditions existed there that did not exist in the other three
cities; in combination with the changed law, these resulted in more
movement down in price than otherwise would have occurred (assum-
ing all other influences on decisions to move up or down to be random
and constant during the study period). Columbus, with its large sup-

The New Home Seller Capital Gains Law
667
ply of new and fairly new higher-valued housing, most of it located in
neighborhoods with nearby schools that are independent of the city
school system, had conditions that were relatively more conducive to
moving down. The rate at which suburban sellers moved up and into
Columbus before the law changed suggests attractive conditions (al-
though the rate was dropping before the change and dropped even
more afterward). City sellers continued to move outward after the
change, but fewer left the city. Cleveland and Dayton did not have the
same conditions, and Cincinnati, although historically a fairly stable
and attractive city, appeared to be beginning to slip. In all four loca-
tions, city conditions and the position of sellers in their life course
were probably the major factors affecting moves; capital gains was a
minor factor. But conditions in Columbus were more compatible with
the opportunity created by the new law.
Seller-buyers who may have been strongly influenced by the change
in the law were not part of our study. These include movers into the
Ohio metropolitan areas from higher-priced housing markets (such as
San Francisco, Chicago, and Boston) who had sold at a level well above
what they needed to purchase (Burns 1998). The million-dollar home
sold in San Francisco could be purchased in Columbus for $400,000
or less, but the old capital gains law pressured the mover into buying
a million-dollar home. The new law relieved that pressure, but our
study, limited to intra-area moves, could not measure this impact, if
any. Also, the new law might have triggered more than a little move-
ment down among high-priced owners in places like Manhattan, down-
town Chicago, and San Francisco. The old $125,000 exemption was not
much when an owner (who would rather live in a smaller unit) was
sitting on a million-dollar capital gain. But the new $500,000 exemp-
tion was a different matter; it made moving down a more reasonable
option (Crenshaw 1998; Harney 1999; Muto 1997; Rozhorn 1997).
Also, our study did not include any owners who shifted to renting as a
result of the new law. In light of the fact that most homeowners who
sell do not then purchase another home, possibly the most significant
impact was among that group.
In conclusion, the repeat home-buyer market is dominated by sellers
moving up; we believe that capital gains neither had nor has any in-
fluence on that group. However, under the old law, some owners in
particular situations probably found the law to be an obstacle to their
preferred housing choice, among them owners who wanted to shift
from owning to renting but were not yet 55, owners who were moving
to a housing market where prices were lower than in their current
location, and owners who simply wanted a less expensive unit. Our
study was not able to document any of those situations, but it did
portray the extent and power of the move-up dynamic.

6
68
Thomas Bier, Ivan Maric, and Winifred Weizer
Future research
Capital gains: Movement down
Our study could not determine whether the new capital gains law
actually affected the moving decisions of any homeowners during the
1
7 months following its enactment, including those who moved down
in the Columbus area. In light of the attention being given to the role
that public policy can play in shaping cities, that determination should
be made nationally. Another two or three years may need to pass before
commencing studies so as to increase the likelihood that owners’ knowl-
edge of the new provisions is widespread. Research questions raised
by our study are as follows:
1
. If any homeowners are being influenced by the freedoms granted
by the new law, what are the characteristics of the households
involved?
2. In what ways are owners’ decisions different from what they
otherwise would have been?
3. What are the implications of any differences for communities?
4. Are moves that have been influenced by the new law associated
with particular market or community conditions?
5
. To what extent are people using, or do they intend to use, the pro-
vision that allows the new gains exemption to be taken every two
years, thus enabling them to buy-rehab-sell repeatedly?
With computerized deed transfer records available in many metropol-
itan areas, sellers can readily be surveyed. Moves can be precisely locat-
ed by means of digital coordinates. Cost might be an obstacle, but by
incorporating the questions on capital gains into a survey designed to
illuminate movement patterns, reasons for moving, and product pref-
erences, it is quite possible that local private, government, and not-
for-profit entities would be interested in supporting such a study. A
survey sent to all sellers (i.e., to the address of the home sold) will be
forwarded by the post office to the seller’s new address. In that way,
repeat buyers and sellers who do not purchase again can be reached
as well (Bier 1998). (The post office will forward mail for one year, but
surveys should be sent no more than nine months after the transfer
date because some moves occur up to several months before the
transfer is recorded.)

The New Home Seller Capital Gains Law
669
Life course: Movement up
The more significant area of research touched by the study concerns
the dynamics of movement up in price. If movement is driven primar-
ily by the progression of the life course (and the economics that support
it), then the old capital gains law, which was essentially an obstacle
to moving down, could not have been a factor in most decisions. If life
course is driving 80 percent of repeat sellers to move up, then nothing
short of serious economic disruption (a recession) or personal misfor-
tune (job loss) will affect that. When a recession occurs, the braking
effect will last as long as the recession: Then movement up resumes.
Life-course movement was flowing through the cities in our study, and
most of that flow, central city and suburban, was outward. Research
questions include the following:
1
2
3
. What is the central city’s position in the metropolitan-wide flow?
. What is each suburb’s position?
. Are most cities inescapably limited to being primarily places to
move up and out from and places for moving down?
4
5
6
. If they are, what are the implications?
. Are some suburbs similarly situated?
. How does a city’s position in its life course (defined partly by the
characteristics of its housing stock) match with the typical phases
of people’s personal life course?
7
. What public policies would serve to support movement up and
toward the center of a metropolitan area?
The relationship between life-course movement and the physical
structure of a city may be one of the more fruitful topics of research
and policy analysis at this time.

6
70
Thomas Bier, Ivan Maric, and Winifred Weizer
Appendix
Move Direction of Sellers
Sellers Who Moved Down in Price (%)
Sellers Who Moved Up in Price (%)
a
a
City Sellers
Suburban Sellers
City Sellers
Suburban Sellers
Farther
Out
Out of Farther
Into
the City
Farther Out of
Farther
Out
Into
the City
Area
the City
Out
Out
the City
Columbus
After Change
66.3
61.5
66.2
69.6
69.9
65.4
69.4
36.9
41.5
39.8
44.9
52.1
35.2
38.9
58.1
60.0
58.0
62.5
55.1
55.5
64.3
33.8
41.3
38.9
31.0
33.7
34.4
34.3
79.5
82.7
81.6
81.2
82.5
83.1
83.8
58.2
62.1
61.7
61.6
61.9
56.8
54.4
74.2* 15.2
Before Change 1
Before Change 2
Before Change 3
Before Change 4
Before Change 5
Before Change 6
Cincinnati
76.3
75.5
74.9
75.2
74.0
73.2
15.9
18.5
17.2
17.1
19.9
21.8
After Change
80.2
83.1
79.0
81.9
83.1
85.1
79.7
55.8
56.3
58.5
58.2
63.6
56.6
59.5
61.9
65.5
61.7
62.4
62.4
63.4
64.9
11.6
12.0
11.5
12.7
7.9
8.6
8.2
91.9
89.4
90.5
91.6
88.8
88.5
88.6
78.3
74.6
73.3
75.6
72.7
74.7
69.9
80.6
79.8
79.9
74.5
74.2
75.9
72.7
3.9
4.2
4.5
5.3
4.4
4.4
4.7
Before Change 1
Before Change 2
Before Change 3
Before Change 4
Before Change 5
Before Change 6
Dayton
After Change
73.9
82.7
74.5
89.6
62.2
73.2
85.0
52.5
44.0
49.3
56.7
44.4
46.0
60.8
61.2
62.6
58.7
63.7
64.4
59.0
65.6
5.7
7.6
97.7
97.0
96.3
96.5
96.0
94.8
96.2
82.3
81.6
80.8
79.9
80.4
78.2
81.4
75.1
77.1
76.2
74.2
71.8
71.6
73.4
1.9
2.0
1.8
2.6
2.9
2.4
1.7
Before Change 1
Before Change 2
Before Change 3
Before Change 4
Before Change 5
Before Change 6
Cleveland
9.2
8.0
6.6
10.0
6.8
After Change
64.4
67.1
79.5
82.4
87.3
73.1
80.8
37.6
45.2
51.0
71.4
58.2
60.4
54.8
65.9
67.3
71.2
68.6
71.5
66.2
71.5
6.5
4.9
4.7
6.9
5.4
6.4
6.0
94.8* 85.0
84.7
83.9
83.8
83.2
85.0
83.3
83.7
0.8
1.2
1.1
0.6
0.7
0.9
1.1
Before Change 1
Before Change 2
Before Change 3
Before Change 4
Before Change 5
Before Change 6
93.3
93.7
93.4
94.1
94.7
95.5
84.4
87.0
87.7
90.2
89.6
89.3
Note: “Before Change 1” refers to the 18-month period just before the change in the capital
gains law; “Before Change 2” refers to the 18-month period before that, and so on.
a
Suburbs of the central county.
p < 0.05.
*
Authors
Thomas Bier is Senior Director of the Housing Policy Research Program at the Levin
College of Urban Affairs, Cleveland State University. Ivan Maric is a Research Assis-
tant at the Levin College of Urban Affairs, and Winifred Weizer is a Ph.D. candidate
at the Levin College of Urban Affairs.

The New Home Seller Capital Gains Law
671
References
Aryeetey-Attoh, Samuel, Hazel A. Morrow-Jones, Frank J. Costa, Charles B. Monroe,
and Gail G. Sommers. 1998. Central City Distress in Ohio’s Elastic Cities: Regional
and Local Policy Responses. Urban Geography 19(8):735–56.
Auten, Gerald E., and Andrew Reschovsky. 1998. The New Exclusion for Capital Gains
on Principal Residences. Proceedings of the 90th Conference on Taxation. Washington,
DC: National Tax Association.
Bier, Thomas. 1998. Sellers of Cleveland Homes: 1988–1996. Unpublished paper.
Cleveland State University, Housing Policy Research Program.
Bier, Thomas, and Stephen R. Howe. 1998. Dynamics of Suburbanization in Ohio
Metropolitan Areas. Urban Geography 19(8):695–713.
Bier, Thomas, and Ivan Maric. 1994. IRS Homeseller Provision and Urban Decline.
Journal of Urban Affairs 16(2):141–54.
Bourne, Larry S. 1981. The Geography of Housing. New York: Wiley.
Budish, Armond D., and Kurt Karakul. 1998. Tax Act Gives Better Break on Home
Sales. Cleveland Plain Dealer, March 28, p. 4, Home Section.
Burman, Leonard E., Sally Wallace, and David Weiner. 1997. How Capital Gains
Taxes Distort Homeowners’ Decisions. Proceedings of the 89th Conference on Taxation.
Washington, DC: National Tax Association.
Burns, Nanette. 1998. A Tax Law’s Saving Grace: Some Sellers Profit from the Flexi-
bility of a Generous Capital Gains Exemption. Washington Post, August 15, p. G1.
Chicago Title and Trust. 1996. Who’s Buying Homes in America. Chicago.
Clark, William A. V., and Frans M. Dieleman. 1996. Households and Housing: Choice
and Outcomes in the Housing Market. New Brunswick, NJ: Center for Urban Policy
Research.
Crenshaw, Albert B. 1998. A Change That Makes Home Sales Less Taxing. Washington
Post, August 30, p. H1.
Crowe, David A., and Elliott Dubin. 1998. Taxpayer Relief Act of 1997 and Its Impact
on Housing. Housing Economics, April, pp. 6–9.
Environmental Law Institute. 1998. Linking Tax Law and Sustainable Urban
Development: The Taxpayer Relief Act of 1997. Washington, DC.
Gray, Kathy Lynn. 1998. New Law May Let Homeowners Step Down. Columbus
Dispatch, January 25, p. 31.
Harney, Kenneth R. 1997. Tax Law Benefits Serial Home Sales. Cincinnati Enquirer,
August 24, p. F7.
Harney, Kenneth R. 1999. Baby Boom Relief Act Fuels Home Sales, Urban Revival.
Chicago Sun–Times, January 22, p. 16N.
Howe, Steven R., David Allor, Thomas Bier, Thomas Finnerty, and Phyllis Green.
1
998. The Shrinking Central City Amidst Growing Suburbs: Case Studies of Ohio’s
Inelastic Cities. Urban Geography 19(8):714–34.

6
72
Thomas Bier, Ivan Maric, and Winifred Weizer
Hoyt, William H., and Stuart S. Rosenthal. 1990. Capital Gains Taxation and the
Demand for Owner-Occupied Housing. Review of Economics and Statistics 72:45–54.
Jackson, Kenneth T. 1985. Crabgrass Frontier: The Suburbanization of the United
States. New York: Oxford University Press.
Jacobs, Gregory. 1998. Getting Around Brown: Desegregation, Development, and the
Columbus Public Schools. Columbus, OH: Ohio State University Press.
Klein, Christine A. 1998. A Requiem for the Rollover Rule: Capital Gains, Farmland
Loss, and the Law of Unintended Consequences. Washington and Lee Law Review
55(2):403–68.
Knapp, Thomas A., and Philip E. Graves. 1989. On the Role of Amenities in Models of
Migration and Regional Development. Journal of Regional Science 29(1):71–87.
Metropolitan Change: Elasticity, Housing, and Policy in Ohio Cities. 1998. Urban
Geography. Special Issue. 19(8).
Muto, Sheila. 1997. Tax Break Could Boost Home Sales. Wall Street Journal, March
19, p. CA1.
Newman, Sandra, and James Reschovsky. 1987. An Evaluation of the One-Time
Capital Gains Exclusion for Homeowners. AREUEA Journal 15:704–24.
Ohio Housing Research Network. 1994. The IRS Homeseller Capital Gain Provision:
Contributor to Urban Decline. Unpublished paper.
Posavac, Emil J., and Raymond G. Carey. 1997. Program Evaluation: Methods and
Case Studies. Upper Saddle River, NJ: Prentice Hall.
Rozhorn, Tracie. 1997. New Tax Law Bringing Homes on the Market. New York Times,
October 9, p. F15.
U.S. Bureau of the Census. 1991. 1990 Census of Population and Housing, Summary
Tape File 1. Washington, DC.
U.S. Bureau of the Census. 1992. American Housing Survey for the Cincinnati
Metropolitan Area in 1990. Washington, DC: U.S. Government Printing Office.
U.S. Bureau of the Census. 1993a. American Housing Survey for the Columbus
Metropolitan Area in 1991. Washington, DC: U.S. Government Printing Office.
U.S. Bureau of the Census. 1993b. American Housing Survey for the United States in
1991. Washington, DC: U.S. Government Printing Office.
U.S. Bureau of the Census. 1994. American Housing Survey for the Cleveland
Metropolitan Area in 1992. Washington, DC: U.S. Government Printing Office.
U.S. Bureau of the Census. 1997. American Housing Survey for the Columbus
Metropolitan Area in 1995. Washington, DC: U.S. Government Printing Office.
Varady, David P. 1991. Segmentation of the Home-Buyer Market: A Cincinnati Study.
Urban Affairs Quarterly 26(4):549–67.
Vaughan, Roger J., and Mary E. Vogel. 1979. The Urban Impacts of Federal Policies.
Vol. 4. Population and Residential Location. Santa Monica, CA: RAND Corporation.

The New Home Seller Capital Gains Law
673
Wells, Rob. 1998. New Tax Law Opens Choices to Homeowners. Dayton Daily News,
January 11, p. F11.