Three Demographic Waves: Exploring The Impacts on House Prices and

Document Sample
Three Demographic Waves: Exploring The Impacts on House Prices and Powered By Docstoc
					Three Demographic Waves:
Exploring The Impacts on House Prices and Rents in Los

John Pitkin
     Analysis and Forecasting, Inc., Cambridge, MA

Prepared for presentation at
The Lusk Center for Real Estate
University of Southern California

September 17, 2004

The author gratefully acknowledges the support of the
California Population Futures project for some of the work
presented here, as well as the able research assistance of
SeongHee Min and Sung Ho Ryu of the USC Population Dynamics
Research Group and the California Housing Futures project.

The author’s email is

3 Demographic Waves, J. Pitkin                                1
A wave of immigration arrived in the Los Angeles region in
the 1970s and had significantly subsided by 1991. This
episode was accompanied by two related, contemporaneous
episodes, a long boom in births that peaked between 1989 and
1996 and a wave of domestic out-migration between 1991 and
1995. Together, these events have profoundly reshaped the
population of Los Angeles and the outlook for its future.
The implications for the housing market, house prices, and
rents are sure to be substantial, but so far poorly

Three Population Waves

1. Immigration

The release of detailed data from the 2000 Census has
afforded new and more precise measures of the size and
duration of the surge of immigration. It rose rapidly in
the late 1970s, reached a peak about 1980 and peaked again
about 1990, and then fell off sharply by 1991. Although Los
Angeles continues to receive large numbers of immigrants,
the annual inflow is down by almost a fourth from the 1985
to 1991 peak, and its impact is now increasingly offset by
out-migration of foreign-born population to other parts of
the United States.

The evidence of post-1991 decline in immigration to the Los
Angeles region comes from decennial census data on the
foreign-born population.

Net 10-year growth of foreign-born population fell. The
foreign-born population reported in the 2000 Census in the
Los Angeles CMSA was 1.09 million more than in 1990. This
increase was 40 percent less than the 1980 to 1990 increase
of 1.81 million and below even the 1.15 million increase
between 1970 and 1980.

Adjusted for changes in census coverage, net 1990-2000
growth fell by half. As the result of fewer persons missed
and an increase in persons who were double-counted, the net
undercount of the population is estimated to have been
substantially smaller in the 2000 Census than in 1990.

There is no an authoritative estimate of the improvement in
coverage of the foreign-born population in Southern
California. However, an increase of at least 5 percent is
consistent with estimates for other, comparable population

3 Demographic Waves, J. Pitkin                                2
groups. This would imply that the actual increase in the
(full count) foreign-born population was not the reported
increase of 1.09 million but closer to .90 million.i

Reported new arrivers fell 23 percent between late 1980s and
late 1990s. The net increase of the foreign-born population
is at best an approximate indication of the number of
immigrants who arrive in any particular interval of time.
For our purposes,
the most reliable         New Im m igrants by Year of Arrival in U.S., L.A. CMSA
source of data on
the annual number          400

                                 Estimated Immigrants Arriving in Year
of new immigrants
is the decennial
census count of the        300

numbers of foreign-
born who residents         200
that say they came
to stay in the
United States in a
particular                  50
     ii iii                                                     5-year average
year. ,     The              0
first chart                  1960        1970        1980         1990        2000
displays the data
on foreign-born
persons by year of arrival in the censuses from 1970 to
2000. The observations for each decade are based on the
counts in the following census and allocated to years within
the decade according to more precise information on exact
year of arrival from the 2000 Census, with an adjustment for
an assumed level of emigration between period of arrival and
the end of the decade.

Note the clear heaping arrivals at the end of decade and
mid-decade arrival years. For this reason, a five-year
moving annual average number of arrivals is also shown.iv

By comparing the flows for periods exactly ten years apart,
we can eliminate much of the differences in the numbers of
temporary residents, emigration, and any consistent bias in
the reported year of arrival. The number of arrivers in Los
Angeles in 1995-1996 as reported in 2000 was 24.5 percent
down from 1985-1986 reported in 1990, and the number in
1997-2000 was down by 17.1 percent from ten years earlier.
If allowance is made for increased coverage in 2000, the
declines were more than 25 and more than 20 percent,

3 Demographic Waves, J. Pitkin                                                       3
L.A.’s Share of U.S. Immigration Down by Half Since 1990.
The wave of immigration to Los Angeles coincided with a
large increase in the share of national immigration that
went to the region after 1968. The share first exceeded 20
percent of the U.S. arrivals in 1971, and then fluctuated
between 20 and 25 percent for 19 years, with local peaks in
1973, 1980, 1984, and 1989. Then the region’s share fell
steeply between 1989
and 1992. The rest of       New Im m igrants by Year of Arrival in U.S., L.A. CMSA
                                                Share of U.S.
the 1990s averaged
just over half of the

                                   L.A. CMSA Share of U.S.

These different             0.15
measures, net change
in foreign-born             0.10
population, number of
new arrivals, and           0.05
share of national
immigrant arrivals,
                                pre- 1964 1969 1974 1979 1984 1989 1994 1999
show a similar                  1960
picture. Los Angeles
became a pre-eminent
destination for immigration to the U.S. during the 1970s and
maintained this position through the 1980s. Since 1993 it
is no longer the mecca it once was but remains an important,
though maturing, gateway for immigration.

Domestic out-migration.

For many decades Los Angeles had been a destination for
domestic migration, but by the beginning of the 1970s, the
Los Angeles region was sending more migrants to other parts
of the U.S. than it received. Although there are official
annual estimates of domestic migration, the data for five-
year intervals are more reliable.v Net domestic migration
can be estimated for ten-year intervals between censuses.
When combined with census data on migration during the
second half of each decade, derived from the census question
on place of residence 5 years earlier, the 10-year migration
estimates give residual estimates of net migration for the
first half of each decade.

These census-based estimates show an increase in net
domestic migration from the first half of the 1970s, when
there was an average annual outflow of over 50 thousand

3 Demographic Waves, J. Pitkin                                                       4
people to the rest
                                                                                                                       Net Dom estic Migration, Los Angeles CMSA, 1970-2002
of the U.S., to
the early 1980s,
when there were                                                                                                100000

                                 Net Annual (Average) Dom estic Migration
small net inflows.                                                                                                     50000
Domestic migration                                                                                                         0
then started to
fall in the late
1980s and then                                                                                 -100000                              5-year averages
                                                                                                                                    estim ated from Census
plunged in 1990-                                                                               -150000                              (see text)
1995 to a net loss                                                                             -200000                                                                    California
over 300 thousand                                                                                                                                                         Departm ent
on average per                                                                                                                                                            of Finance
year between 1990                                                                              -300000                                                                    Annual est.
and 1995. This                                                                                 -350000
wave of out-
migration subsided                                                                                   1965 1970 1975 1980 1985 1990 1995 2000 2005
to a net average
annual outflow of
100 thousand
between 1995 and 2000.

According to these measures, the greater Los Angeles region
lost a net of 2.2 million domestic migrants between 1990 and
2000, including 1.7 million between 1990 and 1995 alone.

When we look at the
net migration flows                                                                                                      Net Dom estic Migration, Los Angeles CMSA, 1970-2002
by nativity, we see
a sharp break in                                                                                                        100000
                                                                            Net Annual (Average) Dom estic Migration

the pattern after                                                                                                        50000
1990. Until that
year, almost all of
the variation in                                                                                                         -50000
net migration                                                                                                           -100000
occurred among the                                                                                                      -150000
population, with
significant numbers                                                                                                     -250000
leaving during                                                                                                          -300000
                                                                                                                                            N a t iv e - bo rn

three 5-year                                                                                                                                F o re ign- bo rn
                                                                                                                        -350000             T o tal
intervals, yet
there were                                                                                                              -400000
                                                                                                                              1965 1970 1975 1980 1985 1990 1995 2000 2005
negligible flows of
the foreign-born in                                                                                                                                                Year
any direction
during these periods.                                     After 1990, by contrast, there were

3 Demographic Waves, J. Pitkin                                                                                                                                                          5
substantial net outflows of foreign-born population, and
changes in the level of out-migration similar to the changes
for the native-born out-migration between the early and late

Annual of the California Department of Finance (DoF) offer
more temporal on the timing of these net population flows
after 1991. These show that out-movement of population from
Southern California was greatest between 1993 and 1995.
However, the DoF underestimated the net migration to the
rest of the

Births follow immigration.

The long wave in the number of births in the Los Angeles
region closely paralleled and is clearly related to the wave
of immigration. Birth registration data afford accurate
annual counts of the number of children born in the region.

Starting from the bottom of                                the “Baby Bust,” the birth rate
inched up from 65.5 in 1974                                to 75.9 in 1987. It then rose
much more rapidly to a peak                                of 90.4 in 1992, when 330
thousand children were born                                in the region. Since then, the
rate has steadily declined,                                reaching 75.0 per thousand women
age 15-44 in 2000.

This wave of births was peculiar to the Los Angeles CMSA.
During the period, the birth rate for the U.S. fluctuated in
a much narrower
range between 65.0              Births per 1,000 Women
and 70.9 per 1,000     100.0
                             Births/1000 Women Age 15-44

women. In 1979          90.0
the rate for Los        80.0
Angeles was just        70.0
2.0 percent above       60.0
the national rate.      50.0                             Los
By 1992 it was          40.0
almost a third
higher and in 2000                                       U.S.
had fallen back to       0.0
12.0 percent above         1970 1980        1990    2000
the national rate.
This baby boom can
be directly tied to the wave of immigration that it followed
so closely. Foreign-born, immigrant, women, particularly

3 Demographic Waves, J. Pitkin                                                            6
Latinas, have had substantially higher fertility rates than
native-born Hispanic women. This has been documented at the
state level by Johnson, Hill, and Heim (2001), who report
total (lifetime) fertility rates of 3.2 for foreign-born
Hispanic women and 2.0 for native-born Hispanic women in
1988. By 1992, the rate for foreign-born Latinas had risen
to 4.5, from which level it gradually declined through at
least 1997, when it was 4.0. The native-born rate also
gradually rose between 1988 and at least 1997.vii

As a result, the overall birth rate rose as the foreign-born
share of the regional population rose and surged most
strongly around 1990, when the flow of immigrants included
large numbers of women admitted under the family unification
provisions of IRCA (Immigration Reform and Control Act of

The post-1992 decline in the average birth rate in Los
Angeles is probably related to the slower pace of
immigration but may also reflect some reduction in the rates
of child-bearing among foreign-born women as they have spent
more time in the U.S.

Implications for the Future.

What are the future implications of the demographic waves
that hit the region in the last quarter century?

It appears that the wave of births and the surge of domestic
out-migration were closely related to if not largely driven
by the wave of immigration. The temporal relationship among
the three waves is a piece of the connection, but only a
piece. The link is clearer for fertility.

The Immigration-Fertility Connection. The 1979-1992
increase in average fertility rates in the Los Angeles
region was caused both by the cumulative, at times rapid,
rise in the number of foreign-born women of child-bearing
age and by the rise in the fertility of foreign-born Latinas
in the late 1980s. Even before the latter rise, in 1985,
the fertility rate of foreign-born Latinas in California was
1.2 children per woman greater than for native-born Latinas
and 1.4 above the rate for native-born white, non-Hispanic
women. Thus, the steady increases in the share of foreign-
born women in the total population had the effect of driving

3 Demographic Waves, J. Pitkin                              7
up the overall fertility rate and number of births in the

The resulting increase was compounded by a sharp (more than
1.0 child) rise in the fertility rate for foreign-born
hispanic women that occurred between 1988 and 1991. The
cause of the rise is not fully understood but may be related
to a shift in the characteristics of immigrants who arrived
in the IRCA period (Johnson, Hill, and Heim).

An Immigration-Domestic Migration Connection? By 1990 Los
Angeles had been receiving large numbers of immigrants for
almost two decades, but it was only after that year that it
became a net source of foreign-born migrants within the U.S.
At least five different though mutually compatible
explanations for this break in the pattern of domestic
migration of the foreign-born can be suggested.

   1. Economic recession. The sharp 1991-1994 contraction
      in the regional economy was surely a main cause of the
      substantial net domestic out-migration of the native-
      born population, an estimated 1.1 million between 1990
      and 1995, and it seems reasonable that the foreign-
      born would respond similarly to the same economic
      incentives. However, this begs the question of why
      the foreign-born population had apparently not
      migrated in response to earlier fluctuations in the
      regional economy.

   2. Immigrant network maturation. Migrants tend to locate
      in the same areas where there are earlier migrants
      from the same country or region who can provide job
      referrals, social support, and other assistance to new
      arrivals. It is proposed that these “networks” of
      migrants that were well established in Southern
      California fairly suddenly reached critical mass in
      other receiving states in the early 1990s, drawing not
      only an exodus of earlier migrants from Los Angeles
      but, as we have seen, attracting a greatly increased
      share of newly arriving immigrants.

   3. Reaction to changes in immigration law. The 1987-1990
      surge in immigration to the L.A. region immediately
      followed the 1986 Immigration Reform and Control Act
      (IRCA) and included large numbers of immigrants who
      became legal residents under the provisions of the
      act. The requirements of the legalization process may

3 Demographic Waves, J. Pitkin                              8
       have enhanced the value of legal services and other
       resources that were more readily available in
       California both for new immigrants and earlier
       immigrants, previously undocumented, who were becoming
       “legalized.” Once this process was complete, they
       were then free to leave the Los Angeles gateway for
       more permanent locations. This effect can help to
       explain the timing and suddenness of the change in
       foreign-born domestic migration but not the longer-
       term shift in the pattern of this migration.

   4. Labor market saturation. Low occupational mobility by
      immigrants and inelastic regional demand for workers
      in the occupations dominated by immigrants coupled
      with continued growth of the foreign-born population
      would eventually lead to a sharp fall in wages of
      immigrants in the region. The hypothesis is that the
      labor market “niche” filled by immigrants reached
      capacity by 1990 and that the resulting declines in
      the wages and employment prospects of immigrants were
      large enough to cause foreign-born out-migration to
      the rest of the U.S.

   5. Rising rents. Since the supply of housing is
      inelastic in the short term, unexpected rapid
      population and household growth resulting from
      immigration would cause increases in housing rents.
      Further, urban location theory (Muth-Alonso) indicates
      that the long-run supply curve for urban housing is
      upward sloping. Similar to the labor-market
      saturation hypothesis, the argument is that the
      increases in rents resulting from either the recent,
      1987-1989, surge in immigration, or the cumulative
      amount of immigration drove up the housing cost
      burdens of foreign-born households and caused net
      foreign-born out-migration to the rest of the U.S.

There is no necessary contradiction among any of these
alternative mechanisms. They may in fact be complementary.
For example, Light (2004) has proposed that the out-
migration of the foreign-born from Southern California was
driven by an increase the ratio of rents relative to
immigrants’ wages, or a combination of labor market
saturation and rising rents. Indeed, all of these factors
may have contributed significantly to the recent, new net
migration of foreign-born from the region. To the extent
that the last two factors had an effect, the shift in

3 Demographic Waves, J. Pitkin                               9
domestic migration would be a direct consequence in
immigration to the region. Moreover, this would also imply
a causal relationship between immigration and out-migration
of the native-born population at least since 1990.

Although I will not attempt to resolve the puzzle of
historic causation for the emergence of net out-migration by
immigrants, I will later present some data on rents, house
prices, and housing consumption during the recent past, and
it should come as no surprise that this data are consistent
with the hypothesis of rising rents. Before I do, however,
I first want to describe in a little more detail the
demographic consequences of the three population waves and
implications for the future of the Los Angeles region.

While the immigration wave emerges as central to
understanding the surges in fertility and domestic
migration, all three of these waves will have major
implications for the future.

The New Demographic Context.

As the result of the three population waves, the composition
of the Los Angeles region’s population has shifted, changing
the demographic context for planning and a wide range of
future-oriented decisions. Clearly, the nativity and race-
ethnic mix of the area’s population now includes larger
shares of foreign-born, Latino, and Asian and Pacific Island
populations. In 2000, 30.6 percent of the population was
born outside of the U.S., 40.4 percent were Latino, and 11.0
percent were Asian or Pacific Islander, non-Hispanic.

More significant for many purposes, the age structure of the
region has been transformed by the waves of domestic
immigration and births. In 1980, the cohorts born in the
U.S. between 1946 and 1965, the Baby Boomers, formed the
largest generation in the Los Angeles region, larger than
the preceding Depression era or the following “Baby Bust”
generations. (Figure, top panel, dark-shaded bars.) At
this time, the wave of immigration was just starting to
arrive on the demographic landscape. (Figure, top panel,
light-shaded bars on left.)

By 2000, the Baby-Boom           generation still retained its
prominence relative to           the immediately preceding and
succeeding generations           but was now outnumbered by both the
generation born in the           1985-2000 wave of births and the

3 Demographic Waves, J. Pitkin                                         10
foreign-born population
                                                        Population of Los Angeles Region, 1980, by Age
that had arrived just
                                                                          and Nativity
since 1980. (Figure,
lower panel.) The                                      90
further evolution of                                   80    Foreign-Born                       Native-Born

                                  Cohort Age (Years)
and interaction among                                  70
these three generations                                60
will greatly shape the                                 50

future of the Los                                      40
                                                       30   Pre-1980
Angeles region.                                             Immigration             Baby Boom
The imprint of the wave
of domestic out-
                                                        -1000       -500        0        500        1000      1500
migration on the
                                                                              Cohort Population, Thousands
demographic structure
of the region is less
immediately apparent                                   Population of Los Angeles Region, 2000, by Age
than that of the                                                         and Nativity
immigration and birth                                  90
waves but can be seen                                  80   Foreign-Born                        Native-Born
                                 Cohort Age (Years)

in the general leftward                                70
shift of the “center of                                60
gravity” of the                                        50
regional population                                    40                        Baby Boom
pyramid above age 30.                                  30
These three cohorts,         10
                                            1986-2000 Birth Wave
defined by birth and          0
immigration, will have        -1000 -500   0       500       1000   1500
lasting effects on the                   Cohort Population, Thousands
population composition
of the region. In order to better quantify these effects
and provide a sounder basis for planning and future-oriented
decision, in general, in conjunction with the California
Demographic Futures and California Housing Futures projects
in the SPPD, I have developed projections of the population
of the greater Los Angeles region with detail by nativity,
period of immigration, ethnicity, and race.

The outlook for migration. For these projections, the
critical assumptions that will affect the regional housing
market in the first two decades of the century are those
about immigration and domestic migration. Before I show the
results, I will therefore explain the bases for the
assumptions on which they are based.

3 Demographic Waves, J. Pitkin                                                                                 11
Now that many other areas of the U.S. are well established
as receiving areas for immigration, Los Angeles’s pre-
eminence as a destination for immigrants from large sending
countries is probably forever broken. Since current trends
in immigration policy, driven by heightened concerns about
security and signaled by the consolidation of the
Immigration and Naturalization Service into the Department
of Homeland Security, are in the direction of restriction
and broader enforcement of both existing and new
regulations, large increases in national immigration,
documented and undocumented, are unlikely in the near

For these reasons, the great wave of immigration that flowed
into Los Angeles starting after 1970 and the related surge
in fertility are assumed to be singular events rather than
precursors of similar episodes in the near future. Future
levels of immigration (and fertility) are therefore assumed
to be close to those observed in the late 1990s.

The same assumption is
made for domestic                                          Population of Los Angeles Region, 2020
migration but for slightly                                    (projected), by Age and Nativity
different reasons. In the
second half of the 1990s,                             90
                                                            Foreign-Born                 Native-Born
the effects of the 1991-94                            80
                                 Cohort Age (Years)

regional economic downturn                            70
as well as any “push” from                                          1980-
                                                                            Baby Boom
the high immigration of                                             1999
1988-90 had diminished;
out-migration of the                                                        1986-2000 Birth Wave
                                                      20      2000-
native-born population was                                    2020
near the long-term (1975-                                     Immigration
1995) average; and that of
the foreign-born exceeded                             -1000      -500       0      500      1000         1500
the pre-1990 experience                                           Cohort Population, Thousands
but was similar to the
native-born rate.

These projections are annual and include detail by age, sex,
race, ethnicity, nativity, period of arrival, and immigrant
generation. The projected population pyramid is shown in
the pyramid.

Because of the details they include and the analysis behind
them, these projections create for the first time a rational
basis for modeling the future implications for the greater

3 Demographic Waves, J. Pitkin                                                                      12
Los Angeles housing market, rents, prices, and unit demand,
of the aging of the Baby Boom generation, the maturation and
progress of the 1980-1999 immigration cohorts, and the entry
of the 1986-2000 birth wave into the housing market.


DiPasquale, Denise, and William Wheaton. 1994. Housing
Market Dynamics and the Future of Housing Prices. Journal
of Urban Economics. (35).

Johnson, Hans P., Mary Heim, and Linda Hill. 2001. New
Trends in Newborns: Fertility Rates and Patterns in
California. California Counts, 3(1). Public Policy
Institute of California.

Saiz, Albert. 2003. Immigration and Housing Rents in
American Cities. Federal Reserve Bank of Philadelphia
Working Paper No. 03-12.

End notes:
  Precise estimates are not available for the foreign-born
population in the Los Angeles region. For the state of
California as a whole, the estimated net coverage rate
increased by 2.6 percent (from 97.3 to 99.9 percent) between
1990 and 2000.[U.S. Census Bureau. 2003. Net Undercount and
Undercount Rate for U.S. and States (1990). Document
downloaded from website 7/10/03. U.S. Census
Bureau. 2003. DSSD A.C.E. Revision II Memorandum Series
#PP-60. A.C.E. Revision II. Adjusted Data for States,
Counties, and Places. April 9, 2003. Document downloaded
from website 7/10/03.] However, because the
foreign-born population in Los Angeles is predominantly
Latino, the estimated net 4.3 percent increase in coverage
of the (national) Hispanic population is more indicative.
Because the undercounts for both legal and undocumented
immigrant populations have historically been high and
because the 2000 Census’s targeted efforts to improve
coverage of these populations, we estimate that the coverage
of the foreign-born population in Los Angeles rose by more
than 4.3 percent.[ U.S. Census Bureau. 2003. Net Undercount
and Undercount Rate for U.S. and States (1990). Document
downloaded from website 7/10/03. A.C.E. Revision
II (two-group) U.S. Census Bureau. 2003. Decision on
Intercensal Population Estimates. March 12, 2003. Document
downloaded from website 7/10/03.]

3 Demographic Waves, J. Pitkin                              13
  There are two main sources of data on the annual numbers
of immigrants who enter the U.S. and live in the Los Angeles
region. The first is the Office of Immigration Statistics
(formerly Immigration and Naturalization Service) data on
the numbers of immigrants newly admitted to permanent
resident status and who intend to live in the Los Angeles
CMSA. These data are flawed by the varying and recently
long delays in processing of the administrative records on
which the data are based, by the large number of admittees
who are living in the U.S. before they adjust their status
to that of “legal permanent resident,” and by the omission
of all foreigners living in the U.S. without permission
(undocumented). In view of the very large numbers of
undocumented foreigners known to live in Los Angeles, the
third flaw is fatal for our purposes.
   It should be noted that these data are subject to their
own inaccuracies. In addition to the omission of people not
counted in the census, these data are flawed by their
inclusion of legal temporary residents, such as students and
temporary workers, who are not immigrants, and by errors in
responses to the question on year of arrival in the U.S.
Because of these shortcomings, care must be exercised in
interpreting these data.
  Note also the regular increases in the final years of the
last 3 decades, apparently reflecting the presence of
temporary, non-immigrant, residents who were counted in the
census in following spring.
   Net domestic migration can be estimated for ten-year
intervals between censuses. When combined with census data
on migration during the second half of each decade, derived
from the census question on place of residence 5 years
earlier, the 10-year migration estimates give residual
estimates of net migration for the first half of each
    The bias in the DoF’s estimate may result from an
estimation that is based on records of California drivers
licenses turned in by out-movers to other counties and
states and new registrations by in-movers. It is possible
that this procedure missed a large number of the estimated
more than 900 thousand foreign-born out-overs between 1990
and 2000. The DoF may also have underestimated migration by
children because, according to Census data, the ratio of
(unlicensed) under-age migrants to (licensed) adult migrants

3 Demographic Waves, J. Pitkin                            14
rose between the 1980s and 1990, when the ratio of
unlicensed to licensed migrants was calibrated.
    The foreign/native gap among Asian women was also large but
much more stable, averaging 1.1 children per woman during
the 1982-1997 period covered in the study.

3 Demographic Waves, J. Pitkin                              15

Shared By: