Upload
others
View
3
Download
0
Embed Size (px)
Citation preview
Housing Finance Policy Center Lunchtime Data Talk The Decline in Geographic Mobility and Implications for the Mortgage Market Sam Khater, CoreLogic Raven Molloy, Board of Governors of the Federal Reserve System September 21, 2016
Declining Migration within the U.S.: the Role of the Labor Market
Raven Molloy, Board of Governors Christopher L. Smith, Board of Governors Abigail Wozniak, University of Notre Dame, IZA, NBER
Disclaimer: Any opinions and conclusions expressed herein are those of the authors and do not indicate concurrence with other members of the research staff of the Federal Reserve or the Board of Governors.
Long-Distance Migration rates within the US are falling
Source: CPS-ASEC
.01
.015
.02
.025
.03
.035
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015Year
Frac
tion
of P
opul
atio
n
Fraction Moved Across States
3
Why do we care?
• Long-distance moves frequently involve a change in one’s local labor market, so there are implications for the labor supply.
• Migration decline could imply that labor markets are less flexible, implying a reduction in aggregate economic efficiency.
• Or, there could be less need to move across labor markets, implying that the aggregate labor market is actually more efficient than it used to be.
4
Why do we care?
• Beyond labor markets, understanding the reason for this decline is essential for determining its implications for wellbeing.
o If the cost of migration has risen, people might not be living in utility-maximizing locations.
o Or the benefits of investing in local social networks might be greater.
5
Why do we care?
• In order to understand the causes of this trend, it would help to know more about the types of people and locations where migration declines have been most pronounced.
6
Agenda for today • Measurement issues
• Show trends for a variety of demographic and socioeconomic groups; and by location.
• Estimate the contribution of observable characteristics to the aggregate trend. o Observables can only explain less than ¼
• Evidence on a number of other common explanations o Empirical support is weak
7
Agenda for today • Draw a connection between the trend in migration and a
concurrent downward trend in labor market transitions.
o The trend in job changing has probably caused the trend in migration.
• We still don’t really understand the causes of these trends, but further work should look to the labor market…
8
Measurement issues Three commonly-used datasets: • Current Population Survey (1948 to present)
o Survey of 60,000 to 100,000 households, each March asks people where they lived 1 year ago
• American Community Survey (2001 to present) o Survey of about 3 million households, asks people where they lived 1
year ago
• Internal Revenue Service (1975 to 2011, 2011 to present) o Tabulated from address changes of tax filers (not a sample) o Disadvantages: Address of the tax filer may not be the residence and
not everyone file taxes
9
Measurement issues
• All show that migration has been falling, although by different amounts.
• Declines seem to have begun in the 1970s or 1980s.
.01
.015
.02
.025
.03
.035
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015Year
CPS ACS IRS
Frac
tion
of P
opul
atio
n
10
Measurement issues
• The level of migration in the CPS is much lower than the ACS. • Odd because the design of the CPS is very similar to ACS.
.01
.015
.02
.025
.03
.035
1950 1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015Year
CPS ACS IRS
Frac
tion
of P
opul
atio
n
11
Measurement issues • Similarity of ACS and IRS leads us to suspect mis-
measurement in the CPS, but there is no easy explanation. o Probably not survey non-response or changes in the survey
instrument/question. o Not related to imputed values.
• We still use the CPS a lot because it provides the longest time-span of individual-level data.
12
Other distances • Migration across shorter distances is falling too.
.07
.08
.09
.1.1
1.1
2.1
3.1
4.1
5
1965 1970 1975 1980 1985 1990 1995 2000 2005 2010 2015Year
CPS ACS
Frac
tion
of P
opul
atio
n
Fraction Moved Within County
13
Demographics and other characteristics • Some types of people tend to move more frequently than
others, so it is natural to expect changes in the composition of the population to affect the trend in aggregate migration. o Older people move less, and their population share is growing o Homeowners move less and (until recently) the homeownership rate
was rising.
14
Demographics and other characteristics • Other research has found only a limited role for demographics
(Cooke 2011; Kaplan and Shulhofer-Wohl, forthcoming; Hyatt and Spletzer 2013).
• Why? Interstate migration has been falling for most demographic and socioeconomic groups.
• It has also been falling in all regions of the US and states.
15
Demographics and other characteristics Interstate Migration by Age
0.0
1.0
2.0
3.0
4.0
5.0
6
1980 1985 1990 1995 2000 2005 2010 2015Year
18-22 23-34 35-44 45-64 65+
Frac
tion
of A
ge G
roup
16
Demographics and other characteristics Interstate Migration by Homeownership
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7
1980 1985 1990 1995 2000 2005 2010 2015Year
Homeowner Renter
Frac
tion
of O
wne
rshi
p G
roup
17
Demographics and other characteristics Interstate Migration by Marital Status
.005
.01
.015
.02
.025
.03
.035
.04
.045
.05
.055
1980 1985 1990 1995 2000 2005 2010 2015Year
Married Single Other
Frac
tion
of M
arita
l Sta
tus
Gro
up
18
Demographics and other characteristics Interstate Migration by Education
.005
.01
.015
.02
.025
.03
.035
.04
.045
.05
.055
1980 1985 1990 1995 2000 2005 2010 2015Year
Less than HS HS Some Coll. Coll. +
Frac
tion
of E
duca
tion
Gro
up
19
Demographics and other characteristics Interstate Migration by Nativity
.005
.01
.015
.02
.025
.03
.035
.04
1995 2000 2005 2010 2015Year
Native Foreign born
Frac
tion
of G
roup
20
Demographics and other characteristics .0
1.0
15.0
2.0
25.0
3.0
35.0
4.0
45
1980 1985 1990 1995 2000 2005 2010 2015Year
Mfg Service Retail FIRE
Frac
tion
of W
orke
rs in
the
Indu
stry
Interstate Migration by Industry
21
Demographics and other characteristics Trend in Interstate Migration by State
NationalAverage
02
46
810
Num
ber o
f Sta
tes
-.002 -.0015 -.001 -.0005 0State-Specific Time Trend
Note. State-specific trends are estimated by regressing the fraction of residents that moved into the state on a time trend. Source: CPS-ASEC.
22
Demographics and other characteristics Trend in Interstate Migration by State
Note. State-specific trends are estimated by regressing the fraction of residents that moved into the state on a time trend. Source: CPS-ASEC.
23
Demographics and other characteristics Trend in Interstate Migration by State: CPS vs IRS
Note. State-specific trends are estimated by regressing the fraction of residents that moved into the state on a time trend. Source: CPS-ASEC and IRS.
Tren
d in
CP
S M
igra
tion
1981
-201
5
Trend in IRS Migration 1976-2011-.0016 -.0014 -.0012 -.001 -.0008 -.0006 -.0004 -.0002 0
-.002
-.0018
-.0016
-.0014
-.0012
-.001
-.0008
-.0006
-.0004
-.0002
0
alabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabamaalabama
arizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizonaarizona
arkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansasarkansas
californcaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforncaliforn
coloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocoloradocolorado
connecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecticonnecti
delawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelawaredelaware
floridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridafloridaflorida
georgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgiageorgia
idahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidahoidaho
illinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinoisillinois
indianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindianaindiana
iowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowaiowa
kansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansaskansas
kentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentuckykentucky
louisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisianlouisian
mainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemainemaine
marylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmarylandmassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachumassachu
michiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichiganmichigan
minnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesotminnesot
mississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississimississi
missourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissourimissouri
montanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontanamontana
nebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraskanebraska
nevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevadanevada
new hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hampnew hamp
new jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jersnew jers
new mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexinew mexi
new yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew yorknew york
north canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth canorth ca
north danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth danorth daohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohioohio
oklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahomaoklahoma
oregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregonoregon
pennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylvpennsylv
rhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode isrhode is
south casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth casouth ca
south dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth dasouth da
tennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennessetennesse
texastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexastexas
utahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutahutah
vermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermontvermont
virginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginiavirginia
washingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingtwashingt
west virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest virwest vir
wisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsiwisconsi
wyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyomingwyoming
24
Demographics and other characteristics • We can formally assess the role of observable characteristics
by regressing a migration indicator on year indicators and observed characteristics.
𝑌𝑌𝑖𝑖𝑖𝑖 = 𝛽𝛽𝑋𝑋𝑖𝑖𝑖𝑖 + ∆𝛿𝛿𝑖𝑖 + 𝜀𝜀𝑖𝑖𝑖𝑖
• The coefficients on the year indicators reveal a counterfactual aggregate migration rate if the characteristics of the population had remained fixed.
• Characteristics: age, homeownership, education, sex, race, martial status, presence of children, real income, employment status, self-employment status, metropolitan status, and Census division.
• Use CPS data 1981-onward.
25
Demographics and other characteristics
-1.5
-1.2
-.9-.6
-.30
.3.6
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011year
No Controls Control for Age and HomeownershipMany Controls
Note. Each lines shows the coefficient estimates on year indicators where the dependent variable is interstate migration. “Many controls” includes age, homeownership, sex education, race, income, marital status, presence of children, employment status, self-employment status, metropolitan status, and Census division. Source: CPS-ASEC.
Interstate Migration
26
Demographics and other characteristics -2
-1.5
-1-.5
0.5
11.
52
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011year
No Controls Control for Age and HomeownershipMany Controls
Within County Migration
Note. Each lines shows the coefficient estimates on year indicators where the dependent variable is within-county migration. “Many controls” includes age, homeownership, sex education, race, income, marital status, presence of children, employment status, self-employment status, metropolitan status, and Census division. Source: CPS-ASEC. 27
Beyond demographics Other common explanations: • Dual-career households • Land use regulation • Telecommuting • Occupational licensing • Health insurance
28
Dual-Career Households • Rise in dual-earner households had ended by the 1980s • Most proxies for dual-career households only find a small fraction of such
households. • Migration fell substantially for other households.
Population Share Interstate Migration Rate
1981-1989 2002-2012 1981-1989 2002-2012 Both spouses employed 30.5 30.1 1.9 0.9 All other 69.5 69.9 3.1 1.9
Both prof/tech occupation 2.1 2.8 3.2 1.4 Other prof/tech occ. 13.5 16.6 3.8 2.2 All other 83.1 79.0 2.6 1.4
Both college or more 4.5 7.9 1.8 1.0 Other college or more 14.2 18.2 2.5 1.6 All other 78.9 70.5 2.9 1.5
Spouses have similar incomes 22.6 28.3 2.5 1.2 All other 77.4 71.7 2.8 1.6 29
Land Use Regulation • Migration has fallen substantially even into states with low house values
and/or relatively lax land use regulation
Population Share Interstate Migration Rate
1981-1989 2002-2012 1981-1989 2002-2012 Median house value
Above average 14.7 14.0 2.2 1.6 Below average 85.3 86.0 2.9 1.7
Land Use Regulation Above average 19.1 16.7 2.0 1.4 Below average 80.9 83.3 3.0 1.7
30
Telecommuting • Advances in mobile technology and communications has made it easier
for people to work remotely, perhaps allowing them remain in the same state even if they start working for a distant employer.
• The fraction of people who telecommute is too small
• In the 2014 ACS, only 4.6 percent of workers reported working at home, up from 2.3 percent in the 1980 Census.
31
Occupational Licensing • The fraction of workers required to hold a state license in order to perform
their job has risen from 5 percent in 1950 to 20 percent in 2008 (Kleiner and Krueger 2013).
• The need to obtain a new license might prevent people from moving across state lines.
32
Occupational Licensing • But the downward trend in migration has not been steeper in states with
more workers that require licenses.
• Other research has found little effect of licensing requirements on mobility (DePasquale and Stange 2016).
South Carolina
Rhode Island
New Hampshire
IndianaKansas
Minnesota
Delaware
GeorgiaVermontMaryland
Colorado
Virginia
OhioWisconsin
District of Columbia
Pennsylvania
Arkansas
MichiganNew York
New Jersey
Maine
California
Alabama
Wyoming
Montana
Missouri
MassachusettsSouth Dakota
North Carolina
Arizona
Louisiana
Idaho
MississippiTennessee
Utah
Texas
Nebraska
IllinoisConnecticut
Oklahoma
Alaska
West Virginia
New Mexico
Oregon
Hawaii
North DakotaKentucky
Florida
Washington
Nevada
Iowa
-.002
5-.0
02-.0
015
-.001
-.000
50
Mig
ratio
n Tr
end
.1 .15 .2 .25 .3 .35Fraction Licensed
Note. Trends are coefficients on a linear time trend where the dependent variable is interstate migration. Fraction licensed is from a 2013 household survey. Source: CPS-ASEC for migration data; Kleiner (2015) for fraction licensed.
33
Health insurance related job lock
Source: CPS ASEC
Population Share Interstate Migration Rate
1981-1989 2002-2012 1981-1989 2002-2012
Employer-provided health insurance in household 64.8 64.3 2.7 1.5
All other 35.2 35.7 3.0 1.8
• The fraction of workers in a household where someone has employer-provided health insurance has not risen.
• Migration declines are similarly-steep for people in households where no one has employer-provided health insurance.
34
Looking toward the labor market
1. Long distance moves most often job related.
2. Labor market transitions are also declining. o This trend is also not related to demographics.
3. Strong cross-state relationship between changes in job
transition rates, migration rates.
35
1. Interstate migration strongly job-related
Source: CPS ASEC
0.0
02.0
04.0
06.0
08.0
1.0
12.0
14.0
16
2000 2002 2004 2006 2008 2010 2012 2014Year
Employment Housing Family Other
Frac
tion
of P
opula
tion
36
2. LM transitions are also declining
Source: CPS ASEC
36
912
15P
erce
nt
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011year
Occupation Change Industry Change Employer Change
37
2. LM transitions are also declining
• Other measures of churning have also fallen o Downtrend in job creation and destruction rates since at least
1990, from BED and JOLTS data (Davis, Faberman, Haltiwanger 2012)
o Downtrend in job creation, job destruction, hiring and separations since 1990s (Hyatt and Spletzer 2013)
• Limited success explaining downtrend
o Aging has modest effect on aggregate job transition rates, can’t explain much of the downtrend (Fallick and Fleischman)
o No role for composition of labor force or employers (Hyatt and Spletzer 2013)
38
2. Decline in job transitions not explained by observables
-3.6
-3-2
.4-1
.8-1
.2-.6
0.6
1.2
1.8
2.4
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011year
No Controls Many Controls
Note. Labor market transition is defined as a change in employer, change in occupation, or change in industry. Source: CPS ASEC 39
3. Cross-state relationship between trends in job transitions, migration
AL
AK
AZ
ARCA
CO
CT
DE
DC
FL
GAHI
ID
IL
IN
IA
KS
KY
LA
ME
MDMA
MI
MNMS
MOMT
NE
NV
NH
NJ
NM
NY
NC
ND
OH
OK
OR
PA
RI
SC
TN
TXUT
VAWA
WV
WI
WY
Change in mobility rate = 0(0) + .25(.03)*(change in fract. changing firms)-.04
-.02
0
Cha
nge
in in
ters
tate
mob
ility
rate
, 198
1/89
to 2
001/
10
-.1 -.075 -.05 -.025 0 .025Change (1981/89 to 2001/10) in fract. changing firms in last year
Source: CPS ASEC 40
3. Cross-state relationship between job transition rate, migration rate
( ). . .st st st s t s t stmig rate avg jobtrans rate X timeβ γ α φ ψ ε= + + + + +
CPS IRS0.06** 0.04**(0.01) (0.01)
0.05 0.04*(0.04) (0.02)
0.05 0.00(0.01) (0.02)
Actual change in mig. (1981-1989 to 2002-2009)
-1.11 -0.43
Decline accounted for by changes in job transitions
-0.50 -0.24
% changing firm since prev. year
% changing industry since prev. year
% changing occupation since prev. year
Note. X includes %<24, %>64, % HS degree or less, %homeowner, wage inequality, %middle-skilled jobs, % mfg, % emp. provided health care, % both spouses employed, %male, %employed, %white, %black, %married, %hh w/ children.
41
In sum, the downward trend in interstate migration has coincided with a downward trend in labor market transitions. What is the direction of causality?
42
Direction of causality?
• Decline in migration probably not causing the decline in job transitions because: o Job transitions are falling even for individuals that remain
in the same state.
43
Direction of causality?
68
1012
1416
mm
jtran
sss
.4.8
1.2
1.6
22.
4
1981 1984 1987 1990 1993 1996 1999 2002 2005 2008 2011year
Labor market transition, state transition (left)No labor market transition, state transition (left)Labor market transition, no state transition (right)
Note. Labor market transitions are a change in employer, a change in occupation, or a change in industry. Source: CPS ASEC 44
Direction of causality?
• Decline in migration probably not causing the decline in labor market transitions because: o Migration flows are too small:
• In the CPS, percentage of population that changed employers fell by about 4 percentage points, while percentage that changed states fell by 1 percentage point.
• Search for explanations rooted in the labor market.
45
Where to from here?
• Factors affecting the slowdown in firm formation.
• Compensation changes within firms.
• Changes in perceived riskiness of working in a new organization.
46
Thank you!
47
48
Demographics and other characteristics Interstate Migration by Census Division
.01
.02
.03
.04
.05
.06
.07
1980 1985 1990 1995 2000 2005 2010 2015Year
New England Mid Atlantic EN Central WN Centeral S Atlantic
ES Central WS Central Mountain Pacific
Frac
tion
of D
ivis
ion
49
Reasons for within county migration
Source: CPS March Supplement
Frac
tion
of th
e po
pula
tion
Year
Family related Job related Housing related Other reasons
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
0
.01
.02
.03
.04
.05
.06
.07
50
0
.01
.02
.03
Frac
t. of
pop
. mov
ing
stat
es in
last
yea
r
1980 1985 1990 1995 2000 2005 2010
Occ. last year was 'low-skill' (Prot. service, food prep, janitorial/cleaning, personal care/service)
Occ. last year was 'mid-skill' (Sales, office/admin, prod., operators)
Occ. last year was 'high-skill' (Managerial, prof., tech,)
A role for changing ind./occ. mix?
Source: CPS March Supplement 51
Role of employer-provided health insurance
• Rising cost of health care o Migration has fallen for households without
employer-provided health insurance.
1981-1989 2002-2012
Interstate migration rate
Household with employer-provided insurance 2.7 1.5
Household with no emp.-provided insurance 3.0 1.8
Share of households with employer-provided insurance 64.8 64.7
52
53
Changing between supervisory and non-supervisory occupations
54
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
The Decline in Homeowner Mobility and Implications for the Market
Sam Khater VP & Deputy Chief Economist
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Overview
56 Source: CoreLogic
• ‘New’ source of demographic data to analyze homeowner mobility, length of tenure, and migration data
• Homeowner mobility significantly decreases during recessions and never recovers
• Mobility increasingly occurring within same metro with large interstate declines
• Sellers leaving high appreciation and buying in lower appreciation areas
• Mobility significantly declined in rising rate environment, but there’s hope
• Economic, mortgage market and policy implications of lower homeowner mobility
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Public Records vs Census Survey Data
57 Source: CoreLogic
• Census - Covers owners and renters and has deep, rich detail. CPS has long history but sample size is small (~ 60k) and only ACS has enough detail for small areas. Both are annual datasets released with lags.
• Matched Public Records – Public records only covers owners/investors, excludes renters. Thin detail on owners/investors. High frequency monthly data, very large data set down to very small areas. Matched public records merged with consumer marketing data to provide identify the sales and purchase transaction.
• Matched public records can serve as a compliment to Census data and used together to show more timely and geographically comprehensive picture of migration patterns for real estate transactions.
• This approach used in medical research for to track exposure to environmental hazard. To my knowledge not used in economic or demographic research.
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
0
1
2
3
4
5
6
7
8
0
2
4
6
8
10
12
14
1995 2000 2005 2010 2015
Number of Years A Home Is Owned (Median)
American Housing Survey - L Public Record - R
Source: American Housing Survey for the United States, various years (difference between survey year and median year householder moved into unit), CoreLogic public records (length of time between recorded sales on same home).
Trends in Length of Tenure in Public Record Similar to Census Surveys
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
The Advantage of Public Record is High Frequency Recent Data Number of Months Between Purchase and Sale
59 Source: CoreLogic
20
30
40
50
60
70
80
90Ja
n-85
Feb-
86M
ar-8
7A
pr-8
8M
ay-8
9Ju
n-90
Jul-9
1A
ug-9
2S
ep-9
3O
ct-9
4N
ov-9
5D
ec-9
6Ja
n-98
Feb-
99M
ar-0
0A
pr-0
1M
ay-0
2Ju
n-03
Jul-0
4A
ug-0
5S
ep-0
6O
ct-0
7N
ov-0
8D
ec-0
9Ja
n-11
Feb-
12M
ar-1
3A
pr-1
4M
ay-1
5
Mon
ths
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Declining Mobility Correlated to Business and Real Estate Cycle Number of Months Between Purchase and Sale
60 Source: CoreLogic
20
30
40
50
60
70
80
90Ja
n-85
Feb-
86M
ar-8
7A
pr-8
8M
ay-8
9Ju
n-90
Jul-9
1A
ug-9
2S
ep-9
3O
ct-9
4N
ov-9
5D
ec-9
6Ja
n-98
Feb-
99M
ar-0
0A
pr-0
1M
ay-0
2Ju
n-03
Jul-0
4A
ug-0
5S
ep-0
6O
ct-0
7N
ov-0
8D
ec-0
9Ja
n-11
Feb-
12M
ar-1
3A
pr-1
4M
ay-1
5
Mon
ths
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Oil Bust in 80’s Led to Much Lower Mobility in Texas Number of Months Between Purchase and Sale in Texas
61 Source: CoreLogic
Mon
ths
20
30
40
50
60
70
80
90
10019
8019
8119
8219
8319
8419
8519
8619
8719
8819
8919
9019
9119
9219
9319
9419
9519
9619
9719
9819
9920
0020
0120
0220
0320
0420
0520
0620
0720
0820
0920
1020
1120
1220
1320
1420
15
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Late 80s Recession Led to Lower Mobility in Massachusetts Number of Months Between Purchase and Sale in Massachusetts
62 Source: CoreLogic
Mon
ths
20
30
40
50
60
70
80
90
10019
8019
8119
8219
8319
8419
8519
8619
8719
8819
8919
9019
9119
9219
9319
9419
9519
9619
9719
9819
9920
0020
0120
0220
0320
0420
0520
0620
0720
0820
0920
1020
1120
1220
1320
1420
15
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Recession & Real Estate Bust Impact on Nevada Mobility More Muted Number of Months Between Purchase and Sale in Nevada
63 Source: CoreLogic
Mon
ths
20
30
40
50
60
70
80
90
10019
8619
8719
8819
8919
9019
9119
9219
9319
9419
9519
9619
9719
9819
9920
0020
0120
0220
0320
0420
0520
0620
0720
0820
0920
1020
1120
1220
1320
1420
15
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
When they Move, Majority Stay in Same Metro Percent of Sellers That Stay in Same Metro
64 Source: CoreLogic
45%
50%
55%
60%
65%Ja
n-00
Aug
-00
Mar
-01
Oct
-01
May
-02
Dec
-02
Jul-0
3Fe
b-04
Sep
-04
Apr
-05
Nov
-05
Jun-
06Ja
n-07
Aug
-07
Mar
-08
Oct
-08
May
-09
Dec
-09
Jul-1
0Fe
b-11
Sep
-11
Apr
-12
Nov
-12
Jun-
13Ja
n-14
Aug
-14
Mar
-15
Oct
-15
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Share Moving Out of State is Declining Percent of Sellers That Move to Another State
65 Source: CoreLogic
20%
22%
24%
26%
28%
30%
32%
34%
36%
38%
40%Ja
n-00
Aug
-00
Mar
-01
Oct
-01
May
-02
Dec
-02
Jul-0
3Fe
b-04
Sep
-04
Apr
-05
Nov
-05
Jun-
06Ja
n-07
Aug
-07
Mar
-08
Oct
-08
May
-09
Dec
-09
Jul-1
0Fe
b-11
Sep
-11
Apr
-12
Nov
-12
Jun-
13Ja
n-14
Aug
-14
Mar
-15
Oct
-15
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Top 10 States for In & Out-migration Ratio of Owners Moving Out to Moving in for 2000 to 2015
66 Source: CoreLogic
0.5
1.0
1.5
2.0
2.5
3.0
NJ CA IL VA CO PA TX FL AZ NC
FL NC TX
AZ FL NJ CA
FL NJ CA
CA CO
FL VA
States in Red are Top 2 Destinations for Out-migrants, States in Green are Top 2 States for In-Migrants
More Owners Moving Out
More Owners Moving In
FL PA AZ
TX
FL WI
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Movers in the Same Metro Pay $50k More for Subsequent Purchase Median Price, Sellers and Buyers within Same Metro
67 Source: CoreLogic
100000
150000
200000
250000
300000
350000
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Sell Same Metro Buy Same Metro
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Movers to Different State Pay Roughly Same for Subsequent Home Median Prices, Sellers that Sell in One State and Purchase in Another State
68 Source: CoreLogic
100000
150000
200000
250000
300000
350000
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Sell in Current State Buy Out of State
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Movers Are Selling High Appreciation Areas & Buying In Low Appreciation Areas Percent Change in Prices by Type of Move and Destination
69 Source: CoreLogic
0%
10%
20%
30%
40%
50%
60%
70%
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015
Sell Same Metro Buy Same Metro Sell in Current State Buy Out of State
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
California and Texas Interstate Movers
70 Source: CoreLogic
$-
$100
$200
$300
$400
$500
$600
Thou
sand
s California In-migrants
Median Sell Price Median Purchase Price
$-
$100
$200
$300
$400
$500
$600
Thou
sand
s
California Out-migrants
Median Sell Price Median Purchase Price
$-
$50
$100
$150
$200
$250
$300
$350
Thou
sand
s
Texas In-migrants
Median Sell Price Median Purchase Price
$-
$50
$100
$150
$200
$250
$300
$350
Thou
sand
s
Texas Out-migrants
Median Sell Price Median Purchase Price
Buyers moving into California face steep prices compared to price of home they sold in the state they sold their home.
Sellers of California properties that move out of state, move to very affordable places.
Buyers moving into Texas are face similar prices relative to prices they sold their homes. Notice the bubble hump from boom states in mid 2000s that moved to Texas
Movers leaving Texas purchase more expensive properties in other states
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Large Drop Off in Sales as Rates Go Out of the Money? Number of Home Sales by Change in Mortgage Rate Between Sales, 1976 - 2016
71 Source: CoreLogic
-
1,000,000
2,000,000
3,000,000
4,000,000
5,000,000
6,000,000
7,000,000
8,000,000
-10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6 7Rate Spread (<0 in the money, >0 out of money)
Sales When Rates Are Higher Sales When Rates Are Lower
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential. Source: CoreLogic
Mobility Rate by Year After Purchase Percent of Sales by Purchase Year Vintage as Share of Remaining Unsold Homes
72
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Overall Sell Rate
Number of Years
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential. Source: CoreLogic
Rising Rates Are a Large Hurdle for Resales Percent of Sales by Purchase Year Vintage as Share of Remaining Unsold Homes For Rising vs Declining Rate Cohort
73
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
Sell Rate for -3% Change Sell Rate for -1.5% Change
Sell Rate for 1.5% Change Sell Rate for 3.0% Change
Number of Years
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential. 74
Source: CoreLogic
-
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
100,000
0
1
2
3
4
5
6
7
8
-6 -5 -4 -3 -2 -1 0 1 2 3 4 5 6
Milli
ons
1976 - 2016 - L 1980 - 1984 - R
Rate Spread (<0 in the money, >0 out of money)
Large Drop Off in Sales as Rates Go Out of the Money? Home Sales by Mortgage Rate for Sale Minus Subsequent Purchase for Sales Between 1976 - 2016
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Mortgage Market Implications
75 Source: CoreLogic
• Lower turnover rates means lower purchase loan originations
• Provides more time to build equity, lower default risk & perhaps more 2nd
lien demand
• More demand for FRM or longer-term hybrid ARMs to minimize owner interest rate risk
• While mobility declines when mortgage rates rise, don’t expect mortgage rates to rise much
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
Current & Future Work
76 Source: CoreLogic
• Migration Using Purchase Mortgage Application
• Single-Family Repeat Rent Index
• Low/Mod Income & Underserved HMDA Real Time Proxy
• Commercial Real Estate Analytics
©2016 CoreLogic, Inc. All rights reserved. Proprietary and Confidential.
The views, opinions, forecasts and estimates herein are those of the CoreLogic Office of the Chief Economist, are subject to change without notice and do not necessarily reflect the position of CoreLogic or its management. The Office of the Chief Economist makes every effort to provide accurate and reliable information, however, it does not guarantee accuracy, completeness, timeliness or suitability for any particular purpose.
CORELOGIC and the CoreLogic logo are trademarks of CoreLogic, Inc. and/or its subsidiaries.
Where to find more information http://www.corelogic.com/about-us/research.aspx @TheSamKhater @CoreLogicEcon
77
All materials from today’s talk can be found here: http://www.urban.org/events