Housing Finance Policy Center Lunchtime Data Talk The ... · Housing Finance Policy Center...

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


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


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


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


• 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

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

Changing between supervisory and non-supervisory occupations

54


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


Public Records vs Census Survey Data


• 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.


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


The Advantage of Public Record is High Frequency Recent Data Number of Months Between Purchase and Sale


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


Declining Mobility Correlated to Business and Real Estate Cycle Number of Months Between Purchase and Sale


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


Oil Bust in 80’s Led to Much Lower Mobility in Texas Number of Months Between Purchase and Sale in Texas


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


Late 80s Recession Led to Lower Mobility in Massachusetts Number of Months Between Purchase and Sale in Massachusetts


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


Recession & Real Estate Bust Impact on Nevada Mobility More Muted Number of Months Between Purchase and Sale in Nevada


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


When they Move, Majority Stay in Same Metro Percent of Sellers That Stay in Same Metro


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


Share Moving Out of State is Declining Percent of Sellers That Move to Another State


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


Top 10 States for In & Out-migration Ratio of Owners Moving Out to Moving in for 2000 to 2015


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


Movers in the Same Metro Pay $50k More for Subsequent Purchase Median Price, Sellers and Buyers within Same Metro


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


Movers to Different State Pay Roughly Same for Subsequent Home Median Prices, Sellers that Sell in One State and Purchase in Another State


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


Movers Are Selling High Appreciation Areas & Buying In Low Appreciation Areas Percent Change in Prices by Type of Move and Destination


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


California and Texas Interstate Movers


$-

$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


$-

$50

$100

$150

$200

$250

$300

$350

Thou

sand

s

Texas In-migrants


$-

$50

$100

$150

$200

$250

$300

$350

Thou

sand

s

Texas Out-migrants


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


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


-

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


Mortgage Market Implications


• 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


Current & Future Work


• Migration Using Purchase Mortgage Application

• Single-Family Repeat Rent Index

• Low/Mod Income & Underserved HMDA Real Time Proxy

• Commercial Real Estate Analytics


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

http://www.corelogic.com/about-us/research.aspx

All materials from today’s talk can be found here: http://www.urban.org/events

Housing Finance Policy Center Lunchtime Data Talk The ... · Housing Finance Policy Center...

Documents

Transcript of Housing Finance Policy Center Lunchtime Data Talk The ... · Housing Finance Policy Center...