Race, Segregation and Postal Employment: New Evidence on Spatial Mismatch Leah P. Boustan UCLA and N

1
Race, Segregation and Postal Employment New
Evidence on Spatial MismatchLeah P. Boustan
(UCLA and NBER) and Robert A. Margo (BU and NBER)
2
The Spatial Mismatch Hypothesis

• Since c. 1970, blacks have been less likely than
  whites to participate in the labor force.
• Blacks tend to live in central cities while
  whites live in suburbs.
• Spatial mismatch As employment moved to suburbs,
  blacks lived further from jobs. Some withdrew
  from LF due to expense of reverse commuting or
  lack of information about openings.
• Kain, 1968 Ellwood, 1986 Raphael, 1998 Ross,
  1998 Weinberg, 2000, 2004a, 2004b etc.

3
Two contributions of this paper(1) Add an
historical dimension

• 1940/50 Blacks, whites and jobs all located in
  cities
• 1960/70 Whites to suburbs employment follows
  blacks remain in cities
• 1980-onward Blacks begin moving to suburbs too

4
Two contributions of this paper(2) Alternative
indicator of mismatch

• Typical approach Focus on LF participation.
  Compare metro areas where blacks are
  segregated/concentrated in city vs. not.
• Our idea Blacks will substitute towards
  centrally located jobs. Postal service has
  historically been and remains centralized.
• Benefit Blacks in segregated areas may be
  negatively selected. Could explain low employment
  rates. Yet, postal work is positively selected.
• (Requires civil service exam for entry. Black
  postal workers have earned above natl median
  from 1940-2000.)

5
We show that.

• Black postal employment increased dramatically
  relative to whites from 1940 to 1970
• This increase was larger in segregated metro
  areas where blacks tend to be concentrated in the
  central city
• This increase occurred primarily in centralized
  sorting and processing, not in mail carrying,
  which is geographically dispersed.
• The relationship between segregation and black
  postal employment, while still present today, has
  declined since 1970.
• We argue that spatial mismatch was an important
  factor in black
• occupational choice from 1950-70, but is less
  important today.

6
Other (Direct) Evidence in Our Favor

• Similar cross-section pattern apparent in 1970
  for other public sector jobs that were highly
  centralized.
• Positive relationship between segregation and
  black postal employment vanishes if CC share of
  overall employment is high enough (1980 IPUMS)

7
We can rule out alternative stories

• Pattern is ALREADY present in 1960 so NOT due to
• Executive Order No. 10925 in 1961 (JFK)
• Post-1963 civil rights legislation
• 1960s riots
• NOT a pre-1960s regional story (voting rights)
  because differences in outcomes between Northern
  and Southern cities were very small

8
Timing Suburbanization of people and jobs

• Share of metro employment in central city
• 1950-60 64 to 61.8 (only manuf., retail,
  some services)
• 1960-70 59.3 to 51.7
• 2000 42.3
• Share of black metro households in central city
• 1950-1970 Above 80
• Constrained by covenants, intimidation, realtors
  and financial inst.
• 1980 68.1
• Fair Housing laws (1968) and related changes

9
Despite employment decentralization, certain
postal jobs remain in central cities

• In early 20th century, inter-city mail
  transported by rail. Mail processing and
  distribution centers (PDC) near central depots
• In 1970s, USPS considered moving some facilities
  to suburban locations near airports and
  interstate exits/exchanges
• Plant closing process is highly contentious and
  subject to legal, regulatory, and political
  delays ? to this day, PDCs remain mostly
  centralized

10
Why current PDCs are sticky

• Legal Closing a PDC constitutes major federal
  action that requires preparation of an
  environmental impact statement. Must consider
  local job loss. Must give advance notice to
  state and local government. Critical case City
  of Rochester v. U.S. Postal Service 541 F.2d967
• Labor Relations USPS has no job loss
  provisions and advance notice requirements in
  its union contracts if it intends to close or
  merge facilities
• Politics Local, state, and federal elected
  officials often lobby against closure
• Dispossession of surplus federal property is
  complicated and time-consuming
• High probability that CC PDC remains in
  operation, undergoing periodic maintenance,
  renovation, and retrofitting (eg. Detroit, 1401
  W. Fort St.) OR the old facility is replaced with
  a new one around the corner (Cardiss-Collins
  Facility, Chicago). SOME post-1970 CC closings
  did occur (Atlanta, Rochester)

11
Location Centralized postal employment

• Micro Census data on place of work. Mean for
  metro areas.
• Same relative pattern in 2000

• Panel B shows figures from 2000 census using
  current addresses for 237 PDCs in 2007 Postal
  Directory
• Neighborhood defined as Census tract with PDC
  and all adjacent

12
Example Cincinnati, OH
2000 Neighborhood 65.7 black County 23.6
1970 Neighborhood 63.8 black County 15.7
13
Figures 1 and 2

• Figure 1 shows growth of postal and intrinsic
  government employment over 20th century
• AA postal employment rate white after WW2
• Figure 2 variation in postal employment rates
  across metropolitan areas in 1970
• White rates are similar BUT AA rates vary
  tremendously. In some metro areas, close to 7.5
  percent of adult black population worked for the
  Post Office in 1970 (!!!)
• Postal employment was (and is) a good job. In
  1970 mean wage of AA postal workers was at 75th
  percentile of AA wage distribution (same in
  2000) for whites, corresponding figure is 46th
  percentile (59th in 2000)

14
Black postal employment increased from 1940-1970
(1) relative to whites and (2) relative to other
public sector
15
Substantial cross-sectional variation in black
postal employment (Not true for whites!)
1970
16
Data and specification

• 1 if work at USPSimt am ß(segregationmt)
  ?(blacki segmt) eimt
• i individual m metro area t decade
• IPUMS 1940-2000 (note USPS only civilian
  employer identified in Census!)
• Hypothesis ? 0 by 1960
• Segregation Dissimilarity index. Ranges from 0
  to 1 (perfect integration to perfect
  segregation).
• Data problems
• Dissimilarity measured for city in 1940-50 BUT
  for SMSA, 1960-2000
• 1960 IPUMS does NOT identify SMSA. We run
  state-level analysis.

17
(2) Black employment at USPS positively related
to metro segregation, especially in 1960 and 1970

• Same results when add year-specific metro area
  fixed effects (Can only identify interaction)
• Same results when restrict to 45 metro areas
  available in every year

18
State results Correlation emerges between 1950
and 1960
Same if consistent set of states is used. Large
effect in 1960 rules out Civil rights
legislation Postal Reorganization Act (1970)
urban riots preferential treatment of Vietnam
vets. Also, blacksegSouth is negative in
1960-70 BUT very small and insignificant, not a
voting rights story.
19
Result not driven by outliers (1970)
20
Consistent cross-section finding Segregation
strongly predicts black postal employment when
employment is decentralized

• Consider city A and city B separated by one SD
  diff in emp in cc (55 to 71)
• Moving from 0?1 on segregation index leads to
• 5.1 point increase in black postal employment in
  city A
• 1.6 point increase in city B
• Similar results using AA living in CC OR
  centralization index (Table 5)

21
(3) This pattern only holds for postal clerks,
not mail carriers
Challenges alternative explanation based on
association between segregation and private
sector racism. If so, we would expect comparable
effects throughout post office and public sector.
22
Robustness Checks (1)

• Expanding sample to include non-workers does NOT
  matter (Table 3)
• Weighting by metro area population does NOT
  matter (Table 7)
• Including race interactions with city-level
  covariates (eg AA, median family income, land
  area, etc. etc.) does NOT affect racesegregation
  interaction (see Table 7)

23
Robustness Checks (2)

• (1) Recall, segregation measured for city in
  1940/50 and for metro thereafter.
• Coefficient cut in half when use city measure in
  1970.
• Yet, even if double 1950 coefficient, still
  substantially smaller than 1970. NOT (solely) a
  measurement issue.

24
Robustness checks (3)

• Table 4, level effect of segregation is negative,
  suggests whites might have increasingly avoided
  postal work in cities that were BECOMING
  segregated.
• Decompose segregation in 1970 into (a)
  segregation in 1940 (b) change from 1940-1970.
• Coefficient of (b) is positive and significant
  BUT so is coefficient of (a)

25
Robustness Check 4 If pattern due to spatial
mismatch, would expect stronger effect among less
educated
1940 1970 2000
26
Conclusions

• We can detect evidence of spatial mismatch using
  a non-standard outcome measure (postal
  employment).
• Mismatch was particularly severe in 1960 and
  1970, as employment moved out to suburbs but
  blacks were yet unable to follow.
• While the housing market is more open today, many
  blacks still live far from the modal job opening.
  The evidence here suggests that space matters.

27
Future agenda

• Postal employment and black-white wage
  convergence
• Between 1940 and 1970, AA employment in
  intrinsic public sector nearly tripled (albeit
  from small base). Half of this growth due to
  postal employment.
• Increase for whites much smaller.
• Large public-private wage premium (esp.
  postal-private) for AA.
• Usual explanation for growth in AA public sector
  employment Demand side
• Voting rights and anti-discrimination legislation
  (eg. Freeman on public school teachers)
• But, AA postal employment increases before Civil
  Rights. The supply side component that we
  attribute to spatial mismatch is an unintended
  (?) jobs program for central city residents.
• ALSO Work in progress on urban topics using
  other jobs with pre-determined locations (State
  Capitals for Brookings-Wharton)

28
Additional Slides

• Boustan-Margo, Blacks and the Post Office

29
(No Transcript)
30
(No Transcript)
31
(No Transcript)
32
(No Transcript)
33
(No Transcript)
34
(No Transcript)
35
(No Transcript)
36
(No Transcript)
37
(No Transcript)
38
(No Transcript)
39
The impact of local black political power on the
location of postal facilities

• Black politicians are vocal opponents of
  relocating PDCs.
• Due to residential segregation, black political
  power is intensified in a ward-based (rather than
  at-large) election system.
• Black employment at the post office is higher in
  cities with ward-based elections.
• Furthermore, higher black population share is
  ONLY associated with postal employment in
  ward-based cities.

40
Relationship between local election format and
black postal employment 1980
41
(No Transcript)
42
USPS Facilities 2004 Data
43
What Happened in Atlanta and Rochester

• Ca. 1970 Atlanta PDC was located in CBD tract,
  surrounded on three sides by census tracts
  ranging from 49 to 98 percent AA
• USPS decides to move the facility closer to the
  Atlanta airport (Hartsfield)
• NAACP files lawsuit in federal district court on
  (a) constitutional grounds (due process clause)
  and (b) residents had a right to advance notice
  and public hearing on the basis of federal law
  (postal service) and associated court rulings
• Case is dismissed in 1975 on grounds that
  plaintiffs legal arguments were incorrect (no
  constitutional property right to access to a
  specific post office, postal law permits hearing
  only if impact on mail service is substantially
  nationwide), National Association for the
  Advancement of Colored People (Atlanta local) et.
  al. v. United States Postal Service (1975)
• A new facility was built close by the airport (miles) on a major road (Crown Road SE) with
  almost immediate access to I-75 road is border
  of a census tract that was already 29 percent AA
  in 1970
• New PDC tract adjacent to another that was 73
  percent AA in 1970
• By 1980, these tracts were, respectively, 87.5
  percent and 97.3 percent AA (in 2000, both were
  98 percent AA)
• Moral even though the Atlanta PDC moved from
  central city to airport-area location, it
  remained physically close to an AA residential
  enclave
• Aside Cleveland PDC is located next to Hopkins
  International Airport, abuts a majority AA (in
  2000) census tract Denver PDC also close to
  airport in a census tract that abuts an AA
  residential enclave
• Rochesters PDC was originally located downtown
  at edge of CBD USPS makes a decision to
  suburbanize sometime in the late 1960s or early
  1970s Rochester sues USPS on grounds that
  service did not properly consider views of other
  government agencies (ICA) nor did it prepare a
  proper environmental impact statement (NEPA)
• Court finds that USPS did violate NEPA and ICA
  but plaintiff (city of Rochester) waited too long
  to file in light of pre-existing knowledge of
  USPS plans to receive injunctive relief however,
  USPS was still required to do a proper impact
  statement before continuing with construction of
  the new facility in suburban Rochester
• Rochester PDC was built near major highway
  (I-390) with easy access to city and airport but
  does NOT abut an AA residential enclave

44
Staying Close By Chicago

• Chicagos current PDC is the Cardiss-Collins
  facility located at 433 W. Harrison Street,
  opened in 1996
• Cardiss-Collins replaced the main PO originally
  built in 1932, located at 433 W. Van Buren Street
  (I-290 and Chicago River, about 2/3rds of a mile
  from Grant Park)
• Walking distance between two facilities 1,760
  feet
• In 1970, main Chicago PO was in a predominantly
  white census tract surrounded (three sides) by
  predominantly AA census tracts
• Cardiss-Collins facility is located in a census
  tract that was 74 percent AA in 2000, adjacent
  tract is 97 percent AA
• Who is Cardiss Collins? AA congresswoman
  (Democrat) who represented the 7th Congressional
  District, IL (majority AA district) from 1973
  until retirement in 1996

45
Some Relevant Technical Change Within the Postal
Service

• Long run march towards automation
• Prior to 1950s mail is sorted into cubbyholes
  (often on railroad cars en route)
• Mechanized letter sorting in 1950s but required
  an operator to key punch address
• Optical character readers introduced in early
  1980s
• Automated sorting of flat parcels and packages is
  relatively recent
• USPS work-sharing (bulk mail, pre-sorting)
  officially begins in 1970s (2nd and 3rd class
  mail is earlier than this)
• Add internet Ebay ? first class mail is
  declining (rapidly)

46
(No Transcript)
47
(No Transcript)
48
(No Transcript)
49
Racial Diversity in Current USPS Workforce

• USPS maintains employee personnel files with
  racial and gender identifiers which is used by
  GAO to periodically prepare reports (for
  Congress) on employee diversity
• 2000 report examines minority and female
  representation in higher occupational ranks,
  includes white male by postal district for all
  career PO employees (around 100 districts, each
  has its own PDC) in 1999
• HUGE spatial variation in white male 5 in LA,
  8 in Chicago v. 78 in Boston
• Nationwide white females are under-represented in
  postal service employment ? suggests spatial
  variation is primarily racial
• Post Office attributes variation in white male
  to successful (or not) implementation of its
  diversity initiatives
• BUT interviews with Chicago-area postal officials
  reveal purported difficulties in recruiting
  Hispanics and women for cultural reasons
  (alleged unwillingness to work nights, required
  of new employees)

50
(No Transcript)
51

•  
• State level results with all states included as
  the enter the sample
• Pooled cross-section contains state fixed
  effects
• Rows 2 and 3 run as OLS not seemingly unrelated
  regressions

52
How to Get a Job at the Post Office

• Take Civil Service exam
• Pool of available qualified applicants is rank
  ordered based on exam scores
• Hiring is subject to the Rule of Three from the
  pool of available qualified applicants when a job
  opening occurs
• Hiring official must choose from the top three
  ranked candidates (HOWEVER veteran status trumps
  rank)
• We emphasize entry because PO employment
  increases by 230 percent from 1940 to 1970
• We emphasize AA supply response because
  (empirically) white postal employment rates are
  relatively constant across metro areas ca. 1970
  (see below)