Embed Size (px)
Citation preview
Preliminary Findings from the Oregon PovertyMeasure Project
David Rothwell, Bruce Weber, & Leanne Giordono
May 23, 2019 | Pacific Northwest Regional Economic ConferenceSeaside, OR
Outline
I Why do we need another poverty measure?I Prelim findings
What is poverty?
I Insufficient resources to meet some level of need over aperiod of time.
I Income to needs = income / threshold
What is poverty?
I Insufficient resources to meet some level of need over aperiod of time.
I Income to needs = income / threshold
Current poverty in Oregon
Lake20
Harney17.5
Malheur25.2
Lane18.8
Grant13.7
Klamath18.7
Douglas17
Baker15.3
Linn16.1
Crook15.3
Umatilla17.8
Coos17.9
Jackson16.7
Wasco13.7
Wallowa13.7
Curry15.5
Union17.4
Deschutes12.1
Morrow14.7
Wheeler20.6Jefferson
20.9
Clackamas9
Gilliam9.9
Polk15.4
Lincoln18.4
Josephine18.6
Clatsop12.2
Marion15.9
Tillamook15.5
Yamhill13.7
Benton20.7
Sherman13.7
Columbia12.3
Washington10.3 Hood River
12.1Multnomah
16.4
Percent in Poverty by County2013-17 (OR: 14.9%)
SOURCE: Census American Community Survey 5-Year Estimates, 2013-2017.
Please let me know of any changes you would like.
2/18/2019
% on Poverty< 10%10.01% - 15%15.01% - 20%20.01% - 25%>25%
Problems with the Official Poverty Measure
1. threshold erosion2. definition of family3. major portions of safety net missing4. geography
A new measure
The Census Bureau introduces the Supplemental Poverty Measure(SPM) in 2009
I Definition of family includes cohabiting and foster children, +I Geographic adjustment for housing costs and tenureI Resources include SNAP, EITC, and consider taxes and other
expenses such as child care, child support, and Medical Out ofPocket Expenses
I Thresholds 3-year moving average of consumer spending,includes geographic adjustment
OPM and SPM over time national
Challenges for understanding poverty in Oregon
I SPMI Based on Census Current Population Survey, Annual Social and
Economic Supplement (ASEC)I Annual sample size in Oregon ~ n=2,600I Recommended to combine at least 3 years of data to estimate
state level poverty ratesI Major counties not identified (exception: Yamhill, Jackson,
Lane, Linn)I Measurement error in reporting income from social assistance
programs (also problem for OPM, see Meyer & Wu, 2018)
The Oregon Poverty Measure Project
I Basic idea : create an SPM-like measure with ability to informlower level geographic unitsI Based on American Community Survey (ACS) - much larger
sample and geographiesI Change family units and create SPM-like thresholdsI Improve resources measurement and policy analysis
I Imputation from administrative data (DHS, OED, OHCS) andother sources (Archived SPM, CPS, TRIM)
I Examples from Wisconsin and California
The Oregon Poverty Measure Project steps
1. Build ACS 5 year sample2. Change family units to match SPM (college students)
3. Create resource bundlesI adjustments and imputations from admin and other data
sources, e.g., school lunches4. Generate thresholds
The Oregon Poverty Measure Project steps
1. Build ACS 5 year sample2. Change family units to match SPM (college students)3. Create resource bundles
I adjustments and imputations from admin and other datasources, e.g., school lunches
4. Generate thresholds
The Oregon Poverty Measure Project steps
1. Build ACS 5 year sample2. Change family units to match SPM (college students)3. Create resource bundles
I adjustments and imputations from admin and other datasources, e.g., school lunches
4. Generate thresholds
Why an Oregon Poverty Measure Matters
Prelim findings - resources - safety net
I Safety net policy compositionI Income assistance (e.g., EITC, TANF, SSI)I Food assistance (e.g., SNAP, WIC)I Housing assistance (e.g., Section 8, LIHEAP)I Caregiving assistance (e.g., ERDC)I Health care (e.g., Medicaid)*
I Safety net policy trendsI Decrease in cash assistance (TANF)I Increase in tax credits (EITC) and noncash benefits (SNAP)I Medicaid expansionI State/federal cooperation – state-level discretion
Oregon safety net and poverty
0.0
0.1
0.2
0.3
1990 1995 2000 2005 2010 2015
year
Hea
dcou
nt p
over
ty r
ate
all
indicator Market income
Source: Archived SPM Columbia U. & CPS
Oregon safety net and poverty
0.0
0.1
0.2
0.3
1990 1995 2000 2005 2010 2015
year
Hea
dcou
nt p
over
ty r
ate
all
indicator Market income poverty SPM poverty
The safety net in Oregon reduces poverty considerably
Source: Archived SPM Columbia U. & CPS
Oregon safety net and poverty
7.5 pp 15 pp
0.0
0.1
0.2
0.3
1990 1995 2000 2005 2010 2015
year
Hea
dcou
nt p
over
ty r
ate
all
indicator Market income poverty SPM poverty
The safety net in Oregon reduces poverty considerably
Source: Archived SPM Columbia U. & CPS
Prelim findings adjustments on participation
Program Participation SNAP TANF SSIPre-TRIM Post-TRIM Pre-TRIM Post-TRIM Pre-TRIM Post-TRIM
% Overall 14.7 26.5% in SPM Poverty 30.1 73.7% Children in SPM Poverty 40.7 92.3Source: CPS 2013-17 pooled sample; Urban Institute TRIM- All differencesAll differences statistically significant at 5% level
TRIM adjusts for under-reporting (Parolin, 2019)
Prelim findings adjustments on participation
Program Participation SNAP TANF SSIPre-TRIM Post-TRIM Pre-TRIM Post-TRIM Pre-TRIM Post-TRIM
% Overall 14.7 26.5 2.0 3.5% in SPM Poverty 30.1 73.7 5.3 14.4% Children in SPM Poverty 40.7 92.3 11.5 32.0Source: CPS 2013-17 pooled sample; Urban Institute TRIM- All differencesAll differences statistically significant at 5% level
TRIM adjusts for under-reporting (Parolin, 2019)
Prelim findings adjustments on participation
Program Participation SNAP TANF SSIPre-TRIM Post-TRIM Pre-TRIM Post-TRIM Pre-TRIM Post-TRIM
% Overall 14.7 26.5 2.0 3.5 4.2 5.2% in SPM Poverty 30.1 73.7 5.3 14.4 8.5 13.3% Children in SPM Poverty 40.7 92.3 11.5 32.0 6.1 15.5Source: CPS 2013-17 pooled sample; Urban Institute TRIM- All differencesAll differences statistically significant at 5% level
TRIM adjusts for under-reporting (Parolin, 2019)
Prelim findings - thresholds
Prelim findings - thresholds
Prelim findings - adjust thresholds at the PUMA level
2 adult 2 children reference family
Area Home w/ mortgage Home w/o mortgage RenterPortland MSA 28106 23995 28021PDX EastPDX S.E.PDX North NEPDX Central EPDX NW SW
Prelim findings - adjust thresholds at the PUMA level
2 adult 2 children reference family
Area Home w/ mortgage Home w/o mortgage RenterPortland MSA 28106 23995 28021PDX East 27245 23376 27164PDX S.E. 28716 24432 28626PDX North NE 29677 25122 29582PDX Central E 30114 25436 30016PDX NW SW 33449 27831 3331
Prelim findings - adjust thresholds at the PUMA level
Median 25672
Portland City (Northwest & Southwest)Portland City (Central East)
Washington County (Central 2)Clackamas County (Northwest 2)
Washington County (Central 1)Portland City (North & Northeast)
Washington County (Southeast)Portland City (Southeast)
Washington County (Northeast)Clackamas County (Northwest 1)
Multnomah County (East)Portland City (East)
Washington County (West)Clackamas County (South & East)
Deschutes CountyJackson County (Central)
Lane County (West Central)Lane County (East Central)
Jackson County (Outside Medford & Central Point)Linn & Benton Counties
Marion County (West Central 2)Lane County (Outside Eugene & Springfield)
Columbia, Loincoln, Clatsop & Tillamook CountiesYamhill & Polk Counties
Marion County (Outside Salem & Keizer)Marion County (West Central 1)
Josephine, Coos & Curry CountiesDouglas County
North Central OregonKlamath, Malheur, Lake & Harney CountiesUmatilla, Union, Baker & Wallowa Counties
0
1000
0
2000
0
3000
0Supplemental Poverty Threshold ($)
Pub
lic U
se M
icro
data
Are
a (P
UM
A)
tenureOwn with mortgageOwn without mortgageRent
OR Poverty Measure: Base Thresholds by PUMA
Oregon Poverty – What we are working on
I Demographic variation (e.g., age*geography; migration)
I Oregon policies (e.g., minimum wage changes, Medicaidexpansion, Oregon Earned Income Credit)
I Impacts of Oregon conditions (e.g., labor markets, childcare,housing)
Oregon Poverty – What we are working on
I Demographic variation (e.g., age*geography; migration)I Oregon policies (e.g., minimum wage changes, Medicaid
expansion, Oregon Earned Income Credit)
I Impacts of Oregon conditions (e.g., labor markets, childcare,housing)
Oregon Poverty – What we are working on
I Demographic variation (e.g., age*geography; migration)I Oregon policies (e.g., minimum wage changes, Medicaid
expansion, Oregon Earned Income Credit)I Impacts of Oregon conditions (e.g., labor markets, childcare,
housing)
Wrap up
I The Oregon Poverty Measure will be a more valid measurethan OPM or SPM and provide more detail
I With a new measure we can have more confidence in rate andbetter explanations
I Study compositional changes; poverty depth and severity;target resources
Thank you
Web space in progress
Appendices
1. Threshold erosion
2. Definition of family
3. Policies are missing
4. Geography
Differences in rates 2015
−10
0
10
20
30
overall child oldera rural
Pov
erty
rat
es
OPM poverty SPM poverty Difference
Source: Archived SPM Columbia U. & CPS
Differences in composition 2015
Supplemental Poverty Measure
Official Poverty Measure
0 25 50 75 100
Age group
child workingage olderadults
Source: Archived SPM Columbia U. & CPS
Child care costs
Child Care Costs as % of Income Center Home
Married, 2 children 28.4 21.2Married, 2 children, poverty 94.0 70.0
Source: Child Care Aware
-Plus child care deserts in 36 counties (Pratt et al., 2019)
Oregon health insurance access
I Health insurance access (ACS 2016)I Oregon 93.8% (US 91.4%)I Medicaid expansion
I Health Inclusive Poverty Measure (Korenman and Remler,2016; Remler and Korenman, 2017)
I 1 pp impact of MA Medicaid expansion on poverty using HIPM(Zewde, 2019)
Geography matters
Housing affordability
