Supplemental Poverty Measure: Alternative Geographic ...

Supplemental Poverty Measure:Alternative Geographic Adjustments

Trudi RenwickAssistant Division Chief for Economic Characteristics

Social, Economic and Housing Statistics DivisionU.S. Census Bureau

DRB-FY19-ROSS-B0121

May 20, 2019

This presentation was developed to promote research and advancements in our understanding of poverty measurement. In that spirit and to encourage discussion and thoughtful feedback at early stages of our work, this paper has undergone a more limited review than official Census Bureau reports. All views and any errors are solely those of the author and do not necessarily reflect any official position of the Bureau. All comparative statements in this presentation have undergone statistical testing, and, unless otherwise noted, all comparisons are statistically significant at the 10 percent significance level. Do not cite or distribute without author permission.

The Census Bureau reviewed this data product for unauthorized disclosure of confidential information and has approved the disclosure avoidance practices applied to this release. DRB‐FY19‐ROSS‐B0121

Official and SPM Thresholds for Units with Two Adults and Two Children

$24,858$27,085

$23,261

$27,005

Officialpovertymeasure

Ownerswith a

mortgage

Ownerswithout amortgage

Renters

Supplemental Poverty MeasureThresholds, 2017

Source: Official Poverty Thresholds, <www.census.gov/data/tables/time‐series/demo/income‐poverty/historical‐poverty‐thresholds.html>, Supplemental Poverty Measure Thresholds, Bureau of Labor Statistics (BLS), <https://stats.bls.gov/pir/spmhome.htm>, Geographic adjustments based on housing costs from the American Community Survey 2012‐2016.

Supplemental Poverty Measure Thresholds for Renters, 2016

2

Median Rent Index

3

Thresholdijt=[(HousingSharet×MRIij) +(1‐HousingSharet)]×Thresholdti = statej = specific metro area, other metro or nonmetro area t = tenure: owner with mortgage, owner without a mortgage, renterHousingShare = percent of threshold represented by housing and

utilities which ranges from 40 to 50 percent of total expenditures, depending on tenure status

MRI = Median Rent IndexThreshold = national average dollar value for income below which

consumer units are considered in poverty

Washington DC vs. Nonmetro Mississippi2017– Two Adults Two Children - Renter Washington, DC Mississippi Nonmetro

Areas

Official Poverty Threshold $24,858 $24,858

SPM Threshold: Renters $27,005 $27,005

4

Washington, DC vs. Nonmetro Mississippi

2017 – Two Adults Two Children - Renter Washington DC Mississippi NonmetroAreas



Rent-based Index Using Median Rent Index (MRI) $1,297/972=1.63 $598/972=0.78

Apply to Only Housing Portion of Thresholds 50%*1.63+50%*1.0 50%*0.78+50%*1.0

MRI Index 1.32 0.89

5


2017– Two Adults Two Children - Renter Washington, DC Mississippi NonmetroAreas





Median Rent Index (MRI) 1.32 0.89

Adjusted SPM Threshold $35,512 $24,034

6

Two Adult, Two Child Thresholds: Renters

$24,858 $24,858$27,005 $27,005

$35,521

$24,034

Washington, DC Nonmetro Mississippi

Official SPM‐NGA SPM

7

Percent of Individuals with Change in Poverty Status Due to the Geographic Adjustments

2.7

3.3

2.02.3

3.6

2.5

4.1

U.S. Total Northeast Midwest South West Inside MSA OutsideMSA

8Source: 2018 CPS ASEC

Comparing Poverty Rates by Adjustment Mechanism

14.9

20.7

15.2

21.7

0

5

10

15

20

25

California MississippiOfficial SPM‐NGA

9Source: 2015 and 2016 CPS ASEC


14.9

20.7

15.2

21.720.3

17.0

0

5

10

15

20

25

California MississippiOfficial SPM‐NGA SPM


Media likes to focus on state changes

11

Concerns with the MRI Methodology

• Adjusts only the housing portion of the thresholds – Regional Price Parities produced by BEA could adjust the entire threshold

• BUT they include many expenditure items not included in the SPM thresholds and therefore give a much smaller weight to housing – see Renwick, Aten, Figueroa, and Martin (2014)

– Custom RPPs using only Food, Apparel and Rent (FAR RPPs) • Adjustments using FAR‐RPPS are even “stronger” than using the MRI’s – see Renwick, Figueroa and Aten (2017)

• Simple medians may not capture differences in housing quality – Using housing component of the RPPs – based on a hedonic regression rather than a simple median – BUT lose the simplicity of the median

• Adjust the MRI index to recognize amenities that may reduce the impact of housing costs on economic well‐being

12

Regional Price Parities (RPPs)

• Spatial price indexes produced by the Bureau of Economic Analysis to measure price level differences across regions

• Stage One – price and expenditures inputs collected by the BLS CPI program and the CE – 38 urban areas (weights available for 38 urban areas plus 4 rural regions)

• Stage Two – combined with data from the ACS on housing costs to calculate index values for all metro areas

• Index values applied to the entire threshold

13

Washington, DC vs. Rural Mississippi2017– Two Adults Two Children - Renter Washington, DC Mississippi Nonmetro

Areas



Rent Index from Regional Price Parities 1.70 .51


Index using RPP Rents 1.35 .76


14

2017 Two Adult, Two Child Thresholds: Renters

$27,005 $27,005

$35,521

$24,034

$36,457

$20,389


Not Geographically Adjusted SPM RPP Rent‐Only

15

Washington DC vs Nonmetro Mississippi

2017 – Two Adults Two Children Washington, DC Mississippi – outside metro area



RPP Index – Broad based 1.19 .83

Adjusted SPM Threshold – Broad Based $32,136 $22,414

16


$24,858 $24,858$27,005 $27,005

$35,521

$24,034

$36,457

$20,389

$32,136

$22,414


Official Not Geographically Adjusted2 SPM RPP Rent‐Only Regional Price Parities

17

Regional Price Parities – Narrowly Defined or FAR

• 2014 analysis examined differences between the MRI adjustments and the RPP adjustments.

• Concern that this index includes many goods and services not in the SPM thresholds. Differences in poverty rates driven by differences in expenditure shares.

• BEA developed a special RPP – based solely on Food, Apparel and Rent (FAR).• One advantage of the FAR RPP is that expenditure shares vary by geography. They do not vary by tenure type.

18

Washington, DC vs Nonmetro Mississippi

2017 – Two Adults Two Children Washington, DC Mississippi – outside metro area



RPP Index – Broad based 1.19 .83

Adjusted SPM Threshold – Broad Based $32,136 $22,414

FAR RPP Index 1.36 .70

Adjusted SPM Threshold – FAR RPP $36,727 $18,904

19


$24,858 $24,858$27,005 $27,005

$35,521

$24,034

$36,457

$20,389

$32,136

$22,414

$36,727

$18,903


Official Not Geographically Adjusted2SPM RPP Rent‐OnlyRegional Price Parities Narrow RPPs

20

Adjusting for Amenities

• Differences in prices reflect differences in the ability to purchase a given bundle of goods. This may or may not translate into equal quality of life.

• Amenities: nonmarket goods that improve the quality of life.• MRI will over‐adjust poverty thresholds if places with higher median rents also have higher “amenities”. MRI will under‐adjust poverty thresholds if places with higher median rents have lower amenities or “dis‐amenities”.

• Considerations:– Assumes that amenities are not completely capitalized in rents– Assumes that these amenities are fungible ‐‐‐ can live in a smaller house because the “weather is so wonderful you can spend more time outside”. Some amenities probably not fungible – “can’t eat the scenery”. Still need enough resources to purchase basic bundle of necessities.

– Value of amenities may vary across the income distribution• No clear methodology to incorporate amenities – this analysis arbitrarily cuts the weight of the MRI index in half.

21


2017– Two Adults Two Children - Renter Washington, DC Mississippi NonmetroAreas





Amenities Adjusted Index (AAI) 1.16 0.95


22


$24,858 $24,858$27,005 $27,005

$35,521

$24,034

$36,457

$20,389

$32,136

$22,414

$36,727

$18,903

$31,258

$25,520


Official Not Geographically Adjusted2SPM RPP Rent‐OnlyRegional Price Parities Narrow RPPsAmenities

23

California vs Mississippi: Poverty Rates and Material Deprivation Rates

11.2

18.0

20.3

15.2

14.9

17.2

19.7

17.0

21.7

20.7

0 5 10 15 20 25

Food Insecurity

Multi‐Dimensional Deprivation

SPM

SPM ‐ No Geo Adjustments

Official Poverty

Mississippi California

24

Sources: Poverty rates – 2015‐2016 CPS ASECMDD – Glassman, 2019 forthcomingFood Insecurity – USDA Economic Research Service

https://www.ers.usda.gov/topics/food‐nutrition‐assistance/food‐security‐in‐the‐us/key‐statistics‐graphics.aspx#map

California vs Mississippi: Deprivation Rates for Specific Dimensions

4.0

16.8

8.4

20.7

5.1

6.9

14.1

13.4

10.2

8.9

0 5 10 15 20 25

Health Dimension

Education

Economic Security

Housing Qualilty

Neighborhood Quality

Mississippi California

25

Source: MDD – Glassman, 2019 forthcoming


14.9

20.7

15.2

21.7

California MississippiOfficial SPM‐NGA

26

Source: 2015 and 2016 CPS ASEC


14.9

20.7

15.2

21.720.3

17.0

California MississippiOfficial SPM‐NGA SPM

27



14.9

20.7

15.2

21.720.3

17.0

21.9

14.9

California MississippiOfficial SPM‐NGA SPM RPP‐Rent Only

28



14.9

20.7

15.2

21.720.3

17.0

21.9

14.9

21.7

12.8

California MississippiOfficial SPM‐NGA SPM RPP‐Rent Only RPP Narrow

Source: 2015 and 2016 CPS ASEC 29


14.9

20.7

15.2

21.720.3

17.0

21.9

14.9

21.7

12.8

18.4

16.5

California MississippiOfficial SPM‐NGA SPM RPP‐Rent Only RPP Narrow RPP Broad



14.9

20.7

15.2

21.720.3

17.0

21.9

14.9

21.7

12.8

18.4

16.5

18.720.0

California MississippiOfficial SPM‐NGA SPM RPP‐Rent Only RPP Narrow RPP Broad Amenities



16.9 16.816.417.4

Inside Principal Cities Outside MetroOfficial SPM‐NGA

32

Source: 2016 CPS ASEC


16.9 16.816.417.417.8

13.2

Inside Principal Cities Outside MetroOfficial SPM‐NGA SPM



16.9 16.816.417.417.8

13.2

17.9

12.2

0

2

4

6

8

10

12

14

16

18

20

Inside Principal Cities Outside MetroOfficial SPM‐NGA SPM RPP‐Rent Only



16.9 16.816.417.417.8

13.2

17.9

12.2

17.5

10.8

Inside Principal Cities Outside MetroOfficial SPM‐NGA SPM RPP‐Rent Only RPP Narrow



16.9 16.816.417.417.8

13.2

17.9

12.2

17.5

10.8

17.1

13.3

Inside Principal Cities Outside MetroOfficial SPM‐NGA SPM RPP‐Rent Only RPP Narrow RPP Broad



16.9 16.816.417.417.8

13.2

17.9

12.2

17.5

10.8

17.1

13.3

18.3

15.8

Inside Principal Cities Outside MetroOfficial SPM‐NGA SPM RPP‐Rent Only RPP Narrow RPP Broad Amenities


Correlation of State Level Poverty Rates to Multi‐dimensional Deprivation Index

0.821

0.711

0.755

0.64 0.66 0.68 0.7 0.72 0.74 0.76 0.78 0.8 0.82 0.84

Amenities‐Adjusted (AAI)

Current Methodology (MRI)

Not Geographically Adjusted (NGA)

Correlation to MDDI

Amenities‐Adjusted (AAI) Current Methodology (MRI) Not Geographically Adjusted (NGA)

38Source: 2016 ‐ 2018 CPS ASEC and Glassman, 2019 forthcoming

Correlation of State Level Poverty Rates to Multi‐dimensional Deprivation Index

0.821

0.711

0.755

0.733

0.531

0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9


Current Methodology (MRI)

Not Geographically Adjusted (NGA)

Regional Price Parities ‐ broadly defined

Regional Price Parities ‐ FAR (narrow)

Correlation to MDDI


Correlation of State Level Poverty Rates to Multi‐dimensional Deprivation Index – Individual Dimensions

0.821

0.816

0.561

0.764

0.762

0.390

0.420

Overall Correlation

Official Poverty

Health

Education

Economic Security

Housing Qualilty





0.711

0.575

0.312

0.718

0.539

0.659

0.330

0.821

0.816

0.561

0.764

0.762

0.390

0.420

Overall Correlation

Official Poverty

Health

Education

Economic Security

Housing Qualilty


Current Methodology (MRI) Amenities‐Adjusted (AAI)


Conclusions

• Amenities‐adjusted appears to be more highly correlated with the Multi‐dimensional Deprivation Index

• Suggests that the MRI index may provide an adjustment that is too strong• Need for further research to establish a defensible “adjustment” factor

– Looking at wage/price elasticities – law of one wage

• Open to suggestions as to other mechanisms to incorporate amenities• Need to test against deprivations at the individual level – perhaps using the Survey of Income and Program Participation

42

ReferencesFox, Liana. 2018. Supplemental Poverty Measure: 2017. Current Population Reports P60‐265. September.

Glassman, Brian. Forthcoming. Multidimensional Deprivation in the United States: 2019. U.S. Census Bureau.

Interagency Technical Working Group. 2010. “Observations from the Interagency Technical Working Group on Developing a Supplemental Poverty Measure.” Available at <www.census.gov/hhes/www/poverty/SPM_TWGObservations.

Renwick, Trudi, Eric Figueroa and Bettina Aten. 2017. Supplemental Poverty Measure: A Comparison of Geographic Adjustments with Regional Price Parities vs. Median Rents from the American Community Survey: An Update. SEHSD Working Paper 2017‐36. Paper presented at the 2017 International Statistical Institute World Statistics Congress in Marrakech, Morocco.

Renwick, Trudi, Bettina Aten, Eric Figueroa and Troy Martin. 2014. Supplemental Poverty Measure: A Comparison of GeographicAdjustments with Regional Price Parities vs Median Rents from the American Community Survey. Paper presented at the Allied Social Sciences Association meetings, January 2014.

Renwick, Trudi. 2011. “Geographic Adjustments of Supplemental Poverty Measure Thresholds: Using the American Community Survey Five‐Year Data on Housing Costs. Paper presented at the July 2011 Western Economic Association, San Diego, CA. Available from Census Bureau working papers.

Ziliak, James P. 2011. Cost of Living and the Supplemental Poverty Measure. A Research Forum Submitted to the Office of the Assistance Secretary for Planning and Evaluation U.S. Department of Health and Human Services.

43

Contact Information

• Trudi Renwick– [email protected]– 301‐763‐5133– Census.gov

44


0.755

0.917

0.705

0.702

0.800

‐0.032

0.431

0.711

0.575

0.312

0.718

0.539

0.659

0.330

0.821

0.816

0.561

0.764

0.762

0.390

0.420

‐0.2 0 0.2 0.4 0.6 0.8 1

Correlation to MDDI

OPM

Health

Education

Economic_security

Housing_quality

Neighborhood_quality

Not Geographically Adjusted (NGA) Current Methodology (MRI) Amenities‐Adjusted (AAI)

45Source: 2016 ‐ 2018 CPS ASEC

46

The Multidimensional Deprivation Index Defined

Dimensions How Dimensions are Measured

Standard of living In poverty according to the official poverty measure

Education Aged 19 or older and without a high school diploma or GED; based on head of household educational attainment for those under age 19.

Health Predicted health status is poor. Based on cutoff value of 3 for people under age 65 and 3.5 for people age 65 and over.

Economic security

At least two of the following conditions: lacked health insurance unemployed all of the prior 12 months cumulative hours worked per week for the household was less than 35 hours and no retirement or social security income in the

household.

Housing quality

At least two of the following conditions: lack complete kitchen lack complete plumbing overcrowded housing unit high cost burden.

Neighborhood quality

Lived in a county with at least two of the following: high crime poor air quality poor food environment.

Supplemental Poverty Measure: Alternative Geographic ...

Documents

Transcript of Supplemental Poverty Measure: Alternative Geographic ...