Challenges of Reaching Low Income Households

Transcription

1 Challenges of Reaching Low Income Households Utilizing Cell Phone Rate Center Information in Targeting Low Income Populations Eran Ben Porath David Dutwin Mollyann Brodie ACA leads to increased interest in reaching: HHs under FPL Under FPL Uninsured Correlated with low income Poverty is fairly clustered, but in small areas Low income respondents, more likely to be CPO Landline Solution to Targeting Low Income HHs Sample providers infer demographics by linking exchanges to Census information Allows estimation of income levels for any given exchange Proven track record Can be crossed with directory listings With caution Can Cell Phones Be Targeted? No equivalent exchange to Census data correspondence Known issue with targeting small geographies Billing information usage is ill advised 1

2 The Possibilities of Cell Phone Rate Centers What are rate centers? Billing geographies that (to oversimplify) demarcated the lines between local and long distance calls. Cell phone providers assign every phone number to a home rate center based on the point of purchase of a phone (usually). The Possibilities of Cell Phone Rate Centers The Possibilities of Cell Phone Rate Centers What information can be estimated by rate center? Rate centers can be associated with likely county Census information pertaining to counties, can serve as proxy for rate center information: Prevalence of race/ethnicity Prevalence of income groups Utilizing Rate Centers Research Questions 1. Can rate center information be exploited to increase incidence of low income HHs? 1a. How low? 1b. How consistently 2. How do rate centers compare to landline exchanges in effectiveness? 3. Can this approach be utilized to reach populations such as the unisnured? 2

3 Methods Studies Two KFF surveys Low Income State N=8,762 (CA, TX, MO, Else) Callbacks: 684 LL: 4,170 Cell: 3,908 CA Uninsured: N=2,001 Callbacks and Income Listed HH: 317 LL: 773 Cell: 911 Study 1 Low Income States Design Designed to reach set Ns by three income levels ( FPL; 138% 399% FPL; FPL+) Definitions of strata vary slightly by state 5 LL strata, based on poverty: 5 Cell Phone strata, based on rate center demos Definitions of strata vary slightly by state Total Respondents California VH HI MED LOW VH HI MED LOW V. LOW V. LOW VH HI MED LOW V. LOW VH HI MED LOW V. LOW 3

4 Missouri Texas VH HI MED LOW VH HI MED LOW V. LOW V. LOW VH HI MED LOW VH HI MED LOW V. LOW V. LOW All Other States Study 1 Low Income States Findings Cell phone rate centers were effective in reaching lower income populations The share of low income households was significantly highest in the VH stratum; and higher in the Hi than in the medium The effect took place within multiple geographies The effect is less pronounced in the lower incidence strata In the high strata: the effect is near identical for LLs and cell phones Overall the effect is stronger across five strata for LLs (γ) Cell phones, overall, reach more LI overall an effective way of reaching LI population Need less elaborate stratification VH HI MED LOW V. LOW VH HI MED LOW V. LOW 4

5 Study 2: CA Uninsured Design 4 LL strata, based on poverty: High Poverty: median HH income (in exchange) <$40k Medium Poverty: median HH income (in exchange) $40k to <$60k Low Poverty: median HH income (in exchange) $60k to <$100k Very Low Poverty: median HH income (in exchange) 100K+ 3 Cell Phone strata, based on rate center demos: High Poverty: numbers linked to rate centers where 3 or more of population have annual income <$25k Medium Poverty: numbers linked to rate centers where <3 of population have annual income <$25k Low Poverty: numbers linked to rate centers where less than of population have annual income <$25k Study 2 CA Uninsured Effectiveness Target population CA Uninsured y.o. LL Incidence: 3 Cell phone incidence: 35% Uninsurance among target population: 25% (ACS, 2012) Number of Screens Insured Uninsured Landline Cell Study 2 CA Uninsured Effectiveness Study 2 CA Uninsured Reaching the Harder to Reach Uninsured Incidence Landline Cell LL Hi Pov LL Med Pov LL Lo Pov LL V. Lo Pov Cell Hi Pov Cell Med Pov Cell Lo Pov For both LL and Cell Phone, uninsurance in the high stratum was significantly higher than in any other stratum (surpassing 4) While incidence is maintained: cell phone targeting allows reaching younger (and more male) uninsured 5

6 Utilizing Cell Phone Rate Centers to Effectively Reach the Uninsured Conclusion Cell phones are essential for reaching the uninsured Particularly, to reach young and male respondents Stratifying by cell phone rate center classes, with income as a proxy for uninsurance, allows targeting on par with landline stratification Cell phone stratification is a bit less effective within the lower incidence areas: employ fewer strata Contact Information Eran Ben-Porath SSRS ebenporath@ssrs.com David Dutwin SSRS ddutwin@ssrs.com Mollyann Brodie Kaiser Family Foundation MollyB@kff.org 6