Media & the Politics of Implementation: Explaining Geographic Variation in Local TV Coverage of the ACA
- 1. Media & the Politics of Implementation: Explaining Geographic Variation in Local TV Coverage of the ACA Erika Franklin Fowler & Laura Baum, Wesleyan Colleen L. Barry, John Hopkins Bloomberg Jeff Niederdeppe, Cornell Pinar Karaca-Mandic, Andrew Wilcock & Sarah E. Gollust, UMN
- 2. Funding from: • Robert Wood Johnson Foundation through its State Health Care Access Reform Evaluation (SHARE), Grant ID# 72179 • Wesleyan University
- 3. News as Commodity, not a Mirror • News is a highly selective version of reality (Lippmann 1922; Gans 1979; Hamilton 2004) • Much of political science assumes information transmission (see Althaus et al. 2011) • Even within communication, news production takes a backseat to effects
- 4. Key Insight: Coverage Unlikely to be Uniform • Audience characteristics • Ease of access to story material • Ownership characteristics • Geographic proximity to story • Congruence between media and political boundaries • In other words, production is variable and complex
- 5. Competing Interests Shaping ACA Coverage • News producers • Politicians • Organized interests Interested here in predicting volume (content focus later)
- 6. Data & Measures • Closed captioned searches of local TV news from 10/1/13-4/15/14 • Key words: “health care,” “healthcare,” “Obama care,” “Obamacare” and “Affordable Care Act” • 685 stations in 208 media markets, all evening news (4:30-11:59pm) • American Community Survey (ACS) • Wesleyan Media Project political and insurance advertising data • RWJF insurance market information • BIA Kelsey Station Ownership Information
- 7. Volume of Evening Local TV ACA Hits Oct 1, 2013 – Apr 15, 2014
- 8. Local TV News Hits by Station 02468 FrequencyofStations 0 200 400 600 Number of Closed Captioning Hits
- 9. Analytic Approach • Negative binomial count models • Clustering by state, robust SEs (robust to hierarchical) • All results from one model
- 10. Factors Shaping News: Audience Interest Greater interest in the ACA in a market should drive coverage up • Markets with more uninsured: more coverage • Markets with more direct purchasers: more coverage • Markets with high poverty: unclear • Partisanship: unclear • Political competition: more coverage • Controls for news consumer demographics: age, sex, education, race
- 11. Factors Shaping News: Audience Interest Coeff. (S.E.) Uninsured (% 18-64 NCI population) 1.278** (0.597) Indiv. health ins. enrollees (% 18-64 NCI pop) -0.206 (1.414) Poverty (%) -3.093** (1.316) White (%) -0.227 (0.205) College grad or more (%) 0.982 (0.734) Male (%) -17.73*** (3.353) Ages 20-64 (%) 1.070 (1.684) Ages 65+ (%) -1.107 (1.032) Obama vote in DMA (%) 0.175 (0.266) Margin of victory (50 – abs(Obama vote)) 0.214 (0.415)
- 12. Predicted Count by Proportion Uninsured 100150200250 PredictedNumberofACAHits .04 .08 .12 .16 .2 .24 .28 .32 Proportion of 18-64 CNI Population w/o Health Insurance (ACS 2013 5-Yr Est)
- 13. Ease of Access to Sources & Story Materials Availability of materials for ACA stories should increase coverage • Accessibility of spokespeople should increase coverage • Volume of ACA insurance and political ads should increase coverage • Type of insurance marketplace: state-based should have more stories
- 14. Ease of Access to Sources & Story Materials Coeff. (S.E.) Consumer assistance funds (per uninsured) -0.00415 (0.00263) Insurance ad volume (000s) 0.0195** (0.00900) Political ad volume (000s) 0.0642 (0.0417) Partnership exchange 0.200*** (0.0665) State-based exchange 0.162** (0.0711)
- 15. Predicted Count by Exchange Type 140160180200220 PredictedNumberofACAHits Federally-Facilitated Partnership State-Based
- 16. Station & Ownership Characteristics Stations are enabled and constrained by their network affiliates and ownership groups choices and priorities for programming • Fox affiliates (more sports, fewer newscasts): less coverage • Ownership profits: unclear
- 17. Station & Ownership Characteristics Coeff. (S.E.) CBS affiliate 0.0293 (0.0415) Fox affiliate -0.228*** (0.0473) NBC affiliate 0.162*** (0.0417) Second tier (revenue) -0.114** (0.0490) Lowest tier (revenue) -0.158*** (0.0483)
- 18. Predicted Count by Owner Revenue Tier 150160170180190200 PredictedNumberofACAHits Largest Profits Mid-Tier Profits Lowest Profits
- 19. Structural Factors Affecting Production Structural set-up of media markets (population served and geographic congruence) matter • Number of states within DMA: decrease coverage • Population of the media market: decrease coverage • Insurer competition: increase coverage (expect negative sign)
- 20. Structural Factors Affecting Production Coeff. (S.E.) # of states within DMA 0.0316 (0.0203) DMA total population (Ms) -0.0321* (0.0174) HHI (insurance competition) -2.02e-05 (1.28e-05)
- 21. Contribution & Future Directions • Variance in news coverage of the ACA is predictable • Systematic analysis of local TV across the country • Future work • Over time • Assessing content