Food for Thought: The ACA Medicaid Expansion and Food Security

By Annika Krovi, J. L. Mann Academy, Greenville, SC, USA

Abstract

The varied adoption of the ACA Medicaid expansion by states has contributed to disparities in care and outcomes, especially for low-income families with dependents. For this reason, this study explored the effect of the Medicaid expansion on food security levels in all households. Prior research established that the Medicaid expansions had significant benefits, but dispute over the true effect on overall food security remained unclear. Through a difference-in-differences analysis using secondary data from the USDA, this study concluded that the Medicaid expansion did not significantly affect food security rates overall. This finding provided clarity on the true effect of the Medicaid expansion and added research that included households with dependents to the existing body of literature. However, future studies may investigate the expansion’s effects in specifically low-income households with dependents in order to better evaluate the outcomes of the population most affected by this issue. This study’s findings, and future research in this field, can be utilized to continue informing public health policies in the U.S. to improve global health and life outcomes for all.

Introduction

In 2010, the Patient Protection and Affordable Care Act (ACA) was enacted in the United States (U.S.). A key aspect of this legislation required the expansion of Medicaid health insurance coverage to all residents above 18 with incomes up to 133% of the federal poverty level. However, after a 2012 Supreme Court ruling, the Medicaid expansion was made optional for states and only twenty-four states and the District of Columbia implemented the expansion on January 1, 2014 (Mitchell & Bencic, 2018). As of November 2022, only forty states and the District of Columbia (considered a state for this study) have expanded Medicaid fully under the ACA. Therefore, in Wyoming, Wisconsin, Kansas, Tennessee, North Carolina, South Carolina, Mississippi, Alabama, Georgia, Florida, and Texas, there is no such expansion in place (KFF, 2022). Due to this disparity, it is essential to study the full effect of the Medicaid expansion to better understand resulting health inequities.

One key metric to focus on is food insecurity. In 2021, the U.S. Department of Agriculture (USDA) found that 10.2% (13.5 million) of U.S. households were food insecure (FI), defined as being “uncertain of having or unable to acquire enough food to meet the needs of all their members”. Moreover, 3.8% of households were marked as having very low food security (VLFS), where “normal eating patterns of one or more household members were disrupted and food intake was reduced” (Coleman-Jensen et al., 2022). Additionally, these authors found that household food insecurity specifically affected 12.5% of households with children in 2021. Therefore, food insecurity has been clearly identified as a major disruptor in many households with children. Literature has consistently found food insecurity to be negatively associated with outcomes for children (Gundersen & Ziliak, 2015; Ke & Ford-Jones, 2015), meaning it is a significant problem that affects global health because disparities in access to food lead to inequity in health outcomes, especially for children.

While many programs exist to address this issue, it is possible that expanding Medicaid coverage has decreased the prevalence of food insecurity amongst low-income households. The ACA Medicaid expansion has already been largely proven to have positive effects on coverage, access to care, and self-reported health (Guth et al., 2022). In regards to food security, a 2019 paper found that the expansion was found to be associated with a “a 12.5% relative reduction” (Himmelstein, 2019). Yet, this finding contrasted with findings from 2018 that indicated that food insecurity rates declined in non-expansion and expansion states at comparable rates (Moran, 2018). Therefore, a gap in current research on this subject was identified. For this reason, this study aimed to explore: do households (including those with children) in states that adopted the ACA Medicaid expansion have higher levels of food security compared to non-expansion states?

Material and Methods

Approach and Alignment:

This study aimed to explore the relationship between the ACA Medicaid expansion (independent variable) and state-level rates of food security in households, including those with children (dependent variable). This study employed a difference-in-differences (DID) method, a quasi-experimental design which uses pre-collected data to estimate a causal effect when certain assumptions are met. Notably, this approach is known to limit the influence of biases that stem from differences in the treatment and control group. Therefore, DID allowed for the use of longitudinal data from before and after the 2014 Medicaid expansion in order to compare the changes in outcomes in non-expansion (control) and expansion (treatment) states (Columbia University, 2023).

The data for this study was obtained from the 2013 and 2017 Current Population Survey (CPS) Food Security Supplement (from the USDA's 'Household Food Security in the United States' annual reports). Food security levels were determined based on questionnaire responses, including eight additional questions for households with children.

Assumptions:

To assume a causal relationship using the DID analysis, four assumptions (as explained by Columbia University, 2023 and Fredriksson & Oliveira, 2019) were met.

First, the Medicaid expansion was not implemented based on specific outcomes of states (supported by Moran, 2018), proving the exchangeability assumption was met.

Next, the parallel trends assumption was satisfied, as Figures 1 and 2 (see below) show that the trends in VLFS and FI were relatively constant before the expansion.

Figure 1: Parallel Trend Assumption for Household Food Insecurity (2007-2013) (Coleman-Jensen et al. 2012, 2014; Nord et al. 2010).

Figure 2: Parallel Trend Assumption for Household Very Low Food Security (2007-2013)  (Coleman-Jensen et al. 2012, 2014; Nord et al. 2010).

Third, the USDA data used for the study was randomly sampled, ensuring stable group composition (Coleman-Jensen et al., 2022). In addition, the following states were excluded from the analyses because their expansion of Medicaid was enacted during the studied time frame but after January 2014: Michigan, New Hampshire, Pennsylvania, Indiana, Alaska, Montana, and Louisiana.

Lastly, the Medicaid expansion in one state did not influence other states' decisions, supporting the assumption of no spillover effects and allowing for causal analysis.

Analytical Procedure:

Variables for the DID analysis included pre and post 2014 ACA Medicaid expansion percentages of food security levels (VLFS and FI) for the years 2011-2013 and 2015-2017, respectively. Two DID tests were performed to determine the relationship between the January 2014 ACA Medicaid expansion and both FI and VLFS. The equation used for the tests was

Y= β0 + β1*[Time] + β2*[Expansion] + β3*[Time*Expansion] +  ε.

Y was the food insecurity level (FI or VLFS) in each group before and after treatment. β0 was the average FI or VLFS rate in the non-expansion group pre-intervention. β1 was the coefficient of the [Time] dummy variable. β2, the coefficient of the [Intervention] dummy variable. The coefficient β3 of [Time*Intervention] gave the estimated difference-in-differences effect. ε was the error term (Columbia University, 2023; Date, 2022; Moran, 2018).

The coefficient β3’s p value reflected the significance of the effect of the Medicaid expansion of FI and VLFS rates. Both the p-value and error term (ε) were used to ensure the replicability of the results. To run this analysis, an unadjusted repeated measures ANOVA test was conducted using the PROC GLM procedure in SAS software (Warton et al., n.d.).

Dr. Elizabeth Adams, an assistant professor at the University of South Carolina Arnold School of Public Health, was consulted as an expert advisor.

Results

Food Insecurity:

Figure 3: Key Output from DID Analysis of the Medicaid Expansion and Food Insecurity.

Very Low Food Security:

Figure 4: Key Output from DID Analysis of the Medicaid Expansion and Very Low Food Security.

Discussion

Major Findings:

For the FI analysis, the magnitude of the difference-in-differences was therefore calculated to be -0.17% (2.09% - 2.26%). Yet, the p-value of 0.7776 on β3 indicated that the change in food insecurity between expansion states and non-expansion states is not significant at the α = 0.05 level. Therefore, there is not enough evidence to prove that the Medicaid expansion caused a decrease in food insecurity.

For the VLFS analysis, similar conclusions were reached. The difference-in-differences value for VLFS was -0.08% (0.75% - 0.83%), but the p-value of 0.7698 on the β3 term was also greater than the α = 0.05 level, indicating that there is no statistically significant causative relationship between the Medicaid expansion and rates of very low food security.

Therefore, based on the results from the DID analysis, the ACA Medicaid expansion did not cause any significant change in FI or VLFS rates in U.S. households, including those with dependents. These findings indicate that households (including those with children) in states that adopted the ACA Medicaid expansion do not have higher levels of food security compared to non-expansion states.

Limitations:

While measures were taken to limit the effects of bias and improper analysis in this study, key limitations can be identified. First, this study did not look strictly at low-income households or only households with children; rather, a holistic approach was taken to ascertain the effects of the expansion on all U.S. households. Doing so may have broadened the population of the sample too far, since the Medicaid expansion only affected adults with incomes up to 133% of the federal poverty level. Therefore, restricting this study to specifically to low-income households with dependents may have yielded different results that would better explain the effect of the expansion on the target population.

However, while conducting this analysis on all U.S. households was a significant limitation, it was also an important step in filling the gap in current understandings of the broader effect of the Medicaid expansion on all U.S. households, while using the resources available.

Implications:

The findings of this study provide clarity on research in this field by supporting the conclusions of Moran (2018) and disputing the findings of Himmelstein (2019). That is, these results imply that the ACA Medicaid expansion did not cause any significant change in food security levels in U.S. households overall, demonstrating that Medicaid may not alter overall food security as significantly as previously believed. However, it is important to note that a more nuanced approach that analyzes changes in food security in smaller subsects of this population due to the expansion is yet to be conducted to further explore this topic.

By analyzing all U.S. households and not just low-income and childless adults, this study provides further insight into the indirect effects of the Medicaid expansion. These findings may help guide lawmakers as they consider the expansion of Medicaid and food assistance programs. While the Medicaid expansion has proven to be significantly beneficial for insurance coverage, health, life outcomes, etc. (Guth et al., 2022), these results show that policymakers cannot rely on this program for reducing overall struggles with food security. Therefore, this analysis proves that further steps need to be taken to address the issue of food insecurity in the US, beyond reliance on the indirect consequences of policies like the Medicaid expansion.

Suggestions for Future Research:

The key suggestion identified by this report for future research on the Medicaid expansion and food insecurity is a narrower focus on low-income households with dependents. Understanding more about the effects of this program on low-income children specifically will help policymakers better understand and address the issue of food insecurity in this vulnerable population.

Moreover, future studies should continue to explore the effects of the ACA Medicaid expansion on various indirect factors, such as maternal, mental, and rural health. As states continue to deliberate on the expansion of Medicaid and programs similar to it, continued research into its broader effects will develop more informed decisions to improve constituent wellbeing.

References

Albouy, D. (n.d.). Program evaluation and the Difference in Difference estimator. Economics 131. Retrieved January 30, 2023, from https://eml.berkeley.edu/~webfac/saez/e131_s04/diff.pdf

Cha, P., & Escarce, J. J. (2022). The Affordable Care Act Medicaid expansion: A difference-in-differences study of spillover participation in SNAP. PloS one, 17(5), e0267244. Retrieved December 3, 2022, from https://doi.org/10.1371/journal.pone.0267244

Coleman-Jensen, A., Nord, M., Andrews, M., & Carlson, S. (2012). (rep.). Household Food Security in the United States in 2011. United States Department of Agriculture. Retrieved February 21, 2023, from https://www.ers.usda.gov/webdocs/publications/45020/30967_err141.pdf?v=1429.9

Coleman-Jensen, A., Gregory, C., & Singh, A. (2014). (rep.). Household Food Security in the United States in 2013. United States Department of Agriculture. Retrieved February 21, 2023, from https://www.ers.usda.gov/webdocs/publications/45265/48787_err173.pdf?v=1633.7

Coleman-Jensen, A., Rabbitt, M. P., Gregory, C. A., & Singh, A. (2018). (rep.). Household Food Security in the United States in 2017. United States Department of Agriculture. Retrieved February 21, 2023, from https://www.ers.usda.gov/webdocs/publications/90023/err-256.pdf?v=3093

Coleman-Jensen, A., Rabbitt, M. P., Hales, L., & Gregory, C. A. (2022, October 17). Overview. USDA ERS. Retrieved December 3, 2022, from https://www.ers.usda.gov/topics/food-nutrition-assistance/food-security-in-the-u-s/

Columbia University Mailman School of Public Health. (2023, January 10). Difference-in-difference estimation. Columbia Public Health. Retrieved January 30, 2023, from https://www.publichealth.columbia.edu/research/population-health-methods/difference-difference-estimation

Date, S. (2022, August 13). Introduction to the difference-in-differences regression model. Time Series Analysis, Regression and Forecasting. Retrieved February 21, 2023, from https://timeseriesreasoning.com/contents/introduction-to-the-difference-in-differences-regression-model/

Fredriksson, A. & Oliveira, G.M.d. (2019), "Impact evaluation using Difference-in-Differences", RAUSP Management Journal, Vol. 54 No. 4, pp. 519-532. Retrieved January 30, 2023, from https://doi.org/10.1108/RAUSP-05-2019-0112

Gundersen, C., & Ziliak, J. P. (2015). Food insecurity and health outcomes. Health Affairs, 34(11), 1830–1839. Retrieved December 3, 2022, from https://doi.org/10.1377/hlthaff.2015.0645

Guth, M., Garfield, R., & Rudowitz, R. (2022). The Effects of Medicaid Expansion under the ACA: Updated Findings from a Literature Review. Kaiser Family Foundation. Retrieved December 3, 2022, from  https://files.kff.org/attachment/Report-The-Effects-of-Medicaid-Expansion-under-the-ACA-Updated-Findings-from-a-Literature-Review.pdf

Himmelstein G. (2019). Effect of the Affordable Care Act's Medicaid Expansions on Food Security, 2010-2016. American journal of public health, 109(9), 1243–1248. Retrieved December 3, 2022, from https://doi.org/10.2105/AJPH.2019.305168

Ke, J., & Ford-Jones, E. L. (2015). Food insecurity and hunger: A review of the effects on children's health and behaviour. Paediatrics & child health, 20(2), 89–91. Retrieved December 3, 2022, from https://doi.org/10.1093/pch/20.2.89

KFF. (2022, November 9). Status of state Medicaid Expansion Decisions: Interactive Map. KFF. Retrieved December 3, 2022, from https://www.kff.org/medicaid/issue-brief/status-of-state-medicaid-expansion-decisions-interactive-map/

Legal Information Institute. (n.d.). National Federation of Independent Business V. Sebelius (2012). Legal Information Institute. Retrieved December 3, 2022, from https://www.law.cornell.edu/wex/national_federation_of_independent_business_v._sebelius_(2012)

Medicaid Eligibility. Medicaid. (n.d.). Retrieved December 3, 2022, from https://www.medicaid.gov/medicaid/eligibility/index.html

Miller, S., & Wherry, L. R. (2019). Four Years Later: Insurance Coverage and Access to Care Continue to Diverge between ACA Medicaid Expansion and Non-Expansion States. AEA Papers and Proceedings, 109, 327–333. Retrieved December 3, 2022, from https://www.jstor.org/stable/26723965

Mitchell, A., & Bencic, S. (2018, December 3). Overview of the ACA Medicaid expansion. Congressional Research Service. Retrieved December 3, 2022, from https://sgp.fas.org/crs/misc/IF10399.pdf

Moran, M. (2018). The Affordable Care Act’s Medicaid Expansion and Food Insecurity Rates, Muskie School Capstones and Dissertations, 136. Retrieved December 3, 2022, from https://digitalcommons.usm.maine.edu/muskie_capstones/136

Nord, M., Coleman-Jensen, A., Andrews, M., & Carlson, S. (2010). (rep.). Household Food Security in the United States, 2009. United States Department of Agriculture. Retrieved February 21, 2023, from https://www.ers.usda.gov/webdocs/publications/44776/7024_err108_1_.pdf?v=4799.2

Ratcliffe, Caroline, et al. How Much Does the Supplemental Nutrition Assistance Program Reduce Food Insecurity? American Journal of Agricultural Economics, vol. 93, no. 4, 2011, pp. 1082–98. JSTOR. Retrieved December 3, 2022, from http://www.jstor.org/stable/41240383

Warton, E. M., Parker, M. M., & Karter, A. J. (n.d.). How D-i-D you do that? basic difference-in-differences models in SAS®. Lex Jansen. Retrieved March 2, 2023, from https://www.lexjansen.com/wuss/2016/49_Final_Paper_PDF.pdf