The Demand for Preventive and Restorative Dental Services among Older Adults

Abstract

Many older Americans have poor access to dental care, resulting in a high prevalence of oral health problems. Because traditional Medicare does not include dental care benefits, only older Americans who are employed, have post-retirement dental benefits or spousal coverage, or enroll in certain Medicare Advantage plans are able to obtain dental care coverage. We seek to determine the extent to which poor access to dental insurance and high out-of-pocket costs affect dental service use by the elderly. Using the 2007–2015 Medical Expenditure Panel Survey and supplemental data on dental care prices, we estimate a demand system for preventive dental services and basic and major restorative services. Selection into dental and medical insurance is addressed using a correlated random effects panel data specification. Consistent with prior studies of the nonelderly population, dental service use was not sensitive to out-of-pocket prices. However, private dental insurance increased preventive service use by 25%, and dental coverage through Medicaid increased basic and major service use by 23% and 36%, respectively. The use of services was more responsive to dental insurance for women than men. These estimates suggest that a Medicare dental benefit could significantly increase dental service use by older Americans.

1. Introduction

Regular dental care is imperative in maintaining good oral health and for diagnosing both oral related problems and symptoms of other systemic conditions (IOM, 2011). Unfortunately, many older Americans do not receive timely, adequate or appropriate dental care. This is largely due to high out-of-pocket costs faced by the elderly and a perceived lack of need (Yarbrough et al. 2014a). The former often results from poor access to dental insurance. Because traditional Medicare does not include dental care benefits, only employed older Americans or those with post-retirement dental benefits or spousal coverage are typically able to obtain dental care coverage.¹ Although some Medicare Advantage plans include dental benefits, coverage is often limited. Additionally, dental coverage is an optional benefit under Medicaid (GAO, 2000a and 2000b). In 2016, 33 states plus the District of Columbia provided dental coverage to adults, 13 states only offered coverage for pain relief or emergency dental services, and 4 states did not provide any coverage (KFF, 2016). In states that do provide Medicaid dental coverage, low levels of dental care use are correlated with low provider payment rates (Decker and Lipton, 2015).

We seek to determine the extent to which poor access to dental insurance and high out-of-pocket costs limit the use of dental services among older Americans. Our approach builds on previous work by Meyerhoefer et al. (2014; hereafter MZM) to estimate the dental care demand among privately insured non-elderly adults. Dental services are classified into one of three categories based on common cost-sharing tiers found in a survey conducted by the National Association of Dental Plans (2008): preventive services (e.g. exams, cleanings, sealants, x-rays), basic restorative services (fillings, extractions, periodontics, endodontics, oral surgery), and major restorative services (crowns, bridges, root canals, dentures). We jointly estimate the probability of preventive dental care and both basic and major restorative dental services (omitting orthodontia services) using a correlated random effects panel data model (Chamberlain, 1980) that controls for endogenous dental and health insurance coverage, endogenous out-of-pocket prices and other control variables.

In addition to using a sample of older Americans, our data and model differ from MZM in several important ways. First, whereas MZM used Medical Expenditure Panel Survey (MEPS) data from 2001–2006, we use the more recent 2007–2015 MEPS. Second, we use a more recent data source to calculate out-of-pocket costs for dental services. Last, we include state-level data to capture varying levels of Medicaid dental coverage and reimbursement generosity by Medicaid and private dental plans.

Consistent with MZM’s findings for the under-65 population, our estimates suggest that dental service use is not responsive to differences in out-of-pocket prices. Earlier foundational studies of the price elasticity of demand for dental care likewise found that dental service use is not responsive to price across a broad range of prices (Mueller and Monheit, 1988; Manning et al., 1986; Hay et al., 1982). We also find that private dental insurance is a significant determinant of preventive service use, and that Medicaid dental coverage increases the demand for basic and major restorative services. Prior studies also suggest that dental service use among the elderly increases as a result of enrollment in dental insurance, but do not concurrently examine Medicaid coverage or price effects (Manski et al., 2011; Kreider et al., 2015; Manski et al., 2015).

2. Methods

2.1. Empirical model

Each equation of our K-equation Probit demand system (K=3 dental services, T=2 years) is specified as

$$m_{kit}={\alpha }_k+\sum _h{\eta }_{kh}{z}_{hit}+\sum _l{\gamma }_{kl}{p}_{lit}+{\beta }_k\log Y_{it}+c_i+{\epsilon }_{kit},$$ (1)

where $m_{kit}$ is a binary variable indicating whether the individual had any preventive dental visits, or episodes of basic and major treatment during year t, $z_{it}$ are sociodemographic variables, $Y_{it}$ is family income, $c_i$ is a stochastic time-invariant individual-specific effect measuring unobserved heterogeneity, and the index $l=1,\mathrm{...},L$ runs over medical services such that $L=K$. We interpret $c_i$ as unobserved oral and physical health status and propensity to consume treatment, which is potentially correlated with dental insurance, prices and several other regressors. Because our model includes a measure of self-reported physical health that varies over time, we believe that it is reasonable to assume the unobserved component of health status is time invariant.

We model prices, $p_{kit}$, using out-of-pocket costs divided by the total payment (i.e. the co-insurance rate).² Following MZM, we use the average out-of-pocket price for preventive care, and note that the average and marginal prices of basic and major care are the same. If the vector of disturbances ${\mathbf{\epsilon }}_{it}=({\epsilon }_{1it},\mathrm{...},{\epsilon }_{Kit}{)}^{\prime }$ is jointly distributed $N(0,{\sigma }_{\epsilon }^2)$, the error terms of the system are correlated through both $c_i$ and ${\mathbf{\epsilon }}_{it}$. In order to account for the heterogeneity parameter we use the correlated random effect (CRE) model originally derived by Chamberlain (1980) and applied to dental care by MZM. Under this specification, the coefficients of all variables specified as correlated with the random effect are identified using variation within individuals over time.

2.2. Data

We pooled panels 12 through 19 of the 2007–2015 Medical Expenditure Panel Survey MEPS and restricted our sample to persons aged 65 and over with Medicare coverage.³ The final sample includes 13,546 persons with two observations each. We merged onto the MEPS mean 3-digit zip-code level dental procedure list price data obtained in 2017 from a Delta Dental on-line pricing tool (Dental Care Cost Estimator, 2017). Appendix figures A1 – A3 present price distributions weighted by MEPS sample weights for 3 representative procedures corresponding to preventive, basic, and major services, respectively: regular cleaning, amalgam 2 surface, and porcelain crown. These prices represent dentists’ list prices and do not reflect discounts negotiated by dental plans.

While prices were observed directly in MEPS for individuals using each of the three types of dental services, we used the Delta Dental data to impute prices for non-users. Specifically, we used an OLS regression-based predictive-mean matching imputation method. This involved separate imputations for each type of service using Stata’s Multiple Imputation (MI) procedures, with 128 imputations for each service. The price imputation regressions contained control variables for sociodemographic characteristics, physical and oral health, medical and dental insurance and deciles of the list price for a standard cleaning from the Delta Dental data.⁴ List charges for other procedures were not predictive of out-of-pocket price, so they were not used for imputation.

The demand model includes several variables that were merged to the MEPS for each year at either the county or state level: the number of the dentists per capita (HRSA, 2018), two composite measures of dental provider reimbursement generosity in 2013 under Medicaid and private plans (Nasseh et al., 2014), median per capita income, unemployment rate, and percent of the population with a bachelor’s degree, (Census, 2016), the annual percent of school-aged children participating in school lunch programs, (USDA, 2017), and a measure of Medicaid managed care health plan penetration, (Mathematica Policy Research, 2017).

Finally, we also constructed from several sources a four category yearly measure of state Medicaid generosity of adult dental coverage for the elderly (McGuinn-Shapiro 2008; Huh, 2017; KFF, 2017; Yarbrough et al., 2014b, 2014c). The generosity levels are: 1) no coverage; 2) emergency services only; 3) partial, limited to some diagnostic, preventive, and minor restorative procedures with an annual per person cap of $1,000 or less; 4) extensive coverage, with a per person cap exceeding $1,000. We combine levels 3 and 4 because preliminary models yielded similar marginal effects for these categories. Indicator variables for generosity of adult Medicaid dental coverage are interacted with an indicator variable for Medicaid enrollment in equation 1. As a result, the marginal effects of coverage are measured relative to the sample of older adults not enrolled in Medicaid.⁵ Appendix table A1 lists descriptive statistics for all the variables used in the demand model and indicates which variables are specified as correlated with the random effect in the CRE specification.⁶

We used the MEPS longitudinal sample weights in all of our analyses and adjusted the standard errors to account for the complex survey design and the first stage price imputations using balanced repeated replications (BRR) (Williams 2000; AHRQ 2017).

3. Results

Table 1 contains information on the use and cost of dental care for the full sample as well as for the sub-samples with different types of insurance coverage. The overall level of dental care utilization among Americans 65 years and older is relatively low. Only 38% of individuals had a preventive care visit during the year, and only 18% and 12% received basic and major dental services, respectively. Clearly, however, there is heterogeneity in access to dental care, with rates of preventive care use varying between 15% and 56% across sub-populations. Furthermore, the rate of preventive dental care use for those enrolled in Medicare and Medicaid (15%) is less than half the rate for seniors with just Medicare (37%), despite the fact that some Medicaid plans include adult dental coverage. Out-of-pocket costs are similar for those with supplemental dental coverage and Medicaid enrollees, but substantially higher for seniors without any dental coverage.

Table 1.

Dental care use and prices (2015 USD).

	Percentage with any use	No. visits conditional on use	Mean $OOP	Mean $Total	$OOP/$Total	Mean coinsurance
Panel A: Full Sample
Preventive	38.2%	1.76	89	147	0.60	0.61
Basic	17.6%	1.94	386	527	0.73	0.68
Major	12.3%	2.25	909	1,299	0.70	0.68
Panel B: Medicare + Private Dental Insurance Coverage (Full or Part of Year)
Preventive	55.8%	1.81	50	153	0.33	0.33
Basic	23.8%	1.97	274	516	0.53	0.47
Major	17.4%	2.12	663	1,270	0.52	0.51
Panel C: Medicare Only (No Private Dental Coverage or Medicaid)
Preventive	37.4%	1.75	106	146	0.73	0.73
Basic	17.2%	1.93	435	536	0.81	0.77
Major	11.7%	2.28	1,055	1,332	0.79	0.77
Panel D: Medicare + Medicaid (Medicaid Dental Coverage Varies)
Preventive	14.8%	1.71	59	119	0.50	0.48
Basic	9.6%	1.92	329	473	0.69	0.54
Major	7.8%	2.44	543	1,120	0.48	0.45

Notes: Out-of-pocket costs for preventive services are the average amount paid per visit, while out-of-pocket costs for basic and major service are the amount paid for the first episode of care.

Average co-insurance rates for those with private dental coverage are much higher than for medical coverage, and likely reflect the fact that many dental plans have relatively low caps on total benefits. We also note that coinsurance levels for those without dental coverage are only between 73% – 77%, which is likely an indication of enrollment in Medicare Advantage plans with dental benefits. It could also reflect under-reporting of private dental coverage.⁷

Elasticity estimates for the key variables in our dental care demand system are reported in table 2, while the full set of coefficients for the model are contained in appendix table A2. An important finding is that dental care demand is not responsive to price for any of the three types of dental services. None of the own-price or cross-price elasticities of demand are precisely estimated, and the point estimates are small in magnitude. The semi-elasticity for full year private dental insurance coverage indicates that having coverage increases the probability of preventive care by 25 percent in the CRE model, but has no detectable effect on the demand for basic or major services.

Table 2.

Elasticity estimates from CRE Probit demand system.

	Full sample			Women only			Men only
	Any Preventive	Any Basic	Any Major	Any Preventive	Any Basic	Any Major	Any Preventive	Any Basic	Any Major
High school diploma	0.395***	0.319***	0.250***	0.330***	0.240***	0.196**	0.464***	0.363***	0.291**
	(0.056)	(0.064)	(0.067)	(0.072)	(0.085)	(0.084)	(0.082)	(0.092)	(0.121)
Some college	0.815***	0.628***	0.558***	0.728***	0.542***	0.530***	0.908***	0.700***	0.543***
	(0.084)	(0.089)	(0.088)	(0.106)	(0.116)	(0.120)	(0.146)	(0.130)	(0.148)
BA degree or higher	1.345***	0.662***	0.464***	1.225***	0.576***	0.286**	1.490***	0.721***	0.615***
	(0.123)	(0.096)	(0.097)	(0.170)	(0.129)	(0.121)	(0.187)	(0.122)	(0.144)
Income	−0.003	−0.030*	0.005	−0.009	−0.047*	−0.003	0.008	0.003	0.027
	(0.012)	(0.018)	(0.017)	(0.013)	(0.024)	(0.023)	(0.016)	(0.024)	(0.035)
Poor/fair health	−0.159***	0.039	0.138**	−0.159***	0.080	0.202**	−0.142**	−0.018	0.060
	(0.045)	(0.060)	(0.068)	(0.048)	(0.079)	(0.087)	(0.068)	(0.084)	(0.084)
Dental insurance, full yr.	0.246**	0.080	0.101	0.430**	0.355	0.385	0.078	−0.165	−0.177
	(0.126)	(0.156)	(0.187)	(0.186)	(0.237)	(0.286)	(0.192)	(0.227)	(0.259)
Dental insurance, partial yr.	0.135	0.158	0.050	0.232	0.373	0.270	0.077	−0.073	−0.155
	(0.093)	(0.147)	(0.158)	(0.155)	(0.229)	(0.264)	(0.139)	(0.176)	(0.198)
Medicaid dental coverage × Medicaid enrollment
No coverage	0.000	0.109	0.112	0.021	−0.078	−0.213	−0.082	0.580	0.457
	(0.179)	(0.293)	(0.383)	(0.182)	(0.360)	(0.351)	(0.290)	(0.424)	(0.492)
Emergency services only	−0.074	0.232	0.247	−0.132	0.345	0.464*	−0.023	0.026	−0.291
	(0.110)	(0.174)	(0.197)	(0.138)	(0.238)	(0.273)	(0.164)	(0.227)	(0.220)
Extensive/partial coverage	0.060	0.233*	0.362**	0.059	0.282*	0.487**	0.061	0.146	0.141
	(0.071)	(0.130)	(0.163)	(0.094)	(0.171)	(0.226)	(0.107)	(0.196)	(0.218)
Private medical insurance	0.070	−0.010	0.067	0.023	−0.071	−0.016	0.098	0.037	0.158
	(0.068)	(0.088)	(0.101)	(0.078)	(0.107)	(0.122)	(0.092)	(0.127)	(0.148)
Medicare Advantage plan	−0.053*	−0.081	−0.040	−0.026	−0.024	−0.016	−0.104**	−0.141*	−0.115
	(0.032)	(0.055)	(0.061)	(0.039)	(0.065)	(0.077)	(0.047)	(0.080)	(0.093)
Preventive treatment price	0.026	0.071	−0.091	0.051	0.129	−0.014	0.013	0.025	−0.173
	(0.081)	(0.109)	(0.132)	(0.101)	(0.146)	(0.162)	(0.128)	(0.155)	(0.193)
Basic treatment price	−0.015	−0.009	0.057	−0.015	0.032	0.038	−0.013	−0.019	0.079
	(0.082)	(0.104)	(0.124)	(0.096)	(0.144)	(0.167)	(0.116)	(0.152)	(0.195)
Major treatment price	−0.026	0.069	0.068	0.000	0.089	0.099	−0.059	0.040	0.017
	(0.067)	(0.105)	(0.125)	(0.096)	(0.134)	(0.148)	(0.094)	(0.154)	(0.185)
Number of observations (N × T)	27,092	27,092	27,092	15,472	15,472	15,472	11,620	11,620	11,620

Notes: Standard errors in parentheses are adjusted for the complex design of the MEPS and first stage price imputation.

Significance level:

^***p < 0.01,

^**p < 0.05,

^*p < 0.1.

Public dental coverage through Medicaid is also an important determinant of dental care use, but only for basic and major care and only when states offer partial or more extensive coverage. Being enrolled in a Medicaid plan with partial or more generous dental coverage increases the likelihood of using basic restorative services by 23% and major services by 36% relative to not being enrolled in Medicaid. Finally, we find that dental service use among the elderly is strongly affected by education level and health status.

When we estimate our models separately by gender we find that the responsiveness of dental service use to private dental insurance and Medicaid dental benefits identified in the full sample is due to a demand response to insurance coverage by women. Although the failure to find any demand response to insurance among men could be due to a smaller sample size, the precision of the estimates increases significantly in the sample of men, while the sample size is only 25% smaller than for women.. However, we do find that the marginally significant negative effect of enrollment in a Medicare Advantage plan on preventive service use estimated in the full sample is driven by men. In appendix table A3 we report results stratified by age (66–74 and 75+), which show that the demand response to Medicaid dental benefits is only precisely estimated in the younger age group.

4. Discussion

Similar to prior research on working age adults by MZM, we find that dental service use among seniors is not responsive to out-of-pocket costs. We also find that the use of preventive dental care is increased by enrollment in private dental coverage to a similar extent for seniors as previous studies have found for working age adults. For example, we estimate that the semi-elasticity of private dental insurance is 0.25, while MZM estimate an elasticity of 0.23. In contrast to MZM, however, we do not find that private dental insurance increases the use of basic or major dental services.

The finding that consumers are not price-sensitive, conditional on insurance enrollment, is in contrast to what has been found for medical care (see, for example, Ellis et al. 2017; Meyerhoefer and Zuvekas 2010). Nonetheless, MZM note that even older studies of dental care demand found that consumers were not price-sensitive over a broad range of out-of-pocket costs. While this could be related to multiple factors, one difference between medical and dental care is that individuals often seek the latter when an oral health problem results in significant pain (Sun et al. 2015; Vargas et al. 2003; Devaraj and Eswar 2012). As we discuss below, it is also possible that measurement error in prices could make it difficult to detect low levels of price responsiveness.

We are not aware of any prior studies that evaluate the impact of Medicaid enrollment on the demand for dental services using a comprehensive demand framework. As a result, the fact that we find that Medicaid is an important determinant of the demand for basic and major services is notable. From a policy standpoint, a key issue is whether incorporating a dental benefit into the Medicare program would increase the demand for dental services and potentially address a currently unmet need for dental care among the elderly. Taken together, our estimates of the effect of public and private dental insurance suggest that a Medicare dental benefit would increase the use of dental care among the elderly.

Our study has some limitations that should be acknowledged. First, we impute prices for MEPS respondents without dental care use, which could introduce measurement error in the price variable. We do note, however, that when we estimate our model on the sub-sample of individuals with use that the coefficients on price are still small and statistically insignificant. Additional measurement error could arise from inaccurate self (or proxy)-reporting of out-of-pocket dental payments in the MEPS. Second, our empirical approach to address the endogeneity of price and insurance coverage does not account for time-varying unobservable factors. Finally, we find evidence that some MEPS respondents may have failed to report their dental coverage, and such under-reporting could attenuate the estimated effects of private dental coverage on the demand for dental services.

Appendix

Figure A1.

Price distribution for regular cleaning.

Figure A2.

Price distribution for amalgam 2 surface, silver.

Figure A3.

Price distribution for porcelain crown.

Table A1.

Descriptive statistics.

	Full Sample		Medicare + Private Dental Insurance		Medicare Only		Medicare + Medicaid
	Mean	Std. Err	Mean	Std. Err	Mean	Std. Err	Mean	Std. Err
Any Preventive visit	0.382	0.007	0.558	0.013	0.374	0.008	0.148	0.008
Any Basic visit	0.176	0.003	0.238	0.009	0.172	0.004	0.096	0.006
Any Major visit	0.123	0.003	0.174	0.007	0.117	0.003	0.078	0.006
Number of Preventive visits	0.674	0.014	1.007	0.030	0.654	0.016	0.253	0.019
Number of Basic visits	0.341	0.010	0.469	0.026	0.333	0.011	0.185	0.016
Number of Major visits	0.276	0.009	0.369	0.020	0.265	0.011	0.190	0.019
Demographic variables
Hispanic	0.072	0.005	0.043	0.005	0.055	0.004	0.214	0.018
Black	0.087	0.005	0.086	0.008	0.072	0.005	0.172	0.012
Asian	0.037	0.006	0.041	0.013	0.026	0.004	0.094	0.011
White	0.787	0.010	0.811	0.018	0.832	0.008	0.497	0.020
Female	0.564	0.004	0.507	0.009	0.563	0.005	0.661	0.010
Age 65 – 74	0.532	0.008	0.657	0.013	0.505	0.009	0.483	0.013
Age75 – 84	0.334	0.006	0.267	0.012	0.353	0.007	0.338	0.012
Age 85+	0.134	0.005	0.076	0.007	0.142	0.006	0.179	0.011
Less than high school degree	0.192	0.006	0.085	0.006	0.175	0.006	0.460	0.016
High school diploma	0.337	0.007	0.280	0.010	0.365	0.008	0.266	0.014
Some college	0.217	0.005	0.248	0.011	0.222	0.006	0.144	0.008
BA degree or higher	0.247	0.008	0.384	0.014	0.233	0.008	0.112	0.011
Married	0.555	0.009	0.687	0.014	0.562	0.010	0.302	0.016
Household size	1.893	0.014	1.929	0.030	1.872	0.014	1.959	0.039
Urban residence	0.814	0.017	0.874	0.016	0.797	0.019	0.816	0.019
Northeast census region	0.190	0.009	0.198	0.015	0.179	0.009	0.242	0.017
Midwest census region	0.224	0.010	0.242	0.016	0.235	0.011	0.138	0.012
South census region	0.371	0.013	0.319	0.019	0.385	0.014	0.372	0.020
West census region	0.214	0.009	0.241	0.020	0.201	0.010	0.248	0.019
Any daily living limitation	0.230	0.005	0.160	0.008	0.218	0.006	0.413	0.013
Any activity limitation	0.203	0.005	0.139	0.008	0.187	0.006	0.397	0.014
No. of chronic conditions	3.085	0.024	2.941	0.042	3.047	0.030	3.530	0.055
No teeth	0.216	0.005	0.124	0.007	0.216	0.005	0.361	0.012
Main respondent	0.424	0.005	0.430	0.010	0.421	0.006	0.429	0.013
Ln(HH income / sqrt(HH size))*	10.181	0.023	10.623	0.034	10.191	0.023	9.420	0.044
Poor/fair physical health*	0.289	0.005	0.212	0.008	0.267	0.005	0.532	0.012
Year 2007	0.057	0.002	0.055	0.004	0.058	0.003	0.060	0.005
Year 2008	0.116	0.003	0.109	0.006	0.120	0.004	0.107	0.006
Year 2009	0.112	0.003	0.105	0.005	0.116	0.003	0.102	0.006
Year 2010	0.117	0.003	0.108	0.005	0.121	0.003	0.109	0.006
Year 2011	0.129	0.003	0.128	0.006	0.130	0.003	0.126	0.007
Year 2012	0.133	0.003	0.137	0.006	0.131	0.003	0.142	0.007
Year 2013	0.130	0.003	0.134	0.006	0.127	0.003	0.140	0.007
Year 2014	0.134	0.003	0.143	0.007	0.131	0.004	0.136	0.006
Year 2015	0.072	0.002	0.082	0.005	0.068	0.003	0.079	0.005
Insurance coverage and dental service prices
Dental insurance, full year*	0.154	0.005	0.800	0.009	-	-	-	-
Dental insurance, partial year *	0.038	0.002	0.200	0.009	-	-	-	-
Medicaid dental coverage × Medicaid enrollment
No coverage*	0.009	0.001	0.003	0.001	0.000	0.000	0.071	0.010
Emergency services only*	0.034	0.003	0.008	0.002	0.000	0.000	0.267	0.020
Extensive/partial coverage*	0.083	0.005	0.016	0.003	0.000	0.000	0.662	0.022
Private medical insurance*	0.522	0.008	0.952	0.007	0.473	0.010	0.118	0.009
Medicare Advantage plan*	0.403	0.009	0.326	0.016	0.425	0.010	0.396	0.015
Preventive treatment price*	0.602	0.005	0.322	0.007	0.689	0.005	0.557	0.006
Basic treatment price*	0.661	0.004	0.457	0.007	0.725	0.003	0.620	0.005
Major treatment price*	0.665	0.003	0.511	0.005	0.719	0.003	0.598	0.005
State/area –level variables
No. dentists per 1000 capita	0.596	0.009	0.630	0.012	0.583	0.009	0.617	0.013
HPSA dental shortage area	46.402	0.635	46.546	0.937	46.306	0.594	46.720	0.750
State Medicaid reimbursement	106.243	0.482	106.922	0.619	105.768	0.500	107.876	0.572
State Private reimbursement	7.798	0.075	7.860	0.102	7.750	0.079	7.974	0.076
State unemployment rate	10.663	0.007	10.666	0.009	10.656	0.007	10.696	0.009
State per capita income	28.528	0.165	28.953	0.195	28.372	0.176	28.739	0.218
State % with BA degree	0.561	0.003	0.550	0.004	0.564	0.004	0.556	0.004
State NSLP participation	0.725	0.008	0.735	0.011	0.726	0.007	0.707	0.009
No. of observations (N × T)	27,092		4,664		17,384		5,044

Notes: The variables with an asterisk are assumed to be correlated with the random effect in the CRE model.

Table A2.

Coefficient estimates from CRE Probit demand system.

	Any Preventive		Any Basic		Any Major
	Coeff.	S.E.	Coeff.	S.E.	Coeff.	S.E.
Demographic variables
Constant	−0.846	1.013	−2.228	1.602	−3.503***	1.210
Hispanic	−0.203***	0.057	−0.052	0.043	−0.061	0.052
Black	−0.459***	0.053	−0.210***	0.044	−0.133***	0.051
Asian	−0.272***	0.062	−0.018	0.055	0.028	0.058
Female	0.192***	0.023	−0.006	0.026	0.026	0.031
Age75 – 84	0.057*	0.029	0.055*	0.030	0.026	0.030
Age 85+	0.023	0.051	0.110**	0.047	−0.030	0.060
High school diploma	0.321***	0.042	0.203***	0.038	0.143***	0.037
Some college	0.558***	0.044	0.336***	0.043	0.290***	0.040
BA degree or higher	0.796***	0.048	0.360***	0.045	0.247***	0.046
Married	0.332***	0.033	0.155***	0.033	0.139***	0.036
Household size	−0.159***	0.021	−0.103***	0.020	−0.104***	0.020
Urban residence	0.066	0.050	0.006	0.044	0.022	0.040
Northeast region	−0.054	0.057	−0.023	0.053	0.025	0.055
Midwest region	0.119**	0.051	0.085*	0.046	0.020	0.049
South region	−0.039	0.057	−0.004	0.043	−0.019	0.044
Any daily living limitation	−0.046	0.033	−0.091***	0.034	0.023	0.045
Any activity limitation	−0.144***	0.037	0.000	0.039	−0.018	0.039
No. chronic conditions	0.010	0.007	0.026***	0.007	0.017**	0.008
No teeth	−1.128***	0.043	−0.686***	0.036	−0.228***	0.037
Main respondent	−0.154***	0.027	−0.130***	0.030	−0.089***	0.025
Log family income	−0.002	0.010	−0.016	0.012	0.003	0.010
Poor/fair health	−0.136***	0.038	0.047	0.038	0.080**	0.038
Year 2008	0.052	0.060	0.060	0.058	0.016	0.060
Year 2009	−0.052	0.075	−0.008	0.078	−0.030	0.083
Year 2010	−0.098	0.074	−0.015	0.084	0.001	0.088
Year 2011	−0.104	0.074	−0.097	0.078	−0.055	0.077
Year 2012	−0.154*	0.079	−0.132*	0.077	−0.027	0.081
Year 2013	−0.061	0.064	−0.011	0.065	−0.019	0.071
Year 2014	0.031	0.062	−0.031	0.063	0.006	0.065
Year 2015	0.038	0.067	−0.044	0.065	−0.070	0.072
Insurance coverage and dental service prices
Dental insurance, full year	0.195**	0.093	0.063	0.096	0.058	0.104
Dental insurance, partial year	0.111	0.073	0.079	0.085	0.029	0.089
Medicaid dental coverage × Medicaid enrollment
No coverage	0.000	0.154	0.055	0.172	0.063	0.205
Emergency services only	−0.065	0.099	0.122	0.098	0.135	0.100
Extensive/partial coverage	0.051	0.059	0.181**	0.074	0.194**	0.079
Private medical insurance	0.060	0.056	0.058	0.057	0.039	0.059
Medicare Advantage plan	−0.046*	0.028	−0.038	0.036	−0.024	0.036
Preventive treatment price	0.022	0.070	0.034	0.070	−0.054	0.077
Basic treatment price	−0.013	0.070	0.040	0.067	0.033	0.073
Major treatment price	−0.023	0.058	0.121*	0.067	0.040	0.073
State/area-level variables
No. dentists per 1000 capita	0.037	0.081	0.106*	0.056	0.078	0.073
Medicaid reimbursement rate	−0.002	0.002	−0.001	0.001	−0.002	0.001
Private reimbursement rate	−0.002	0.003	0.000	0.003	0.003	0.003
Unemployment rate	0.021*	0.012	0.009	0.012	0.003	0.015
Per capita income	−0.292	0.419	0.268	0.696	0.659	0.484
Percentage with BA degree	0.021***	0.006	0.009	0.007	−0.005	0.008
NSLP participation	0.339	0.299	−0.096	0.279	−0.320	0.308
Medicaid managed care penetration	−0.232***	0.085	−0.101	0.073	0.098	0.080

Correlated random effect parameters
Log family income λ1	0.016*	0.009	Log family income λ2	0.035***	0.009
Poor/fair health λ1	−0.037	0.031	Poor/fair health λ2	−0.090***	0.035
No Medicaid dental coverage λ1	−0.250	0.174	No Medicaid dental coverage λ2	−0.424***	0.139
Emergency Medicaid coverage λ1	−0.076	0.077	Emergency Medicaid coverage λ2	−0.116	0.093
Extensive/partial Medicaid coverage λ1	−0.091	0.066	Extensive/partial Medicaid coverage λ2	−0.090	0.064
Private medical insurance λ1	0.073	0.048	Private medical insurance λ2	0.007	0.048
MA plan λ1	0.065**	0.027	MA plan λ2	0.009	0.026
Dental insurance, full year λ1	−0.074	0.083	Dental insurance, full year λ2	0.166*	0.093
Dental insurance, partial year λ1	−0.018	0.068	Dental insurance, partial year λ2	0.049	0.077
Preventive treatment price λ1	−0.011	0.056	Preventive treatment price λ2	0.023	0.061
Basic treatment price λ1	0.048	0.068	Basic treatment priceλ2	−0.009	0.064
Major treatment price λ1	0.022	0.057	Major treatment price λ2	−0.016	0.056

Notes: Standard errors are adjusted for the complex design of the MEPS and first stage price imputation.

Significance level:

^***p < 0.01,

^**p < 0.05,

^*p < 0.1.

Table A3.

Elasticity estimates from CRE Probit demand system by age.

	Age 65–74						Age 75 +
	Any Preventive		Any Basic		Any Major		Any Preventive		Any Basic		Any Major
	Elasticity	S.E.	Elasticity	S.E.	Elasticity	S.E.	Elasticity	S.E.	Elasticity	S.E.	Elasticity	S.E.
High school diploma	0.427***	0.073	0.210**	0.091	0.141	0.100	0.386***	0.082	0.413***	0.088	0.304***	0.098
Some college	0.736***	0.111	0.535***	0.114	0.442***	0.112	0.911***	0.145	0.692***	0.142	0.563***	0.137
BA degree or higher	1.378***	0.164	0.641***	0.125	0.429***	0.125	1.221***	0.174	0.666***	0.133	0.397***	0.132
Income	0.005	0.015	−0.003	0.023	−0.013	0.025	−0.004	0.016	−0.038	0.024	0.031	0.027
Poor/fair health	−0.183***	0.065	−0.055	0.090	0.024	0.098	−0.111*	0.059	0.129	0.090	0.264**	0.110
Dental and medical insurance
Dental insurance, full yr.	0.190	0.139	0.110	0.188	0.223	0.244	0.445	0.291	0.177	0.308	0.125	0.334
Dental insurance, partial yr.	0.177	0.119	0.110	0.160	0.042	0.187	0.162	0.169	0.357	0.298	0.366	0.325
Medicaid dental coverage × Medicaid enrollment
No coverage	0.274	0.229	0.149	0.234	0.491	0.419	−0.031	0.192	0.188	0.384	−0.076	0.482
Emergency services only	−0.057	0.141	0.237	0.239	0.081	0.263	−0.048	0.206	0.385	0.339	0.554	0.340
Extensive/partial coverage	0.078	0.092	0.393**	0.182	0.575**	0.234	0.078	0.122	0.116	0.182	0.245	0.216
Private medical insurance	0.185**	0.081	0.128	0.108	0.194	0.124	−0.080	0.095	−0.174	0.116	−0.084	0.145
Medicare Advantage plan	−0.031	0.048	−0.085	0.065	−0.010	0.074	−0.074	0.055	−0.078	0.081	−0.073	0.091
Dental service prices
Preventive treatment price	0.013	0.099	0.063	0.123	−0.152	0.159	0.051	0.123	0.079	0.170	−0.046	0.201
Basic treatment price	−0.039	0.091	−0.029	0.123	0.062	0.163	0.018	0.127	0.033	0.158	0.046	0.191
Major treatment price	−0.019	0.081	0.069	0.144	0.061	0.155	−0.036	0.105	0.063	0.158	0.078	0.187

Notes: Standard errors are adjusted for the complex design of the MEPS and first stage price imputation.

Significance level:

^***p < 0.01,

^**p < 0.05,

^*p < 0.1.