2016 Survey of Health Insurance Marketplace Assister Programs and Brokers
The Kaiser Family Foundation 2015 Survey of Health Insurance Marketplace Assister Programs and Brokers was designed and analyzed by KFF researchers and administered by Davis Research. This nationwide survey was conducted through an online questionnaire from February 11, 2016 through March 4, 2016.
To recruit Assister Program survey participants, we asked officials CMS and from States operating SBM or FPM Marketplaces to provide contact information for the directors of their certified Assister Programs. In addition, we requested contact information for the directors of enrollment assistance activities in each of the FQHCs from HRSA. All Assister Programs received an email with a link to the survey inviting the director to participate. In the event the person receiving the survey was not the appropriate person to complete it, they were asked to provide the contact name and email for the appropriate person within their organization.
To analyze results, we assigned Assister Programs to one of four types based on their primary source of funding. The first type, Navigators, were those identified by Marketplace officials contracted with and received grant funding directly from the Marketplace. The second type, FEAP, were those identified by CMS as contractors that operate in certain FFM states and that otherwise act as Navigators. We tracked FEAP responses separately in the survey, but for most data analysis presented in this report we combined responses of FEAPs and Navigators. The third type, FQHCs, were those that received grant funding from HRSA to provide enrollment assistance. We identified FQHCs using the contact list provided by HRSA. A small percentage of FQHC Programs receive both HRSA grants and Marketplace Navigator grant funding; these were categorized as Navigators for our analysis. All other Assister Programs certified to provide assistance in Marketplaces were designated as CACs.
A total of 5,094 Programs were invited to participate in the study, and 688 Programs responded and were included (for a response rate of 13.5%). Because response rates varied by Program type, data were weighted to reflect the distribution in the initial sample by Program type and Marketplace type; for our analysis, FFM and FPM Marketplaces were grouped together. (FFM + FPM, and SBM). Weighted and unweighted proportions of the final sample by Program type are shown in the table below.
|Unweighted % of total
|Weighted % of total
Using responses provided by Assister Programs in the study, we were able to estimate the number of Assister Program staff and the number of consumers they helped with eligibility and enrollment in Medicaid/CHIP and Qualified Health Plans during the second Open Enrollment period nationwide, by extrapolating response data to the national level. Survey participants were asked to provide the number of full-time equivalent Assisters in their Program and the number of consumers helped. Respondents who did not provide a numeric value for the number of consumers helped were asked to estimate a number using a range of options. In making our calculation, we used the midpoint value for responses that provided a range of numbers of consumers helped. Non-responses were imputed based on the type of Assister Program. A limitation of our national-estimates methodology is that outliers in our response data (i.e. assister programs that helped over 10,000 people during open enrollment, or who had more than 100 staff), when extrapolated to the national level may have an outsize influence on our estimates of total helped and total assister staff nationwide.
We also surveyed the work of Assister Programs outside of Open Enrollment as they helped people apply for Special Enrollment Periods, report mid-year changes to the Marketplace, and resolve post-enrollment problems. Using response data provided by returning Assister Programs, we were able to estimate the number of people nationally who received help from Assister Programs between the first and second Open Enrollment periods with each of these types of issues.
To recruit brokers in the Federally-Facilitated Marketplace (FFM) states, we obtained contact information from a file of brokers in the FFM states, made publicly available through healthcare.gov.1 To obtain broker contact information from the SBM and FPM states, we asked Marketplaces to provide contact information, and when that was not provided, compiled contact information that was publicly available on Marketplace websites. As we estimate that there are tens of thousands of brokers selling non-group Marketplace policies nationwide, we drew a sample of 9,432 brokers based on their distribution by Marketplace type (FFM, FPM, or SBM). Our general sampling rule was to randomly select 10% of all contacts in each state; we oversampled in ten states where we had fewer than 500 contacts to begin with. Because we did not have a complete sample of Marketplace brokers in all states, we were not able to compute national estimates of the numbers of consumers helped by brokers.
Out of the 9,432 brokers who were invited to participate in the study, 418 responded and were included (for a response rate of 4%).
Toplines and Margin of Sampling Error
Survey toplines with overall frequencies of both Assister Programs and Brokers for all survey questions are available at https://www.kff.org/health-reform/report/2016-survey-of-health-insurance-marketplace-assister-programs-and-brokers.
The sample size and margin of sampling error (MOSE) for the total sample and key subgroups of Assister Programs are shown in the table below. All statistical tests of significance account for the effect of weighting.
|+/-4 percentage points
|+/-5 percentage points
|+/-7 percentage points
|Navigator and FEAP
|+/-8 percentage points
|+/-5 percentage points