By Molly Warren | May 2016
Building on a successful inaugural year of the Get Covered Connector, an online appointment-scheduling tool for consumers seeking help applying for health coverage, Enroll America worked to refine the tool’s functionality for assisters, added new training focusing on ways to leverage the tool for maximum impact, and expanded its reach during the third open enrollment period (OE3).1,2
Overall, 2,000 in-person assisters at 400 organizations across 38 states offered nearly 1 million assistance appointments which led to 97,000 appointments scheduled, 62,000 recorded appointments attended, and nearly 40,000 recorded enrollments during OE3.3 This report describes the changes and growth of the Connector in OE3, as well as insights for the enrollment community on consumers’ behaviors and needs regarding in-person assistance. This report also provides an early look at the success of the Connector at engaging limited-English-proficient and rural populations, two groups that remain uninsured at a disproportionately high rate.
Changes and Growth of the Connector in OE3
After the first open enrollment period that the Connector was in use, Enroll America assessed the tool’s successes and areas for improvement — identifying models for using the tool, and means to better support assisters and consumers. Additionally, Enroll America worked to expand to new areas in order to reach and connect more consumers to assistance, and determine ways to encourage more data reporting to enable more findings and quality improvement. This section reviews lessons learned in OE2 and the changes made in OE3.
Figure 1. Get Covered Connector During OE3
Maximizing In-Person Assistance With the Get Covered Connector
In-person assistance emerged as a key component to maximizing enrollment in OE1 and was the driving force behind Enroll America’s decision to create the Get Covered Connector for OE2.4 The tool was created to improve awareness of free in-person help and make it easier for consumers to find and schedule appointments with local assisters, and reduce administrative burden for assister groups.
The Connector builds upon Enroll America’s OE1 “Locator” tool that connected consumers looking for help getting health coverage to local listings of in-person assistance locations. In addition to these location listings, which remain in place across the country, the Connector allows consumers, in places with subscribing assisters, to view upcoming appointments and schedule an appointment directly through the tool. The backend portion of the Connector allows these subscribing assisters to upload their appointment schedules, send appointment reminders to consumers, track their work, and create program reports.
Even after three successful marketplace enrollment periods, in-person assistance remains an important resource, particularly for those who have little experience with health insurance. Going into OE3, the vast majority (71 percent) of uninsured Americans said it would be important to talk with someone one-on-one to help understand their options, and 64 percent of Spanish-speaking Latinos, a group disproportionately likely to be uninsured, indicated it would be “very important.”5
Many returning enrollees also want in-person assistance to help them understand and weigh their options in the still relatively new and dynamic marketplaces. From year to year, coverage can be affected by many factors, including an individual or family’s income and circumstances, as well as external factors such as the premiums of their plan and the second-lowest-cost silver plan in their area.6 Helping consumers return to the marketplace to update income and family information, get a new financial help estimate, and reassess the available options every year is critical to ensure satisfaction and retention. In a survey of consumers on Enroll America’s email list, almost half (49 percent) of marketplace enrollees renewing their coverage in OE3 reported getting in-person assistance from an assister or agent/broker.7 Furthermore, more of the renewal population who got in-person assistance reported being confident in their health plan selection, with 84 percent reporting being “somewhat” or “very” confident compared with 78 percent of renewing consumers who did not get in-person assistance.
Looking ahead, continuing to strengthen partnerships and bolster outreach to consumers about the availability of in-person assistance and the ability to easily find and schedule appointments in their communities will be paramount to ensuring new consumers have the support they need to successfully obtain to coverage, and that current enrollees maintain coverage that continues to meets their needs and budget each year.
The front-end appearance of the Connector remained constant for public users from OE2 to OE3. (See Figure 1.) However, behind the same interface, the Connector was upgraded to be faster, have additional consumer-friendly features, and support greater flexibility for assisters, including notable changes to the reporting features.
During OE3, Enroll America added several features to help consumers find appointments that best meet their needs. One of the additions was adding a banner for locations with appointments available in the near future (i.e. “Appointment available this week”). Another new feature, for partners who used their own consumer interface, was the option to add visual indicators for application-assistance sites trained to work with the LGBT community and have the default sort place those locations at the top of a consumer’s search.
Enroll America also worked to improve the Connector’s reporting features. Consistent and uniform reporting is critical for the outreach and enrollment community — it can help relevant stakeholders better understand what is working and what is not, and adapt programs to operate as efficiently and effectively as possible. Based on input from assisters, who found it redundant to enter data about appointments into the tool if it could not also be used for federal or state grant reporting, the Connector included additional fields that mirror the relevant reporting requirements either from the state-based marketplace (SBM) or the Centers on Medicare and Medicaid Services (CMS) for federal marketplaces, as well as some additional partner-specific customization.
These changes contributed to improved reporting rates compared to the previous year. In OE2, 58 percent of scheduled appointment results were recorded compared with 83 percent in OE3. While these changes represent a significant step at improving the reporting functionality, Enroll America will continue to improve the usability of the reporting features, including adding requirements from the Health Resources and Services Administration, and additional customization for groups with unique needs for OE4.
Figure 2. Get Covered Connector Use During OE3, Nov. 1, 2015 – Jan. 31, 2016
Taking cues from OE2 users, Enroll America identified a model to maximize the benefits of the Connector for assister organizations. In particular, successful partners often used the Connector to improve existing outreach and assistance efforts and systems, implemented universal usage across the organization, and devoted staff time to training and ongoing management. They also recognized the Connector’s limitations, namely that the tool by itself would not attract consumers without sufficient outreach work and appointment offerings.
Many of these lessons and insights were published in the April 2015 issue brief “Making the Connection: How Centralized Scheduling Benefits Consumers, Assisters, and the Enrollment Community.”8 The issue brief covered the value of centralized scheduling tools to assisters, insights into consumer behavior and needs, and recommendations for the enrollment and assistance community based on lessons learned from the inaugural year of the Connector.
For OE3, Enroll America took this model of success and incorporated it into the training offered through the Get Covered Academy, a new Enroll America initiative created in 2015 that formalized outreach training into an ongoing, structured program for participating partners. Academy trainings related to the Connector included methods of conducting outreach using the tool, using data to expanding assisters’ reach, and ongoing coaching during open enrollment to better leverage the tool to create a more effective and efficient outreach and assistance program. The OE3 Academy class included 118 organizations, more than half of which used the Connector, including 30 that also used the Connector in OE2.
Growth in OE3
The Connector was more widely used in more communities and by more partners in OE3 compared to OE2. The Connector grew across all metrics — expanding to offer appointments in 14 new states, bringing the total to 38 states and growing in the raw numbers of organizations, assisters, and appointments, as well as looking at the appointments offered, scheduled, and enrolled per organization and per assister. Additionally, for appointments with outcome information, enrollment rates increased by 30 percent (from 49 percent in OE2 to 64 percent in OE3). (See Figures 2 and 3.)
Figure 3. Get Covered Connector Growth from OE2 to OE3
In addition to the increase in raw numbers, there were some notable changes in the makeup of the organizations and individual assisters using the tool. Many of the organizations (43 percent) that used the Connector in OE3 also used the tool in OE2. And while in both years, just over half of the individual assisters using the tool were Navigators (or the Navigator equivalent in SBMs), the proportion of Certified Application Counselors (CACs) using the tool increased modestly from OE2 to OE3 (from 35 to 40 percent). Conversely, the share of agents and brokers declined (from 9 to 6 percent).
Partner Spotlight: Connect for Health Colorado
Colorado’s SBM, Connect for Health Colorado, adopted the Connector just before OE3 and quickly incorporated the tool into their outreach and application assistance work, leveraging its value through smart implementation and a robust outreach effort.
Overall in Colorado in OE3, consumers scheduled 12,000 Connector appointments and attended 9,500 appointments, which led to 6,000 enrollments. More impressive than these raw numbers are the process statistics that show the high level of reporting in the state — 95 percent of appointments had recorded outcomes, far outpacing the overall reporting rate of 83 percent — and the high level of appointment attendance — 81 percent attendance rate for appointments, which also outpaces the average of 71 percent.
Part of Colorado’s success can be attributed to the widespread, systematic adoption of the tool among its Health Coverage Guides (Colorado’s term for Navigators) and CACs across the state. In Connect for Health Colorado’s role as the state’s marketplace, they required the assisters they fund to use to the tool — which created a consistent Connector program across the state for consumers — and had central administration and uniform protocols — which created efficiencies with initial training and troubleshooting, and led to high rates of data reporting.
Another key to their success was their strong outreach effort, including partnering with a variety of community organizations and employers who referred their clients and employees to appointments, and a paid digital media effort run by Enroll America that drove Coloradoans seeking health coverage information and in-person assistance to the Connector to schedule appointments. Overall, 7 percent of consumers who saw ads for in-person assistance clicked to find out more (versus a 2 percent industry standard), and certain versions and keywords had 18-23 percent click-through rates from the ads to the Connector. This means that costs were kept low while still reaching thousands of Coloradans.
Insights in Meeting Consumers’ Needs: Appointment Purpose and Popular Times
On an average day in OE3, assisters using the Connector offered more than 10,000 appointments, and consumers scheduled over 1,000 appointments. This section looks at appointment trends and consumers’ behavior to help assistance organizations tailor their outreach and program plans to better match realities and meet consumers’ needs — from the basic understanding of likely outcomes and issues to when the most popular times are to schedule appointments.
In OE3, assisters reported the appointment attendance and enrollment outcomes of about 57,000 Connector appointments. More than 60 percent of these appointments led to enrollment for 2016 coverage (either new coverage or renewal), up from 49 percent in OE2.
- Seven in 10 of all appointments attended (enrollments or renewals) were for marketplace coverage. Three in 10 were for Medicaid.9
- The reasons consumers did not enroll were not specified in the vast majority of cases, but “Medicaid coverage gap” was the most common reason given when it was, encompassing 17 percent of all consumers who did not enroll.10
- Renewals represented 14 percent of all enrollments on the Connector, including 16 percent of marketplace enrollments and 10 percent of Medicaid enrollments.
Nationwide, about half of all consumers who seek assistance are getting marketplace coverage, with many others getting Medicaid coverage and another segment not enrolling. Understanding the likely outcomes and issues, through consistent reporting, can help assisters with expectations and preparation to better serve the range of consumers they will likely see.
Popular Times and Days of the Week for Appointments
When appointments were offered and scheduled varied considerably across times of the day, days of the week, and weeks during open enrollment. The appointment times assisters offered more often and the appointment times consumers selected more frequently did not always match, however.
- Assisters offered most appointments on weekday afternoons (noon-4:00 p.m.), and then weekday mornings (8:00 a.m.-noon); weekday evenings (4:00-8:00 p.m.) and weekends (any time) appointments were offered far less often. In fact, weekdays had over three times as many appointments offered per day compared with a weekend day (with 172,000 on an average weekday versus 53,000 on an average weekend day throughout OE3).
- The timeslots with the highest numbers of scheduled appointments were on weekday mornings. More specifically, Tuesday at 10:00 a.m. was the most common appointment time in OE3.
- Mornings (except Sundays) and weekend afternoon appointments were the most popular times. In other words these appointments have the highest uptake (as in, the proportion of appointments that consumers schedule compared with all that assisters offer) among appointments offered. Saturday mornings outpaced all other times — 19.4 percent of Saturday morning appointments were scheduled compared with the average uptake of 10.1 percent. (See Figure 4.)
Figure 4. Appointment Uptake by Day of Week and Time of Day, Nov. 1, 2015 – Jan. 31, 2016
The data shows that assisters are offering consumers the most appointments on weekday afternoons, but consumers are more interested in morning and weekend appointments. To better meet consumers’ schedules, assisters should consider offering more appointments in the mornings, particularly Saturday mornings, as well as weekend afternoon appointments.
Appointment Demand Over the Open Enrollment Period
All weeks during open enrollment are not created equal — indeed consumers often seem to be driven to engage by deadlines and to disengage around holidays. The appointment patterns across OE3 diverged somewhat from OE2, and may suggest a maturing of the marketplaces as enrollees return to renew coverage and avoid coverage gaps. (See Figure 5.) There was still a distinct increase in the number of appointments attended and those ending with enrollment during the last week of open enrollment — but there was also a longer, sustained interest leading up to the December 15 deadline (for coverage starting January 1, 2016). This pattern tracks with the overall enrollment trends nationally.11
Figure 5. Consumer Appointments and Enrollment Rates During OE3, Nov. 1, 2015 – Jan. 31, 2016
- About half of all attended appointments during OE3 happened before the December 15 deadline for January 1 coverage.
- Deadlines still matter to consumers: December 15 was the single day with the most attended appointments, followed by January 30.
- Similar to OE2, there is distinct drop in attended appointments near Thanksgiving, Christmas, and New Years.
Understanding these key times of demand (before deadlines as well as generally between Thanksgiving and the December 15), allows assisters to plan ahead to ensure consumers’ needs are met. Specifically, weekends right before or on deadlines tended to still offer fewer appointments than adjacent weekdays. Strategically increasing appointments on these weekends could be used to help additional consumers who may not be available during the week.
Reduce Flake Rates: Send Reminder Text Messages
A certain proportion of appointments scheduled ultimately fail to materialize, which is common across many industries — hospitality, health care, etc. And, while the Connector has comparatively high attendance rates, minimizing the number of no-shows and cancellations (the “flake rate”) helps ensure assisters are helping as many consumers as possible.12 The Connector was set up to provide automated email or text message reminders; some assisters additionally make reminder calls to further encourage consumers to attend appointments. During OE3, text messages appeared to be an effective way to increase appointment attendance rates: Consumers who received text message reminders were about 20 percent more likely to attend their appointment than those who did not receive a text reminder.13
Enroll America also sought to determine if certain messages were more motivating than others by conducting an experiment varying language in reminder texts sent to individuals who scheduled appointments between January 13 and February 1, 2016.
The language variations were based on ideas from electoral campaigns and social science research, with each appealing to a different element of consumers’ motivation to act. The reminders included two separate text messages sent in immediate succession on the day prior to the consumer’s appointment. The first message included one of four language variations:
- Control (no additional motivational language): “Remember! Your appointment for free, expert help with your health insurance application is coming up.”
- Bandwagon: “Hey there, lots of people in your neighborhood are getting free, expert help with their health insurance applications. Don’t miss out!”
- Empathy: “Hey there, we know money’s tight. Get all your questions answered at your appointment for free, expert help with your health insurance application.”
- Positive identity: “Thanks for taking charge of your health and your wallet! Remember your appointment for free, expert help with your health insurance application.”
All recipients then received a second text message comprising their appointment details: “Your appt is at [time] on [date]. [Location name], [street address], [city], [state]. To change, call [location phone number].”
In this test, the bandwagon and empathy messages had noticeably higher appointment attendance rates than either the control or the positive identity. (See Figure 6.) Due to the lack of true randomization in this experiment, however, more rigorous testing is needed to make definitive conclusions about the cause-and-effect relationship between text message language and consumer behavior; these results do show a potentially promising trend.14
Figure 6. Attendance Rates of Appointments With Different Reminder Texts Compared with the Control Message, Jan. 13 – Feb. 1, 2016
Reaching More Consumers: Considering Assistance Appointments in Non-English Languages and Non-Metro Areas
Racial and ethnic minority, immigrant, and rural populations continue to have among the highest uninsured rates in the United States.15 Figuring out effective ways to reach these groups and connect them to coverage was a priority of Enroll America and many others in the enrollment community in OE3. To further this objective, Enroll America sought to expand Connector appointment offerings that reach some of these constituencies — specifically introducing the Spanish-language Connector, and growing the number appointments offered in non-English languages and in rural areas.16 This section looks at these appointments to better understand the efficacy of this effort and help inform future programming.
Appointments in Non-English Languages
In OE3, assisters using the Connector held appointments in 18 languages besides English. The vast majority of available and scheduled appointments were still in English, with Spanish-language appointments being the second most popular. Spanish-language appointments were offered in 21 states, although the vast majority of appointments were in just a handful of states (Arizona, Colorado, Illinois, Florida, North Carolina, and Texas); appointments in other languages was scattered across 16 states.17 These appointments had higher attendance rates, and match or exceed enrollment outcomes compared with English-language appointments.
- Spanish-language appointments made up 7 percent of all scheduled appointments and 9 percent of recorded enrollments. The starkest contrast between English-language and Spanish-language appointments was the attendance rates: 84 percent of scheduled Spanish-language appointments led to an attended appointment versus 71 percent of English-language appointments. Spanish-language appointments also had slightly higher enrollment rates than appointments conducted in English (64 percent of attended appointments led enrollment versus 61 percent). (See Figure 7.)
- Appointments in the 17 languages besides English and Spanish made up just under 1 percent of scheduled appointments and just over 1 percent of attended appointments and enrollments. The enrollment rates for these appointments was equal to English-language appointments but attendance for these appointments was extremely high — with 91 percent of these appointments attended (which is 20 percentage points higher than English-language appointments).
- After English and Spanish-language appointments, Arabic, Vietnamese, and Creole had the most appointments (with 100-200 scheduled appointments each across the country).
Figure 7. Attendance Rates of Scheduled Appointments, Nov. 1, 2015 – Jan. 31, 2016
While the data do not speak to the reasons why non-English-language appointments have higher attendance rates, it seems most likely that there is pent up demand in these communities for access to in-person assistance, that these consumers value in-person assistance more than other groups, or that the relative scarcity of these appointments means the value of each appointment is greater. The underlying reason why Spanish language appointments would have a higher enrollment rate is even less clear since this group as a whole are likely to have more barriers during the enrollment process, like difficulty with identity and data-matching. However, it seems plausible that these consumers with a more difficult enrollment process are more dependent on in-person assistance than the rest of the population, or are more likely to finish the process with an assister than decide to make a final decision at another point.
While the underlying reasons for these trends is uncertain and needs to be monitored in future years, it does strongly suggests that in-person assistance is highly valued in the these communities, that the appointments are successful, and increasing the availability of appointments in non-English languages is an essential step towards serving these populations.
Appointments in Non-Metro Areas
In OE3, the number of locations offered on the Connector increased, both covering additional states and areas within states, including more rural areas. Across the country, the average distance for a Connector appointment remained approximately the same, from 4.5 miles in OE2 to 4.6 miles in OE3.
In analyzing the available locations using standard metro/non-metro measures, the vast majority of appointments were in metro areas, however from OE2 to OE3 the number of non-metro appointments scheduled more than doubled and the proportion of non-metro appointments increased by one-third.18 Like non-English language appointments, non-metro appointments matched or exceeded other appointments’ completion and enrollment outcomes. The scheduling uptake rates of non-metro appointments overall was slightly lower but there were areas of extremely high uptake as well.
- In OE3, 10 percent of scheduled appointments, 12 percent of attended appointments, and 13 percent of recorded enrollments were in non-metro areas.
- The scheduling uptake of offered appointments in non-metro areas was slightly lower (8 percent) compared to metro appointments (10 percent). The appointment uptake rate within non-metro areas varied, however, with the most rural locations within the non-metro definition having a slightly higher uptake rate (12 percent) and the most metro of these areas having the highest uptake of all areas (with 23 percent uptake) compared with the other non-metro categories.
- The attendance rate of non-metro appointments was higher than metro areas: There was an 82 percent attendance rate for appointments in non-metro areas compared with 70 percent in metro areas. (See Figure 7.)
- The enrollment rates of non-metro appointments was slightly higher than metro areas — 65 percent of attended appointments led to enrollment in non-metro appointments versus 60 percent in metro areas.
Like non-English appointments, non-metro appointments appear to have higher attendance rates and equivalent enrollment outcomes to other appointments, suggesting that in-person assistance appointments through the Connector are a successful way to engage this population in applying for health coverage. Based on the available data, an area that needs further exploration is uptake of appointments in non-metro areas, which was overall a bit lower than metro areas. However, the uptake in non-metro areas was not uniform, which suggests that additional outreach may be needed in some areas as well as an assessment of assistance locations to ensure that the locations are situated in places that are accessible for consumers. Tracking the uptake of different locations can help determine where to focus efforts while ensuring assisters’ times are being used best. While finding non-metro consumers may be harder, they are more likely to show up once they have an appointment.
Six Tips for OE4 Assister Programs
Insights from the Connector can help organizations design assistance programs that are more accessible for consumers and use assisters’ time more efficiently. Here are some findings to keep in mind while planning assisters programs in OE4:
1. Outreach is essential to engaging consumers about the availability of in-person assistance in their communities, and the steps they can take to access help. In addition to traditional outreach methods, one strategy to consider is a paid digital media campaign to drive consumers to make appointments.
2. Be prepared to help all kinds of consumers. While most consumers are marketplace eligible, there are many that are Medicaid eligible or fall into the Medicaid gap in non-expansion states.
3. Certain days and times are more popular among consumers. Consider offering more appointments in the mornings, particularly Saturday mornings, as well as weekend afternoon appointments.
4. Offer more appointments during key times of demand during open enrollment including before deadlines as well as generally between Thanksgiving and December 15 deadline. In particular, the number of appointments on weekends right before or on deadlines is an opportunity to reach consumers at their peak interest who may be unavailable during the week.
5. Remind consumers (via text or phone call) who make appointments for in-person help about the time and location of their appointment. Consumers who received text reminders were more likely to attend their appointments — and certain messages appear to be more motivating than others.
6. In-person assistance may be a valuable tactic in reaching certain constituencies. Appointments in non-English-languages and rural areas have good success with high attendance rates and equal or better enrollment outcomes. Offering appointments in non-English languages and more rural areas may help reach these populations who have disproportionally low rates of insurance.
In OE3, the Get Covered Connector grew in the number of assisters using the tool, appointments offered, the geographic reach of the tool, and, most important to maximizing enrollment, consumers assisted. By using the insights from Connector data outlined in this issue brief, assister programs can tailor their efforts to be more efficient and better meet consumers’ needs. However, the key lesson from the first year of the Connector remains indispensably important — the Connector, as well as the availability of in-person assistance, financial help, and coverage more generally, depends on smart partnerships and strong outreach efforts to educate and drive consumers to schedule appointments. There are still millions of uninsured Americans who need coverage and millions more marketplace enrollees who need to keep coverage, both of which depend on sustained outreach and assistance efforts in OE4 and beyond.
This piece was written by Molly Warren, Senior Policy Analyst, Best Practices Institute.
Assistance was provided by Jennifer Sullivan, Director, Best Practices Institute and Sophie Stern, Deputy Director, Best Practices Institute.
The author wishes to thank the following individuals from Enroll America:
- Andy Anderson, Ernie Anderson, Ed Coleman, André Crombie, Lauren Deitz, Jake Gysland, Kim Lehmkuhl, Lizzy Salinas, and Laura Timmerman for their input and guidance.
- William Tomasko and Cynthia Youngblood for their editorial and design support.
2 The third open enrollment period ran from November 1, 2015, through January 31, 2016.↩
3 States using the Connector during OE3: AK, AL, AR, AZ, CA, CO, DC, FL, GA, IL, IN, KS, KY, LA, MD, MI, MO, MS, MT, NC, ND, NE, NJ, NM, NV, NY, OH, OK, OR, PA, SC, SD, TN, TX, UT, VA, WA, WV, and WY.↩
4 Enroll America, In-Person Assistance Maximizes Enrollment Success, March 2014. Available online: https://s3.amazonaws.com/assets.enrollamerica.org/wp-content/uploads/2013/12/In-Person-Assistance-Success.pdf.↩
5 PerryUndem Research, Understanding the Uninsured Now, June 2015. Available online: http://www.rwjf.org/en/library/research/2015/06/understanding-the-uninsured-now.html.↩
6 In addition to any changes in plans’ cost from year to year, financial help on the marketplace also fluctuates. Financial help is determined on a sliding scale based on an individual or family’s income and family size as measured by the Federal Poverty Level (FPL), and the premium of the second-lowest cost silver plan in their rating area. Changes in family income and size, and the premium of second-lowest cost silver plan all affect financial help.↩
7 Enroll America surveyed the Get Covered America consumer email list in December 2015 and received 2,963 responses, of which 2,374 were sufficiently complete and included in the analysis. Questions asked included 2015 and 2016 coverage status, type of coverage, and experiences shopping for coverage in the marketplaces.↩
8 The issue brief is available online: https://www.enrollamerica.org/making-the-connection-get-covered-connector-in-person-assistance/.↩
9 During OE3, there was a mix of Connector states that had expanded Medicaid coverage to adults with incomes up to 138 percent FPL under the Affordable Care Act (ACA) and those who have not expanded Medicaid. Of states with Connector appointments in OE3, twenty-one (AK, AR, AZ, CA, CO, IL, IN, KY, MD, MI, ND, NJ, NM, NV, NY, OH, OR, PA, WA, WV, and WY) have expanded Medicaid, 16 have not (AL, FL, GA, KS, LA, MO, MS, NC, NE, OK, SC, SD, TN, TX, UT, and VA), and one, MT, expanded Medicaid during OE3 (January 1, 2016 start date).↩
10 The “Medicaid coverage gap” refers to individuals who fall above their state’s income threshold for Medicaid eligibility and below the income threshold for financial help on the marketplace (only individuals 100-400 percent FPL are eligible for financial help on the marketplaces). This gap is a result of states not expanding Medicaid under the ACA.↩
11 Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, Health Insurance Marketplaces 2016 Open Enrollment Period: Final Enrollment Report, March 2016. Available online: https://aspe.hhs.gov/pdf-report/health-insurance-marketplaces-2016-open-enrollment-period-final-enrollment-report.↩
12 Making the Connection: How Centralized Scheduling Benefits Consumers, Assisters, and the Enrollment Community, Enroll America, April 2015. Available online at: https://www.enrollamerica.org/making-the-connection-get-covered-connector-in-person-assistance/.↩
13 Out of nearly 47,000 scheduled appointments between November 1 and December 17, 2015, 19,000 appointments had a text message reminder. The overall flake rate for all the appointments was 27 percent while those with text reminder was 22. This is a 21 percent difference.↩
14 Since the Connector was not originally designed with this type of experiment in mind, we could not rigorously randomize text message sends. Instead, as an approximate randomization, we manually changed the copy of the outgoing reminder text message each day, chosen at random, through the Connector admin page.↩
15 Office of the Assistant Secretary for Planning and Evaluation, U.S. Department of Health and Human Services, Health Insurance Marketplace: Uninsured Populations Eligible to Enroll for 2016, October 2015. Available online: https://aspe.hhs.gov/pdf-report/health-insurance-marketplace-uninsured-populations-eligible-enroll-2016.↩
17 Appointments in Spanish were offered in 21 states: AR, AZ, CA, CO, FL, GA, IL, IN, KS, NC, NM, NV, OH, OK, OR, PA, SC, TN, TX, VA, and WA. Appointments in languages besides English or Spanish were offered in 16 states: AK, AZ, CO, FL, GA, IL, MI, NC, OH, OK, OR, PA, SC, TN, TX, and VA.↩
18 In defining metro/non-metro appointment locations, we determined the county the appointment was located and then used the U.S. Department of Agriculture’s (USDA) 2013 Rural-Urban Continuum Codes — which categorizes counties as metro or non-metro as defined by the Office of Management and Budget based on population and commuting criteria — to determine what category that appointment falls under. For more information on these measures, see: http://www.ers.usda.gov/data-products/rural-urban-continuum-codes/documentation.aspx.↩