2009 District of Columbia Health Insurance Survey

Methodology Report
February 2010
Barbara A. Ormond, Tim Triplett, and Sharon K. Long
The Urban Institute

David Dutwin and Robyn Rapoport

Adrian M. Fenty, Mayor

The District of Columbia Department of Health Care Finance contracted with the Urban Institute and its subcontractor, SSRS/Social Science Research Solutions, to conduct the 2009 District of Columbia Health Insurance Survey (DC-HIS). The goal of the DC-HIS is to document health insurance options and coverage and access to and use of health care for the noninstitutionalized population in the District. This report provides information about the methods used to collect and analyze the 2009 DC-HIS data. 1 The report is organized in seven sections. Section I describes the design of the sample used for the 2009 survey. Section II discusses the design of the survey instrument. Section III describes the data collection strategy. Data processing and preparation are described in Section IV. Section V addresses the response rate to the survey. Section VI describes the survey weights and variance estimation. The final section, Section VII, presents estimates of the uninsurance rate in the District of Columbia (DC) in 2009. I. Sample Design The 2009 DC-HIS has a dual sample frame design that combines a random-digit-dial (RDD) telephone sample and an address-based (AB) household sample. The decision to rely on the dual-frame sample for the DC-HIS reflects the changing telephone environment in the United States. Historically, RDD telephone interviewing has been the method of choice for many survey data collection efforts given the strength of its randomization method (random digit dialing), ease of administering complex questionnaires using computerized interviewing systems, excellent survey coverage of the overall population (given that less than 2% of Americans live in a household without telephone service), and relatively low cost. Survey coverage refers to the extent to which the sample frame for a survey includes all members of the target population. A survey design with a gap in coverage raises the possibility of bias if the individuals missing from the sample frame (e.g., households without landline telephones) differ from those in the sample frame. Unfortunately, the coverage of the overall population in RDD surveys is changing as more and more households are relying on cell phones and giving up their landline telephones. Cell phone numbers are typically not called in RDD surveys. Cell phone-only households are increasing rapidly in the United States, with 22.7 percent in the first half of 2009 (Blumberg and Luke 2009), as compared to 20.2% of households estimated to be cell phone-only in the second half of 2008, and 14.7% in 2007 (Blumberg et al. 2009). While there is limited data available on the share of cell phone-only households within each state, a recent model-based approach (combining survey data and synthetic estimates) was used to generate state-level estimates of cell phone-only households using the National Health Interview Survey. That work estimated that 20.0% of households in the District were cell phoneonly in 2007 (Blumberg et al. 2009). In order to capture cell phone-only households in the sample frame for the DC-HIS, the decision was made to include an address-based-sample (AB-sample) along with the RDD-sample for the survey. 2 The AB-sample captures households with landline phones, cell phone-only
The DC-HIS uses the same design as that developed by the Massachusetts Division of Health Care Finance (MADHCFP) and Policy for the Massachusetts Health Insurance Survey (MA-HIS). We are grateful to MA-DHCFP for their generosity in sharing the methods developed for the MA-HIS. 2 There are two alternative strategies to incorporate cell-phone-only households in the DC-HIS. The first is to conduct in-person surveys, which, unfortunately, are quite expensive. The second is to include cell phones as well as landline telephones as part of an RDD sample. While less expensive than in-person interviews, calling cell phones in
1

households and non-telephone households, supplementing the landline sample of the traditional RDD survey. One limitation of both the AB-sample and the RDD-sample is that they will miss homeless persons in the District, an estimated 6,228 persons. 3 Drawing the Samples. For the 2009 DC-HIS, we drew one random sample from the AB-sample frame and one from the RDD-sample frame. The AB-sample was developed in a two-step process. First, a file was generated of all DC residential addresses currently in use based on the United States Postal Service (USPS) Delivery Sequence File (DSF). The DSF is a computerized file that contains information on all delivery addresses serviced by the USPS, with the exception of general delivery. 4 The DSF is updated weekly and contains home and apartment addresses as well as Post Office boxes and other types of residential addresses for mail delivery. Then, that address file was compared to databases from InfoUSA and Experian that include all listed landline telephone numbers in the District to identify addresses with a listed telephone number. We began with a random sample of roughly 19,000 addresses, of which 12,819 had a listed phone number. In order to facilitate the fielding of the survey (discussed below), the final AB-sample was divided into two strata: addresses with a listed landline telephone number and addresses without a listed landline telephone number. Drawing the RDD-sample was also done in two steps. First, a file was generated of all residential telephone numbers currently in use in the District using Marketing Systems Groups GENESYS sampling system. The standard GENESYS RDD methodology produces a strict single stage, equal probability sample of residential telephone numbers to insure an equal and known probability of selection for every residential telephone number in the sample frame. 5 Then, the RDD file was run against reverse directory services (white pages) of all listed landline telephone numbers in the District from InfoUSA and Experian to identify telephone numbers in the RDD-sample that were associated with a known address. The random sample included about 12,540 telephone numbers, of which 5,464 were listed numbers with a known address. As with the AB-sample, the RDD-sample was split into two strata: those with a listed

an RDD survey is much more expensive than a landline-only RDD survey since federal regulations require that cell phone calling be done by hand. In addition, there is a much lower success rate for completing an interview with a person on a cell phone than a landline phone. Further, many cell phone users also have landlines, increasing the number of calls to cell-phones that are needed to identify cell-phone-only households, and many cell-phones are used by those under 18 years of age, who are not eligible for the survey. Finally, and most problematic, people who live in the District but have a cell phone that was purchased in another state will not be dialed in an RDD survey focused on District of Columbia residents since their cell phone number will not have a DC exchange and, conversely, not all cell phone numbers with a DC exchange are used by residents of the District. Concern about the potential size and composition of these two groups of cell phone users led to the decision to rely on the AB-sample to capture cell-phone-only households. 3 Homeless Services Planning and Coordinating Committee, Metropolitan Washington Council of Governments, The 2009 Count of Homeless Persons in Shelters and on the Streets in Metropolitan Washington, May 13, 2009. Available at http://www.mwcog.org/uploads/pub-documents/zVZeVw20090513103355.pdf. 4 See http://pe.usps.gov/text/dmm300/509.htm. 5 Consistent with industry standard procedure, we use list-assisted sampling to improve the efficiency of the RDDsample. In list-assisted sampling, the telephone numbers in the sample frame are divided into groups or banks of 1000 telephone numbers with the same first seven digits (e.g., (123-456-7XXX). These banks of telephone numbers are run through databases with listed household telephone numbers to identify telephone numbers that can be linked to a household with a listed telephone number. When none of the 1000 phone numbers in a bank are found to be listed, that 1000-bank is discarded from the survey sample. Any bank with at least 1 listed number is included in the sample frame.

landline telephone number (and, therefore, a known address) and those without a listed landline telephone number (and, therefore, with no known address). Since the survey relies on two sample frames, households that could be included in both the RDD- and AB-sample frames have a greater probability of being selected for the survey than households that could only be captured in one of the sample frames, all else equal. That is, households with a landline telephone, who are included in both the RDD- and AB-sample frames, have a higher probability of being selected for the survey than households without a landline telephone (including cell phone-only households and non-telephone households), since they are only included in the AB-sample frame. To address this issue, we oversampled households without a listed landline telephone number in the AB-sample frame (as it is this stratum that includes the cell phone-only households and non-telephone households). We also undersampled the households without a listed landline number in the RDD-sample to counterbalance their oversample in the AB-sample. (The goal was to keep the proportion of households without a listed landline telephone number in the combined sample roughly comparable to their share of the overall population.) The sample records in the four strataRDD-sample with a listed telephone number, RDD-sample without a listed telephone number, AB-sample with a listed telephone number, and AB-sample without a listed telephone numberwere randomized and put into small random subsamples or replicates to be released as needed for the study. The initial sample release was designed based on our best estimate of the distribution of District households by telephone ownership. We estimated that 80% of District households had a listed landline telephone, 12% had an unlisted landline telephone, 20% had a cell phone (or cell phones) and no landline telephone, and 2% had no telephone service at all. We anticipated a response rate of 40% for sample households with a known landline telephone and 25% response rate for those without a known landline telephone, for an overall projected response rate of 32.5%. 6 While expected to yield a lower response rate than a traditional RDD survey, the design used here captures information for residents of the District who will be missed in a traditional RDD survey (and, thus, effectively have a response rate of 0). In order to obtain a sample that was generally consistent with the distribution of telephone ownership, we allocated approximately 57% of the sample to the AB-sample and 43% to the RDD-sample. At about four weeks into the field, a supplemental sample release was ordered. The supplemental sample duplicated the proportions of ABS and RDD sample in the initial sample release. The goal of the survey was to obtain interviews with an overall sample of 4,000 households in the District. The final sample released for the survey was 20,445, with 8,697 households (43%) from the RDD-sample and 11,748 (57%) from the AB-sample. (Survey response rates are discussed below.) Table 1 shows the distribution of the final sample by strata. Based on the survey results, we estimate that 15.8% of DC households are cell phone-only households.

The projected response rate is calculated as (40% response rate * 80% of the sample who could be contacted by landline telephone) + (25% response rate * 20% of the sample that could not be contacted by landline telephone).

Table 1: Distribution of the Final Sample Released for the 2009 DC-HIS, by Sample Frame RDD-Sample Total Sample Sample Released Generated With listed landline phone/address With no listed landline phone/address Total Sample 5,464 7,322 12,786 5,218 3,479 8,697 AB-Sample Total Sample Sample Released Generated 12,819 6,548 19,367 5,221 6,527 11,748

II. Instrument Development The survey instrument used for the DC-HIS was an adaptation of the Massachusetts health interview survey, 7 which relies on the Coordinated State Coverage Survey (CSCS) developed by the State Health Access Data Assistance Center (SHADAC). 8 Modifications to that design were made to address issues specific to the District of Columbia. Pretest interviews were conducted as part of the survey design process. After establishing that the household was included in the survey sample frame (which included confirming that the home was a primary residence), the survey asked for a person aged 18 or older who could answer questions about the health insurance coverage of the members of the household. That respondent was asked questions that were used to create a roster of every individual in the household by age, gender, education, work status, and relationship to the respondent. Persons temporarily living away from home (including college students) were included in their usual household. Persons living in group quarters (e.g., dorms, nursing homes, and shelters) were excluded from the study, as were individuals sampled at their vacation homes or second homes. From this roster, the computer-assisted telephone interviewing (CATI) system randomly selected one household member to be the target person for the household. Basic information (including demographic and socioeconomic characteristics and insurance status) was collected on all of the members of each selected household, with more detailed information collected for the target person. Table 2 summarizes the topic areas covered in the DC-HIS for all members of the household and for the household member selected as the target individual for the survey. As noted above, the survey includes a complete roster of all members of the household. The survey instrument is included in Appendix A. Completion of the survey instrument took approximately 18 minutes on average.

For a description of the Massachusetts health insurance survey methodology, see http://www.urban.org/UploadedPDF/411814_Massachusetts_Methodology.pdf. 8 For a description of the CSCS, see http://www.shadac.umn.edu/shadac/survey/cscs.html. We thank Kathleen Call at SHADAC for sharing the CSCS and helping to modify it for the DC-HIS.

Table 2: Summary of Topic Areas Covered in the 2009 DC-HIS, by Household Members Targets Spouse (if present) and Parents (if present and Target age<26)

Topics Demographic characteristics (age, race/ethnicity, gender, marital status) Socioeconomic characteristics (education, employment status) Nativity and citizenship status Length of residency in the District Health insurance coverage Detailed employment questions Availability of employer sponsored insurance Awareness of public insurance options Health status Access to and use of health care Family income Homeownership Household telephone status

Survey Respondent X

All Household Members X

Target Household Member X

X X X X X X X X (Targets parents if Target age <18) X X X X X X X X

III. Data Collection Strategy The 2009 DC-HIS was fielded between August 24, 2009, and November 10, 2009. Data collection relied on two interview modes: telephone and web. Respondents were also given the option of requesting a paper copy of the survey, and SSRS mailed out 53 surveys to DC residents who requested them. 9 Information for the RDD-sample was obtained using traditional telephone interviewing methods, with a web-based option offered to the sample members for whom addresses could be obtained (i.e., the strata RDD-sample with listed telephone number). The survey options were explained to those sample members in advance letters and reminder letters. Advance letters and reminder letters were mailed to all members in the AB-sample offering the
9

Only one of these paper copies was completed; the respondent was in the AB-sample.

options of telephone and web survey models. In addition, for the AB-sample for which listed telephone numbers could be obtained (i.e., the strata AB-sample with listed telephone number), traditional telephone interviewing methods were used as well. The specific steps for the data collection process were as follows. 1. Advance letters were sent to all households for whom we had an address, which included all households in the AB-sample as well as telephone numbers in the RDDsample that had listed addresses. The advance letter invited the household to participate in the study and offered the option of calling in to the survey center using a 1-800 telephone number or completing a web-based survey. The letters to the RDDsample and the AB-sample with a listed telephone number also notified people they would be receiving a call in the next few weeks to complete the survey. These letters were mailed with a live stamp and sent in a 6x9 envelope with a clear address window. 2. Telephone interviews were attempted with all households for which we had a telephone number, including all of the RDD-sample and the households in the ABsample with a listed telephone number. Outgoing calls to the unlisted RDD sample were initiated a week before advance letters were mailed; outgoing calls to RDDsample with an address and AB-sample with a listed telephone number were initiated within a week of the mailing of the advance letters. 3. Post card reminders were sent to all households in the AB-sample and the members of the RDD-sample that had a listed address. These were sent out 12 days after the advance letter was mailed. 4. A final round of mailings was sent to all non-responding households in the ABsample and the RDD-sample that had a listed address. These were sent out two weeks after the post card reminders were mailed. The advance letters and reminder letters were printed on DC Department of Health Care Finance (DHCF) letterhead and signed by Julie Hudman, the DHCF director. All of the letters included a 1-800 toll-free number that the respondent could call for additional information on the survey or to complete the survey. Given the differences in our ability to contact the sample households for the survey, we expected the response rate for the survey to be higher for the RDD- and AB-samples with listed telephone numbers. We anticipated the lowest response rates for the AB-sample with no listed telephone number as we would only be able to contact those households via the mail. The response rates achieved for the survey are described in Section V. Survey Mode. As noted above, data were collected using two modestelephone and web. One individual completed the survey by mail, having requested a hard copy survey. Fiftyseven percent of the interviews were completed on the telephone with a CATI interviewer and 43% on-line using a web-based questionnaire. Of the telephone interviews, 9% were in-bound (toll-free) telephone calls from sample members requesting to complete the survey by telephone. Questions were identical in these two modes. The major distinction between these telephone and web modes is that, in the case of the CATI interviews, a trained interviewer guided the respondent through the process, whereas the web surveys were self-administered.

Although web respondents were completing the questionnaires without the direct assistance of an interviewer, all correspondence with respondents included contact information for project staff who were available to assist respondents with any problems they had completing the survey. Respondents completing the survey on-line were provided access to both staff telephone numbers and a link for emailing for technical support. Languages. The 2009 DC-HIS was administered in two languagesEnglish and Spanish. The CATI survey and web-based survey were available in both languages. All invitation letters and reminders to those in the AB-sample and the RDD-sample with a known address were delivered in English and Spanish. Potential respondents were invited in both languages to either call in to arrange a telephone interview or go to the website to complete the survey. Twenty-eight respondents completed the survey in Spanish. Training Materials and Interviewer Training. CATI interviewers received both written materials on the survey and formal training for conducting this survey. The written materials were provided prior to the beginning of the field period and included: 1. An annotated questionnaire that contained information about the goals of the study as well as detailed explanations of why questions were being asked, the meaning and pronunciation of key terms, potential obstacles to be overcome in getting good answers to questions, and respondent problems that could be anticipated ahead of time as well as strategies for addressing them 2. A list of frequently asked questions and the appropriate responses to those questions 3. A script to use when leaving messages on answering machines 4. Contact information for project personnel Interviewer training was conducted both prior to the study pretest (described below) and immediately before the survey was officially launched. Call center supervisors and interviewers were walked through each question in the questionnaire. Interviewers were given instructions to help them maximize response rates and ensure accurate data collection. They were instructed to encourage participation by emphasizing the social importance of the project and to reassure respondents that the information they provided was confidential. Interviewers were monitored during the first several nights of interviewing and provided feedback where appropriate to improve interviewer technique and clarify survey questions. The interviewer monitoring process was repeated periodically during the field period. Incentives. In order to encourage participation in the survey, all respondents were offered a twenty dollar gift card for either Giant or the Metro or given the option of donating $20 to one of four local charities: Capital Area Community Food Bank, DC Public Education Fund, DC Public Library Foundation, or DC Youth Orchestra. In addition, respondents were notified that through their participation in the study they would be entered into a drawing to win $100. Information on the incentives was provided in all advance letters and reminder letters and in the introduction to the survey. Call Rules for the CATI Interviews. For all RDD-sample members and AB-sample members with a listed telephone number, the initial telephone interviewing attempt included one initial call plus six callbacks. If an interview was not completed at that point, we set that telephone number aside for at least two weeks to rest. After that rest period, an additional six 7

callbacks were attempted. Overall, households received at least 12 call attempts. To increase the probability of completing an interview, we established a differential call rule that required that call attempts be initiated at different times of day and days of the week. Refusal Avoidance and Conversion Strategies. With the increased popularity of telemarketing and the use of telephone answering machines and calling number identification (i.e., caller-ID), the problem of non-response has become acute in household telephone surveys. Similarly, the increasing prevalence of unsolicited advertising in the mail (i.e., junk mail) makes it more difficult to conduct surveys using only invitation letters as we are doing here with the AB-sample without a listed telephone number. In addition to the incentives and call rules for the CATI interviews outlined above, we employed several other techniques to maximize the response rate for the survey. In the CATI interviewing, techniques included providing a clear and early statement that the call was not a sales call. In both versions of the survey (telephone and web), the introduction included an explanation of the purpose of the study, the expected amount of time needed to complete the survey, and a discussion of the incentives. A toll-free 1-800 number was provided to all respondents with a known address. In an effort to maximize the response rate in the interview phase, respondents were given every opportunity to complete the interview at their convenience. For instance, those refusing to continue at the initiation of or during the course of the telephone interview were offered the opportunity to be re-contacted at a more convenient time to complete the interview. They were also offered the opportunity to complete the survey on-line or to call into the 1-800 toll-free telephone number to complete the survey at their convenience. Those completing the interview on the web were able to complete the survey at their own speed and stop and re-start as needed. A progress meter was included in the web version of the survey in an effort to reduce the number of web suspends that remain unresolved at the conclusion of the field period. A key way to increase responses rates is through the use of refusal conversions. Though all of SSRSs interviewers regularly go through refusal aversion training, refusals are still a regular part of survey research. For all of the RDD-sample and the AB-sample with a listed telephone number, SSRS used a core group of specially-trained and highly-experienced refusal conversion interviewers to call all who initially refused the survey in an attempt to persuade respondents to complete the survey. Completed Interviews. Table 3 shows the number of completed interviews for the 2009 DC DC-HIS across the two sample frames by survey mode and by the telephone status of the household (cell phone only, landline phone and cell phone, and landline phone only). We completed surveys with 554 cell phone-only households, 3,215 landline and cell phone households, 789 landline-only households, 15 households without a phone, and 144 households with an undetermined telephone status. Survey response rates are discussed below.

Table 3: Total Number of Completed Interviews for the RDD- and AB-Samples for the 2009 DC DCHIS, by Survey Mode and Household Telephone Status RDD-Sample With Known Address 5,218 1874 With No Known Address 3,479 225 AB-Sample With With No Total Listed Listed Sample Landline Landline Telephone Telephone Number Number 5,221 1598 6,527 1020 20,445 4,717

Total

Total

Total Sample Released Total Completed Interviews Survey mode Phone-Outbound Phone-Inbound Internet Mail* Household Phone Status Cell phone only Cell phone and Landline phone Landline phone only No phone Phone status unknown

8,697 2,099

11,748 2,618

1,337 123 639 0

1,113 122 639 0

224 1 0 0

919 296 1402 1

917 129 552 0

2 167 850 1

2,256 419 2041 1

26 1,622 401 2 48

24 1,462 341 2 45

2 160 60 0 3

539 1,618 391 16 54

51 1,207 306 2 32

488 411 85 14 22

565 3,240 792 18 102

* Hard-copy survey requested by AB-sample respondent.

IV. Data Processing and Preparation Two analytical data files were created from the raw unedited survey data: (1) a personlevel file that includes all data elements collected for all persons in the household and data on the characteristics of the household and (2) a target-level file that includes all data elements collected for the target person in the household along with data on the characteristics of the targets family and household. Table 2, above, provides a summary of the variables included in each file. CATI range and logic checks were used to check the data during the data collection process. Additional data checks were implemented as part of the data file development work, checking for consistency across variables and family members, and developing composite measures of family and household characteristics.

Missing values for key demographic variablesage and race/ethnicity, health and disability status, and family income have been replaced through imputation in both files. For the variables for which imputed data were created, the data files include both the original variable (with missing values) and a new variable that includes the imputed values for cases that had missing values. In general, the item nonresponse was quite low (usually less than 2%); however, item nonresponse for family income was somewhat higher. Approximately 9.4% of the sample is missing all data on the income questions, while another 5.0% provided information on whether family income was above or below 200% of the federal poverty level but not any additional information. 10 In imputing values for the variables, we rely on one of two methods: a hot deck imputation method (Ford 1983) and a regression-model-based imputation method. In hot deck imputation, the missing value for an individual who failed to respond to a question is replaced by the value reported by a similar respondent for that question. A similar individual is identified based on observed characteristics that are strong predictors of the value of the variable with high levels of nonresponse. In regressionbased imputation, the missing value is replaced by the best linear unbiased predicted value from a regression model plus a random error term (to preserve the underlying distribution of the variable). We do not impute values for missing data for the remaining variables in the file. We have, however, incorporated edits across related variables and across family members to ensure consistent survey responses. While item non-response is relatively low in the survey, it is higher for questions that ask for more detailed information. For example, when asked about the number of employees in their firm, only 3.5% of non-elderly adults were unable to answer a question based on whether the number was less than 51 or 51 or more workers, while 6.01% were unable to answer when more detailed categories were asked about (less than 10, 11 to 50, 51 to 100, 101 to 500, 501 to 1000, or 1000 or more workers). In general, item non-response is higher for questions that seek more detailed information and measures that require combining responses across individual family members. V. Response Rates Response rates are one method used to assess the quality of a survey, as they provide a measure of how successfully the survey obtained responses from the sample. The American Association of Public Opinion Research (AAPOR) has established standardized methods for calculating response rates (AAPOR 2008). This survey uses AAPORs response rate definition RR3, with an AAPOR-approved alternative method of addressing ineligible households (described below). We report the overall response rates achieved for the RDD- and AB-samples, and for the combined RDD- and AB-sample (hereafter referred to as the combined sample). Before presenting those estimates, we describe our methods for calculating the response rates. Defining the Response Rate. The response rate is the number of people who completed the survey divided by the number of people in the sample who were eligible for the survey. The challenge in RDD- and AB-samples is in determining the number of people in the sample who were eligible for the survey. While eligibility for the survey can be determined for many sample members in both RDD- and AB-samples, there are parts of both samples for which eligibility cannot be determined directly. For example, in an RDD-sample, some telephone numbers are
10

These levels of item nonresponse appear to be much lower than those obtained in the MA-HIS.

10

never answered, while in an AB-sample with address-correction requested, some addresses that do not respond to the survey may be vacant or second homes. In estimating the response rate for the 2009 DC-HIS, we define four categories of sample records based on AAPOR response rate calculations. There are two categories of eligible sample records: those that resulted in a completed interview (AAPOR category one) and those that did not (AAPOR category two). The latter includes persons who refused to be interviewed and those who broke-off the interview part way through, as well as any sample record that was determined to be a household (e.g., an answering machine indicated that it was a household and not a business). Another category, AAPOR category four, includes sample records that are known to be ineligible, such as business numbers, fax machine numbers, non-working numbers, and vacant or second homes. Finally, AAPOR category three includes all sample records for which eligibility is unknown, such as RDD-sample records that result in a no answer (continuous ringing with no answering machine) or AB-sample records for which no response or address correction information is obtained. It is this latter category that is difficult to deal with in constructing the response rate since it is inevitable that the category includes both sample members who are eligible for the survey and those who are not. The size and composition of the category with unknown eligibility varies across the four survey strata. In general, we expect there to be relatively few ineligible addresses in the ABsamples since the USPS updates the DSF listing weekly to maintain its mail routes. Within the RDD-sample, we expect more ineligible telephone numbers within the unlisted numbers (e.g., businesses) than the listed numbers. Therefore, we calculate response rates separately for each of the survey strata. To generate response rates for the RDD- and AB- samples, we average the response rates from the separate strata within each sample after multiplying the response rate of each stratum by the inverse of their sampling rate. To generate an overall response rate for the full sample we average the overall RDD and AB-sample response rates in proportion to their overall released sample size. To allocate the unknown eligibility group for the RDD-samples, we use the survival method developed at Westat by Brick, Montaquila, and Scheuren (2002) to allocate sample members in AAPOR category three (eligibility is unknown) to eligible and ineligible status. This method has been used in a number of RDD surveys, including, for example, AHRQs Consumer Assessment of Healthcare Providers and Systems (CAHPS), the California Health Interview Survey, and the National Survey of Americas Families. To allocate the unknown eligibility group for the AB-samples, we rely on research conducted by Michael Brick and colleagues at Westat using an in-person follow-up to an ABsample in California. They found that 90% of AB-sample addresses that did not respond to the initial survey were residential addresses and, thus, eligible for the survey. 11 Compared with a Massachusetts survey based on a similar survey instrument and methodology, a higher proportion of the DC-HIS prenotification letters were returned as undeliverable; thus we assumed that 85% of those in AAPOR category three for the AB-samples are eligible for the survey in calculating the response rates for those strata.
Personal communication with Michael Brick, November 2008. Because the AB-sample design is used less often that the RDD-sample design, there is not yet an established standard on how to estimate a response rate for an AB sample. We anticipate that this study, along with other recent studies using a similar design, will generate improved procedures for estimating response rates for AB samples.
11

11

As expected, the response rate was higher for the RDD-sample than for the AB-sample. Not surprisingly, the response rate for the AB-sample without a listed landline telephone number, which is the strata that includes the cell phone-only households, was the lowest, at 19.7%. The response rate for the combined RDD- and AB-samples was 34.1%. Additional information on the sample disposition is provided in Table 5. Final response rates are summarized in Table 4. The overall response rate for the survey was 34.1%; for the RDD- and AB-samples, the overall rates were 43.7 % and 27.3%, respectively.
Table 4: 2009 DC-HIS Response Rates Sample Combined RDD- and AB-Sample RDD-Sample RDD-Sample with a listed landline telephone number RDD-Sample without a listed landline telephone number AB-Sample AB-Sample with a listed landline telephone number AB-Sample without a listed landline telephone number Response Rate 34.1% 43.7% 50.0% 29.1% 27.3% 36.1% 19.7%

The lack of standardization in reporting response rates makes it difficult to compare response rates across surveys (Brick, Ferraro, Strickler, and Rauch 2003). In calculating response rates here, we have made conservative assumptions about the share of ineligible households in the 2009 DC-HIS and have required that the respondent complete the bulk of the survey to be counted as a completed interview, which will tend to make the response rates we report look low relative to other surveys. As noted above, in calculating the response rates for the DC-HIS, we rely on a method developed by Westat to allocate sample members in AAPOR category three (eligibility unknown) to eligible and ineligible status. While response rates provide an indicator of potential bias in a survey (which can arise when survey nonrespondents are significantly different than respondents), lower response rates are not, in and of themselves, an indicator of survey quality since lower response rates do not necessarily increase nonresponse bias in surveys (Groves 2006, Groves and Peytcheva 2008). This issue has been addressed in a number of recent studies, including, for example, Keeter and colleagues (2000), who compared the results of a 5-day survey fielding period (response rate of 36%) to the results from fielding the same survey for 8 weeks (response rate 61%), and found no significant differences between the two surveys in the outcomes of interest.

12

Table 5: Disposition of the 2009 DC DC-HIS Sample RDDABABSample Sample Sample with with with No Listed Listed Listed Landline Landline Landline Telephone Telephone Telephone Disposition Number Number Number Eligible, completed survey (AAPOR category 1) Completed survey 1,874 1,598 1,020 Eligible, did not complete survey (AAPOR category 2) Refusal 462 404 21 Break off 38 26 2 Respondent never available 113 93 16 Web suspends/said will do on web, not complete 168 157 29 Physically or mentally unable/incompetent 36 31 0 Language problem 93 97 0 RDDSample with No Listed Landline Telephone Number
225 130 3 21

Total ABSample
2,618 425 28 109

Total RDDSample
2,099 592 41 134

Combined Sample
4,717 1,017 69 243

21

186

189

375

14 21

31 97

50 114

81 211

Unknown eligibility, did not complete survey (AAPOR category 3) Telephone always busy 5 1 0 1 No answer/no response 1,340 Not eligible (AAPOR category 4) Fax/data line 185 Non-working number 486 Undeliverable mail 186 Identified as nonresidence Business, government office, other organizations 111 No eligible respondent 121 Total Sample 5,218

2,369 0 0 220

4,796 0 0 605

1,521 412 708 0

7,165 0 0 825

2,861 597 1,194 186

10,026 597 1,194 1,011

115 110 5,221

0 38 6,527

378 24 3,479

115 148 11,748

489 145 8,697

604 293 20,445

13

VI. Survey Weights and Variance Estimation Survey Weights Survey data are weighted to adjust for differential sampling probabilities, to reduce any biases that may arise because of differences between respondents and non-respondents (i.e., nonresponse bias), and to address gaps in coverage in the survey frame (i.e., coverage bias). Survey weights, when properly applied in surveys can reduce the effect of nonresponse and coverage gaps on the reliability of the survey results (Keeter et al. 2000, Groves 2006). We constructed analytical survey weights for the DC-HIS using standard procedures. Separate weights are created for all persons and for the target-person in the household. The weights can be used to produce adult- and child-level population estimates as well as estimates of the total population in the District of Columbia. Since the survey used a dual-frame sample, weights were constructed separately for the RDD- and AB-samples and then combined to generate weights for the combined sample. Constructing the Household Base Weights. The first step in the weighting process for each sample is to create a household weight for each completed survey. That household weight is used to construct weights for each person in the household and for the target-person in each household. We begin with the households base weightthe inverse of the probability of selection of the sample telephone number for the RDD-sample and the inverse of the probability of selection of the address for the AB-sample. We first adjust the base weight so that all the households have the same probability of selection. In the RDD-sample, this adjustment corrects for the increased probability of selecting households that have more than one telephone number (based on information on the existence of multiple telephone numbers collected in the survey) and for the under-sampling of unlisted telephone numbers from the RDD-sample frame to compensate for their inclusion in the AB-sample. In the AB-sample, this adjustment corrects for the oversampling of addresses for which we were unable to obtain a listed telephone number. We then applied a non-response adjustment. In the RDD-sample, this adjustment corrects for the lower response rate from households that had unlisted telephone numbers, while for the AB-sample it corrects for the lower response rate from addresses for which we were unable to obtain a listed telephone number. Finally, we examined the distribution of the resulting weights and determined that there was no need to implement trimming 12 rules as there were few weights above 4 or below 0.2, and none well above 4 or well below 0.2. Constructing the Person Weights. To create a person weight for each person in an interviewed household we started with the household weight and then post-stratified so that our weighted sample population totals equaled population control totals based on data for the District of Columbia. Specifically, we aligned the sample to current Census population estimates for DC for the age, race/ethnicity and gender of the population, rates of home ownership, and ward. The demographic information and homeownership data came from the Current Population Survey, March Supplement, while the ward data came from Claritas. Respondent ward status was determined by Marketing Systems Group (M-S-G) using geographic information supplied by the respondents. We examined the distribution of the resulting person weights and since the largest person weight was 6.7, we again decided not to trim the final post stratified person weights.
In many household surveys, a final step is to trim very large weights, which are usually a byproduct of the post stratification adjustment. Inordinately large weights tend to substantially increase sampling variance.
12

14

Constructing the Target-Person Weights. The basic target-person weights are the product of the final household weight and the inverse of the probability that the person was selected from among all the persons in the household. As with the person weight, this initial target-person weight was post-stratified to population control totals based on the Census data outlined above. We examined the distribution of the resulting target-person weights and since the largest weight was 6.8 we again determined that there was no need to implement trimming rules. Creating Weights for the Combined Sample. The critical issue in constructing the weights for the combined sample is to adjust the weights so that all households have an equal probability of being included in the sample. Since households with landline telephones are included in both the RDD- and AB-samples, while non-landline-telephone households (i.e., cell phone-only households and non-telephone households) are only included in the AB-sample, this adjustment means reducing the weights for landline-telephone households and increasing the weights for non-landline-telephone households. To make this adjustment, we first estimate the percentage of District households and residents who live in non-landline-telephone households based on the information collected in the survey for the AB-sample. We then adjust the household weight, person weight and the target-person weight for the combined RDD- and AB-sample so that the percentage living in households without a landline telephone number in that sample equals the estimate from the ABsample. These new combined-sample weights are then post-stratified to the Census control totals described. Variance Estimation and the Average Design Effect Complex survey designs and post-data collection statistical adjustments affect variance estimates and, as a result, tests of significance and confidence intervals. Variance estimates derived from standard statistical software packages that assume simple random sampling are generally too low, which leads significance levels to be overstated and confidence intervals to be too narrow. The impact of the survey design on variance estimates is measured by the design effect. The design effect describes the variance of the sample estimate for the survey relative to the variance of an estimate based on a hypothetical random sample of the same size. In situations where statistical software packages assume a simple random sample, the adjusted standard error of a statistic should be calculated by multiplying by the design effect. Each variable will have its own design effect. The average design effect for estimates for the target person in the household is 1.85 (effective sample size = 2,549). As with all surveys, results based on this sample may differ from what would have been obtained if we had surveyed the entire DC population. Based on the effective sample size of 2,549, the 95 percentage confidence interval for estimates for outcomes that occur for about 50% of the sample will be +1.94 percentage points and for outcomes that occur for 90% or 10% of the sample, the 95 percentage confidence interval is +1.16 percentage points. Variance estimation procedures have been developed for most standard software packages to account for complex survey designs. We provide a replicate stratum (strata) and primary sampling unit (psu) variable on the survey data files that can be used with the appropriate weight variable to obtain corrected standard errors using a Taylor series approximation (or other related linearization method). Users interested in using a linearization method can choose to use SUDAAN, the SVY commands in Stata, the PROC 15

SURVEYMEANS and PROC SURVEYREG commands in SAS, or the CSELECT complex samples procedures in the SPSS complex samples module. VII. Estimates of the Uninsurance Rate for the District of Columbia Cell phone-only Households. As noted earlier, the goal of the AB-sample was to capture information on the cell phone-only households in the District, as they are not covered in an RDD-sample. We estimate that 15.8% of households in DC were cell phone-only households at the time of the survey. Consistent with the estimates for the nation as a whole by Blumberg and Luke (2008), we found that the uninsurance rate among cell phone-only households in the District was greater than that of households with landline telephones (8.4% versus 5.9%), as shown in Table 6, although this difference is not statistically significant.
Table 6: Estimates of the Uninsurance Rate in the District of Columbia, by Household Telephone Status (2009) Uninsurance Rate 95% Confidence Interval Cell phone-only households Landline-telephone households Total Population 8.4% 5.9% 6.2% 5.6% -11.2% 5.0% - 6.8% 5.3% - 7.1%

Uninsurance Estimates in the AB-Sample and RDD-Sample. In Table 7, we compare the uninsurance rate estimates generated from the RDD-sample, the AB-sample, and the combined sample. As shown, the estimates of the uninsurance rate were similar for the AB- and RDD-samples, ranging from 6.7% based on the AB-sample to 5.1% based on the RDD-sample. The uninsurance rate in the combined sample was 6.2%, which translates to 36,774 uninsured residents in the District at the time of the survey. This estimate of the number of uninsured residents is lower than estimate of 9.8% from the 2007-2008 Current Population Survey (CPS) (March Supplement) 13 and the estimate of 7.8% from the 2008 American Community Survey (ACS). 14 The differences are likely due in part to refinements in the DC-HIS including the inclusion of the Alliance as an insurance choice and the dual sampling frame. In addition, there are differences in survey years, survey methodology, and the survey instruments. The DC-HIS estimates of employer-sponsored insurance (55.0%) and public coverage (32.8%) for all residents compare to the CPS estimates of 53.5% and 30.7%, respectively, and to the ACS estimates of 61.0% and 31.9%, respectively.
Table 7: Estimates of the Uninsurance Rate in 2009 in the District of Columbia, by Sample Uninsurance Rate RDD-Sample AB-Sample Combined Sample 5.1% 6.7% 6.2% 95% Confidence Interval 3.7% - 6.4% 5.4% - 8.0% 5.3% -7.1%

13 14

Kaiser State Health Facts Online, http://www.statehealthfacts.org/profileind.jsp?ind=125&cat=3&rgn=10. http://sda.usa.ipums.org/cgi-bin/sdaweb/hsda?harcsda+2008.

16

References
American Association for Public Opinion Research. Standard Definitions: Final Dispositions of Case Codes and Outcome Rates for Surveys. 5th edition. (2008) Lenexa, Kansas: AAPOR. Behavioral Risk Factor Surveillance System. 2007 Summary Data Quality Report. Centers for Disease Control and Prevention, 2008. Available at: ftp.cdc.gov/pub/Data/Brfss/2007SummaryDataQualityReport.pdf. Blumberg, S.J., and J.V. Luke. Wireless Substitution: Early Release of Estimates Based on Data from the National Health Interview Survey, July-December 2006. National Center for Health Statistics. May 13, 2008. Blumberg, S.J., and J.V. Luke. Wireless Substitution: Early Release of Estimates Based on Data from the National Health Interview Survey, July-December 2007. National Center for Health Statistics. May 14, 2007. Blumberg, S.J., et al. 2009. Wireless Substitution: State-level Estimates from the National Health Interview Survey, January-December 2007. National Center for Health Statistics, March 11, 2009. Blumberg SJ, Luke JV. Wireless substitution: Early release of estimates from the National Health Interview Survey, January-June 2009. National Center for Health Statistics. December 2009. Available from: http://www.cdc.gov/nchs/nhis.htm. Brick, J.M., et al., "Estimating Residency Rates for Undetermined Telephone Numbers." Public Opinion Quarterly, 66 (2002), 18-39. Brick, J.M., D. Ferraro, T. Stickler and C. Rauch. 2002 NSAF Response Rates. Report No. 8, NSAF Methodology Reports, 200X. Available at www.urban.org/publications/900692.html. Cervantes, I.F., J.M. Brick and W.S. Edwards. Response Rates Report 4, California Health Interview Survey 2005 Methodology Series, 2007. Available at www.askchis.com/pdf/CDC-HIS2005_method4.pdf. Curtin, R., S. Presser and E. Singer, "Changes in Telephone Survey Nonresponse Over the Past Quarter Century," Public Opinion Quarterly 69, no. 1 (Spring 2005): 87-98. Curtin, R., et al., "The Effects of Response Rate Changes on the Index of Consumer Sentiment." Public Opinion Quarterly 64 no. 4 (2000): 413-428. Groves, R.M. "Nonresponse Rates and Nonresponse Bias in Household Surveys," Public Opinion Quarterly 70, no. 5 (2006): 646-75. Groves, R. M. and Emilia Peytcheva. 2008. The Impact of Nonresponse Rates on Nonresponse Bias. Public Opinion Quarterly 72: 167-189. Holbrook, A., et al., The Causes and Consequences of Response Rates in Surveys by the News Media and Government Contractor Survey Research Firms. In Advances in Telephone Survey Methodology, ed. James M. Lepkowski, et al. (New York: Wiley, 2007). Keeter, S., et al., "Consequences of Reducing Nonresponse in a Large National Telephone Survey," Public Opinion Quarterly 67 (Summer 2000): 125-48. Keeter, S., et al., Gauging the Impact of Growing Nonresponse on Estimates from a National RDD Telephone Survey. Public Opinion Quarterly 70, no. 5 (2006): 759-779. Singer, E. "The Use of Incentives to Reduce Nonresponse in Household Surveys," in Survey Nonresponse, ed. Robert M. Groves et al. (New York: Wiley, 2002), Chapter 11, 163-77. Smith, T. "A Review of Methods to Estimate the Status of Cases with Unknown Eligibility." Report prepared for the AAPOR Standard Definitions Committee, May (2004): version 1.1. Visser, P.S., et al., Mail Surveys for Election Forecasting? An Evaluation of the Colombia Dispatch Poll, Public Opinion Quarterly vol. 60 (1996): 181-227.

17

APPENDIX A 2009 DISTRICT OF COLUMBIA HEALTH INSURANCE SURVEY

18

Job #I1100 FINAL CLEAN SURVEY September 10, 2009

District of Columbia Health Insurance Survey- 2009 INTRO1. Hello. My name is __________ and I'm calling on behalf of the District of Columbia. Im with Social Science Research Solutions, or SSRS. (INTERVIEWER SHOULD CONFIRM THAT RESPONDENT IS 18 OR OLDER. OTHERWISE ASK TO SPEAK WITH SOMEONE IN THE HOUSEHOLD WHO IS 18 OR OLDER) We are working on a study about health care and health insurance in the District. As a thank you for your participation in this important research, we will send eligible households a $20 gift card and also enter you in a drawing for a $100 prize. As you may know, DC is a national leader in offering coverage to its residents. The Districts Department of Health Care Finance is sponsoring this survey to get the information it needs to improve its public coverage programs and to find ways to help make private coverage easier to obtain. INTERVIEWER: IF YOU REACH A CHILD, ASK TO SPEAK TO AN ADULT. NEW VERBIAGE INTRO2. FOR RDD sample: Your telephone number was randomly selected from phone numbers in DC. Your participation in this study is voluntary and will be a great help. This study takes only about 20 minutes. (IF NEEDED: It tends to be a bit shorter for smaller households and a bit longer for larger households.) (IF NEEDED, READ: As I mentioned, we will send each eligible household that completes the survey a $20 gift card and we will enter the household into a drawing to win $100 as an additional thank you.) New verbiage INTRO2. FOR USPS sample: Your address was randomly selected from residential addresses in DC. Your participation in this study is voluntary and will be a great help. This study takes only about 20 minutes. (IF NEEDED: It tends to be a bit shorter for smaller households and a bit longer for larger households.) (IF NEEDED, READ: As I mentioned, we will send each eligible household that completes the survey a $20 gift card and we will enter the household into a drawing to win $100 as an additional thank you )

NEW INTRO 3 VERBIAGE INTRO3. Before we start, let me tell you that everything you say will be kept private. Your answers will be combined with those of other people in DC. The study will not be used for marketing purposes and your decision whether or not to participate will not have any effect on anything to do with your insurance coverage, health care, eligibility for health care services, or your relationship with any agencies in DC. No one in the District government or anyone else will be able to link the information that you provide to you. You may skip over questions or stop the interview at any time you wish.

INTRO4. If you have questions about the study, I can give you a phone number now or at the end of the survey that you can call to find out more about the study. (IF NEEDED: For questions about the survey, please call Kathy Langdale 1-800-633-1986. To speak with someone from the DC Department of Health Care Finance , please call Jessica Schubel at 202-442-5889.)

S1.

First of all, is this a second home or vacation home? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK S1B IF S1=1 AND (RDD SAMPLE or USPS SAMPLE WITH LISTED LANDLINE PHONE)) IF REMAINING USPS SAMPLE, READ: Thank you. We are only interviewing people at their main residence. TERMINATE

S1b.Is the telephone number I dialed, xxx-xxx-xxxx, the number I would use to reach you at your main residence? 1 2 Yes No CONTINUE (READ: Thank you. We are only interviewing people at their main residence. TERMINATE) (READ: Thank you. We are only interviewing people at their main residence. TERMINATE) (READ: Thank you. We are only interviewing people at their main residence. TERMINATE)

(DO NOT READ) Dont know

(DO NOT READ) Refused

(ASK EVERYONE: (IF S1b = YES OR S1 = 2, D OR R) S1A. Is your primary residence located in the District of Columbia? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

IF S1A = 1 CONTINUE TO S5 IF S1A = 2, 7 or 9 READ: Thank you. We are only interviewing people whose main residence is in the District of Columbia. THANK & TERM. RECORD AS TQS1A (MOVED PLACEMENT Q.S5 ON QRE, to match program) S5. Is this a house, condominium, apartment, mobile home, some other kind of private residence, or something else? (DO NOT READ) 1 2 House or row house Condominium or apartment 2

3 4 7 D R

Mobile home or trailer Other private residence Other Non-private residence (such as a group home such as an assisted living or continuing care home) (DO NOT READ) Dont know (DO NOT READ) Refused

(IF S5 = 7, D, R READ: THANK YOU. WE ARE ONLY INTERVIEWING PEOPLE IN THEIR PRIVATE RESIDENCE.)

(IF Q.S1 = D OR R OR Q.S1b=2, D OR R TERMINATE AND RECORD AS TQS1. IF Q.S1 = 2 OR Q.S1b=1, CONTINUE) (INTERVIEWER READ) Id like to begin by asking some questions about health insurance coverage for people in your household. S2. Can you answer questions about health insurance for people in this household? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(IF S2=2, D, R; ASK S3) S3. Is another adult available who could answer questions about health insurance? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused GET PERSON ON PHONE AND CONTINUE [SKIP TO INTRO1] SET UP CALL BACK. (THANK AND TERMINATE) (THANK AND TERMINATE)

S4. How many people currently live or stay here? Please include anyone temporarily away for school or the armed services. (PROBE: Include in this number, children, foster children, roomers, or housemates not related to you, college students living away while attending college and National Guard members who are deployed.) (Do not include people who live or stay at another place most of the time, people in a correctional facility, nursing home, or residential facility, or people in the regular Armed Forces living somewhere else.) ________ people (RANGE 1-10) 11 11 or more D (DO NOT READ) Dont know R (DO NOT READ) Refused 3

(THANK AND TERMINATE) (THANK AND TERMINATE)

(INTERVIEWER READ IF Q.S4 = 2+) I need some general information about the people in this house so that one person can be picked at random to talk about their access to health insurance. (INTERVIEWER READ IF Q.S4 = 1) Can you please tell me your age? ASK S6 TO S9 IN SUCCESSION FOR EACH MEMBER OF THE HOUSEHOLD (PN: Questions S6 S9 can be used to create a HH Roster listing each person in HH) S6a. Starting with yourself, what is your age? (INTERVIEWER IF RESPONDENT DK/REFUSES AGE: I understand your reluctance to give your age, but this information is totally confidential. It is very important that we gather this information accurately to help improve health insurance coverage for District families. IF RESPONDENT STILL DK/REFUSES AGE, ASK Q.S6a1) ___________ (AGE) RR (DO NOT READ) Refused (ASK Q.S6a1 IF Q.S6 = DD OR RR) S6a1. Could you please tell me if you are? (READ LIST. ENTER ONE ONLY) 1 2 D R Less than 18 years of age 18 years of age or older (DO NOT READ) Dont know (DO NOT READ) Refused

(IF S6<18 OR Q.S6a1 = 1, D, OR R; ASK S6A) S6A. Is there someone available who is 18 or older? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused GET PERSON ON PHONE AND CONTINUE [SKIP TO INTRO1] SET UP CALL BACK (THANK AND TERMINATE) (THANK AND TERMINATE)

S7a. RECORD RESPONDENT GENDER 1 2 Male Female

INTERVIEWER NOTE: If unsure of respondents gender please confirm. S6(b-j). And the next persons age? (INTERVIEWER IF RESPONDENT DK/REFUSES AGE: I understand your 4

reluctance to give other household members ages, but this information is totally confidential. It is very important that we gather this information accurately to help improve health insurance coverage for District families. IF RESPONDENT STILL DK/REFUSES AGE, ASK Q.S6b1) __________ years (ENTER AGE 1-110) 00 Less than 1 year old DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK Q.S6b1 FOR EACH Q.S6b-j = DD OR RR) S6b1(b-j). Could you please tell me if this person is? (READ LIST. ENTER ONE ONLY) 1 2 D R Less than 18 years of age 18 years of age or older (DO NOT READ) Dont know (DO NOT READ) Refused

IF Q.S6(b-j) = 18+ OR Q.S6b1 = 2, INSERT person and male or female IF Q.S6(b-j) IS <18 OR Q.S6b1 = 1, INSERT child and boy or girl S7(b-j). Is this (child/person) (a boy or a girl/male or female)? 1 2 R Male/Boy Female/Girl (DO NOT READ) Refused

S8(b-j). What is this persons relationship to you? (DO NOT READ. ENTER ONE ONLY. R CAN PROVIDE UP TO ONE PARTNER AND FOUR PARENTS, GUARDIAN AND WARD SHOULD BE CODED BEFORE ANY OTHER RELATIONSHIP EXCEPT PARENT OR STEPPARENT OR CHILD/STEPCHILD/FOSTER CHILD, SO A GRANDPARENT AND GUARDIAN SHOULD BE CODED AS GUARDIAN) (INTERVIEWER IF RESPONDENT REFUSES RELATIONSHIP: I understand your reluctance to give your relationship to other members of your household, but this information is totally confidential. It is very important that we gather this information accurately to help improve health insurance coverage for District families. IF RESPONDENT STILL REFUSES RELATIONSHIP, THANK AND TERMINATE.) 01 02 03 04 05 06 Spouse (wife/husband) Unmarried partner / significant other Child / stepchild / foster child/ward Parent / Stepparent / foster parent/guardian Sibling / Stepsister / Stepbrother Grandparent / Step-grandparent 5

07 08 09 10 11 12 13 14 DD RR

Grandchild / Step-grandchild Son-in-law / Daughter-in-law Father-in-law / Mother-in-law Other relative Employer Employee (maid, nanny, au pair, housekeeper, etc.) Professional caregiver (nurse, aide, etc.) Other non-relative (DO NOT READ) Dont know (DO NOT READ) Refused

ASK IF S6 >/= 16 S9(a-j). What is the highest level of school [you have/she has/he has] completed or the highest degree [you have/she has/he has] received? (DO NOT READ. ENTER ONE ONLY) 1 2 3 4 5 6 D R Less than high school (grades 1-11, grade 12 but no diploma) High school graduate or equivalent (e.g. GED) Some college but no degree (incl. 2 year occupational or vocational programs) Associates Degree (not occupational or vocational programs) College graduate (e.g. BA, AB, BS) Postgraduate (e.g. MA, MS, MEng, Med, MSW, MBA, MD, DDs, PhD, JD, LLB, DVM) (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.S7=1 INSERT he; IF Q.S7=2, INSERT she, ELSE INSERT they) (IF S6(a-j) = >16 ASK) S9a(a-j). [Are you /Is she /Is he] currently working for pay? 1 2 D R Yes, working No, not working (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.S4=1, SKIP TO Q.S11) FOR RESPONDENT, INSERT you S10. I will be asking some specific insurance coverage questions about one randomly chosen person from your household. For those questions my computer has selected [you/TARGET]. INTERVIEWER RECORD 1 2 Respondent Target 6

(P.N. IF RANDOM PERSON CHOSEN IS TARGET AND Q.S6b1 = D OR R FOR THAT PERSON OR Q.S8b-j = RR FOR THAT PERSON, THANK AND TERMINATE) S10a. What is the first name or initials of the person I selected? 1 R Answer given (SPECIFY) ______________ Refused

IF AGE >=17, ASK Q.S11 AND IDENTIFY SPOUSE/PARTNER (S11=1 OR 2) IN THE ROSTER (TSPOUSE) INSERT is this person IF Q.S10 = 2 IF TARGET IS RESPONDENTS SPOUSE, GEN IN CODE 1 (RESPONDENTS WHO ARE ALSO TARGETS SHOULD BE ASKED THIS QUESTION IF S8a-j NE 01) IF ONE PERSON HOUSEHOLD, (S4 =1) DO NOT SHOW CODE 2, LIVING WITH PARTNER) S11. Are you (is this person) currently: 1 2 3 4 5 6 D R Married Living with partner Divorced Separated Widowed Never Married (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.S4 = 1, READ VERBIAGE IN PARENS) (I will be asking some specific questions about your insurance coverage) IF TARGET IS THE RESPONDENT, SKIP TO SETUP 1 (ASK Q.S12 OF EVERYONE EXCEPT FOR RESPONDENT) S12(b-j). It would be helpful to know the relationship between the other members of your household and (INSERT NAME OR INITIALS FROM Q.S10a OR RELATIONSHIP FROM Q.S8[b-j])? What is the relationship of your (RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j] if multiple members with same relationship code] to the TARGET)? (DO NOT READ, ENTER ONE ONLY, GUARDIAN AND WARD SHOULD BE CODED BEFORE ANY OTHER RELATIONSHIP EXCEPT PARENT OR STEPPARENT OR CHILD/STEPCHILD/FOSTER CHILD, SO A GRANDPARENT AND GUARDIAN SHOULD BE CODED AS GUARDIAN)) 01 Spouse (wife/husband) 02 Unmarried partner / significant other 03 Child / stepchild / foster child/ward 7

04 05 06 07 08 09 10 11 12 13 14 DD RR

Parent / Stepparent / foster parent/guardian Sibling / Stepsister / Stepbrother Grandparent / Step-grandparent Grandchild / Step-grandchild Son-in-law / Daughter-in-law Father-in-law / Mother-in-law Other relative Employer Employee (maid, nanny, au pair, housekeeper, etc.) Professional caregiver (nurse, aide, etc) Other non-relative (DO NOT READ) Dont know (DO NOT READ) Refused

ASK IF S12(b-j) NE 04 AND TAGE<18 S12-1. Are any members of your household the legal guardian or caretaker of (TARGET)? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

ASK IF Q.S12-1=1 S12-2. Which household member (or members) is (TARGETs) legal guardian or caretaker? (DO NOT READ, ALLOW MULTIPLE) 01 02 03 04 05 06 07 08 09 10 11 12 13 14 DD RR Spouse (wife/husband) Unmarried partner / significant other Child / stepchild / foster child/ward Parent / Stepparent / foster parent/guardian Sibling / Stepsister / Stepbrother Grandparent / Step-grandparent Grandchild / Step-grandchild Son-in-law / Daughter-in-law Father-in-law / Mother-in-law Other relative Employer Employee (maid, nanny, au pair, housekeeper, etc.) Professional caregiver (nurse, aide, etc) Other non-relative (DO NOT READ) Dont know (DO NOT READ) Refused

[PN: ALL HH MEMBERS CODED AS GUARDIAN SHOULD = 04 IN S12]

SETUP1: P.N. - Create the following variables to be used in remainder of survey HH_COUNT Number of people in household (S4) TMARR 1 if TARGET is married/partner; 0 otherwise TPAR 1 if TARGET is parent; 0 otherwise TAGE TARGETs age TFEM 1 if TARGET is female; 0 otherwise TFAM_COUNT Number of people in TARGETs family. IF TAGE<19 & TMARR=0 & TPAR=0: TARGET+PARENTS/PARENTS UNMARRIED PARTNER+SIBLINGS<19 FROM ROSTER IF TAGE<19 & (TMARR=1 OR TPAR=1): TARGET+SPOUSE+CHILDREN<19 FROM ROSTER IF TAGE>18: TARGET+SPOUSE+CHILDREN<19 FROM ROSTER If unmarried partner (who can be the respondent or the unmarried partner of the respondent) is not parent of TARGET OR if TARGET has an unmarried partner, create variable: TFAML1 = TFAM_COUNT-1 TFAML1 = Number of people in TARGETs family, not including unmarried partner HEALTH INSURANCE INSERT you/have/do you IF Q.S10 = 1 INSERT NAME/INITIALS/RELATIONSHIP/has/does IF Q.S10 = 2 (ASK ITEMS b and l if TAGE>17) (ASK ITEM c IF H1b < 1 >) (ASK ITEM l IF H1b AND H1c < 1 >) (ASK ITEM M IF H1l AND H1b AND H1c < 1 >) (ASK ITEM k if TAGE>15 AND </=64 AND H1b AND H1c < 1 >) Changes to this list of insurance types H1. I am going to read you a list of different types of health insurance coverage. Please tell me if (you / TARGET) currently (have/has) any of the following types of insurance. Please exclude any health insurance plans that cover only ONE type of service, like plans for dental care or prescription drugs. (Do you/does TARGET) currently have (READ LIST)? IF RESPONDENT ASKS TO SKIP THROUGH INSURANCE QUESTIONS, SAY: Im sorry, but I have to read all of the insurance categories. 1 Yes 2 No D (DO NOT READ) Dont know R (DO NOT READ) Refused a. DELETED 9

b. c. d. f. g. h.

i.

j. k. l. m.

Health insurance through (your/TARGETs) work or union (Probe: This insurance could also be through COBRA, through a former employer or a retiree benefit.) Health insurance through someone elses work or union (Probe: This insurance could also be through COBRA, through a former employer or a retiree benefit.) Medicare (PROBE: Medicare is the health insurance for persons 65 years old and over or persons with disabilities. For many people, this will be a red, white and blue card.) Veteran's Affairs, Military Health, TRICARE or CHAMPUS DELETED Medicaid or DC Healthy Families (PROBE: This is a District program for low- and moderate-income families with children, seniors, and people with disabilities. (You/TARGET) may have coverage under Medicaid or DC Healthy Families through a health insurance plan such as Health Right, Chartered, or Unison.) The Alliance or DC Health Care Alliance? (PROBE: This is another DC program that provides insurance at no cost for low- and moderate-income residents. (You/TARGET) might have coverage under the Alliance through a health insurance plan such as Health Right, Chartered, or Unison.) DELETED A qualifying student health insurance plan? (PROBE: A QSHIP is a health insurance plan that is sponsored by a college or university.) Health insurance bought directly by (you / TARGET) (PROBE: For example, bought directly from CareFirst or another company or bought through an insurance broker.) Health insurance bought directly by someone else

(ASK Q.H1ba IF Q.H1b = 1) H1b. Is this an individual policy or is it a family policy? (READ IF NECESSARY: The health insurance through (your/TARGETs) work or union?) 1 2 D R Individual policy Family policy (covers more than one person) (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK IF H1c=1) Deleted instruction for & TAGE<26 H1ca. Is this an individual policy or is it a family policy? New wording (READ IF NECESSARY: The health insurance through someone elses work or union?) 1 2 D R Individual policy Family policy (covers more than one person) (DO NOT READ) Dont know (DO NOT READ) Refused

New Question (ASK IF H1c=1 & TAGE<26) H1c. Is this through (your/TARGETs) parent or guardian? 10

(READ IF NECESSARY: The health insurance through someone elses work or union?) 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK IF H1m=1) H1mm. Is this an individual policy or is it a family policy? (READ IF NECESSARY: The health insurance bought directly by someone else?) 1 2 D R Individual policy Family policy (covers more than one person) (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK IF H1m=1 & TAGE<26) H1mbm. Is this through your/TARGETs parent or guardian? (READ IF NECESSARY: The health insurance bought directly by someone else?) 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

H1n. DELETED (ASK IF ALL IN H1=2, D, OR R) H2. Do you currently have any other type of insurance? (DO NOT READ, ALLOW MULTIPLE) 1 2 3 4 5 7 N D R Workers compensation for specific injury/illness Employer pays for bills, but not an insurance policy Family member pays out of pocket for any bills Other Non Insurance Payment Source Indian Health Service (IHS) Other Insurance (SPECIFY) __________ No other insurance (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.H1 = 2, D OR R TO ALL AND Q.H2 = 1-4 ONLY, READ:) For the purposes of this survey, well assume that (you do/TARGET does) not have health insurance. NOW GO TO Q.H6 IF H2=5 ONLY SKIP TO H3C 11

(ASK Q.H3 IF Q.H1 = 2, D, R FOR ALL AND Q.H2 = N, D OR R) INSERT you do IF Q.S10 = 1 INSERT NAME/INITIALS/RELATIONSHIP does IF Q.S10 = 2 H3. Just to be sure I have this right, (you do/TARGET does) not have health insurance coverage. Is that correct? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.H3a IF Q.H3 = 2) INSERT you IF Q.S10 = 1 INSERT they IF Q.S10 = 2 H3a What insurance do (you/they) have? (DO NOT READ, ENTER ONE ONLY) (Probe: if you can, it might be helpful to look at (your/their) insurance card to help identify the type of insurance.) 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 97 DD RR DELETED Health insurance through (your / TARGETs) work or union Health insurance through someone elses current work or union Medicare Railroad Retirement Plan Veteran's Affairs, Military Health, TRICARE or CHAMPUS Indian Health Service Medicaid or DC Healthy Families New verbiage The Alliance or DC Health Care Alliance New verbiage DELETED Student health plan Health insurance bought directly by (you / TARGET) Health insurance bought directly by someone else DELETED Workers compensation for specific injury/illness Employer pays for bills, but not an insurance policy Family member pays out of pocket for any bills Other Non Insurance Payment Source Other Insurance (SPECIFY) __________ (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.H3a = 15-18, READ:) INSERT you do IF Q.S10 = 1 INSERT TARGET does IF Q.S10 = 2 For the purposes of this survey, well assume that (you do/TARGET does) not have health insurance. 12

NOW GO TO Q.H6 (ASK Q.H3b IF Q.H3a = DD OR RR) INSERT you/your IF Q.S10 = 1 INSERT they/their IF Q.S10 = 2 H3b When (you/they) go to a doctor, health clinic, or hospital, does anyone else pay for some or all of (your / their) medical bills? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.H3c IF Q.H2=5 OR IF Q.H3a = 07) INSERT you receive/your/you IF Q.S10 = 1 INSERT TARGET receives/TARGETs/they IF Q.S10 = 2 H3c. I understand that (you receive / TARGET receives) services through the Indian Health Service. In addition to this, does anyone else pay for (your / TARGETs) bills when (you/they) go to a doctor or hospital? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.H3c = 2, D, OR R, READ:) INSERT you do IF Q.S10 = 1 INSERT TARGET does IF Q.S10 = 2 For the purposes of this survey, Indian Health Service is not considered comprehensive insurance. For our survey, well assume that (you do/TARGET does) not have health insurance. NOW GO TO Q.H6 (ASK Q.H4 IF Q H3b=1 OR Q.H3c = 1) H4. And who is that? (DO NOT READ, ENTER ONE ONLY) 01 02 03 04 05 06 07 08 09 DELETED Health insurance through (your / TARGETs) work or union Health insurance through someone elses work or union Medicare Railroad Retirement Plan Veteran's Affairs, Military Health, TRICARE or CHAMPUS Indian Health Service Medicaid or DC Healthy Families The Alliance or DC Health Care Alliance 13

10 11 12 13 14 15 16 17 18 97 DD RR

DELETED Student health plan Health insurance bought directly by (you / TARGET) Health insurance bought directly by someone else DELETED Workers compensation for specific injury/illness Employer pays for bills, but not an insurance policy Family member pays out of pocket for any bills Other Non Insurance Payment Source Other Insurance (SPECIFY) __________ (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.H4 = 07 OR 15-18, READ:) INSERT you do IF Q.S10 = 1 INSERT TARGET does IF Q.S10 = 2 For purposes of this survey, well assume (you do/TARGET does) not have insurance. NOW GO TO Q.H6 (ASK Q.H-4a IF Q.H3a = 2 OR 3 OR Q.H4 = 2 OR 3) INSERT your IF Q.S10 = 1 INSERT TARGETs IF Q.S10 = 2 H4a. Is this an individual policy or is it a family policy? (READ IF NECESSARY IF Q.H3a OR Q.H4 = 2: The health insurance through (your/TARGETs) work or union?) (READ IF NECESSARY IF Q.H3a OR Q.H4 = 3: The health insurance through someone elses work or union?) 1 2 D R Individual policy Family policy (covers more than one person) (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.H4b IF Q.H3a OR Q.H4 = 13 AND IF TARGET<26 [Q.S6 <26 OR Q.S6(b-j) <26]) INSERT your IF Q.S10 = 1 INSERT TARGETs IF Q.S10 = 2 H4b. Is this through (your/TARGETs) parent or guardian? (READ IF NECESSARY: The health insurance bought directly by someone else?) 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

14

SETUP2 P.N.: CREATE VARIABLES FOR INSURANCE STATUS TO USE IN REST OF SURVEY TINS 1 TARGET is insured (Q.H1b-f = 1 OR Q.H1h-i/k-m = 1 OR H2 = 7 OR H3a = 02-06, 08-13 OR H4 = 02-06, 08-13, 97) 0 NOT insured {[Q.H1=2, D, or R to all AND (Q.H2 = 1-4 OR Q.H3 = 2, D, or R)] OR (Q.H3a = 15-18) OR (Q.H4 = 07 or 15-18)} TESI = 1 IF TARGET HAS INSURANCE THROUGH OWN WORK OR UNION (Q.H1b = 1) OR (Q.H3a = 02) OR (QNH4 = 02)) TFAM = 1 IF TARGETS FAMILY OR SPOUSE HAS FAMILY COVERAGE THROUGH OWN WORK OR UNION (ADD IN DEFINITION HERE) (ASK IF TINS=1) IF TAGE>=1 USE 1ST VERBIAGE IN PARENS IF TAGE<1 USE 2ND VERBIAGE IN PARENS INSERT Have you IF Q.S10 = 1 INSERT Has TARGET IF Q.S10 = 2 H5. [(Have you/Has TARGET) had insurance coverage for all of the past 12 months?] [Has TARGET had insurance coverage for all of the time since he/she was born?] 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.H6 IF Q.H5 = 2) INSERT you IF Q.S10 = 1 INSERT TARGET IF Q.S10 = 2 H6. How many months during the past 12 months were (you / TARGET) without health insurance coverage? __________ months (RANGE 1-12) 00 Less than 1 month DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK Q.H7 IF TINS = 0) INSERT you IF Q.S10 = 1 INSERT TARGET IF Q.S10 = 2 H7. How long has it been since (you/TARGET) had any health insurance? (PROBE FOR MONTHS IF LESS THAN 2 YEARS)

15

01 02 00 NN DD RR

ANSWER GIVEN IN YEARS __________# (2-50) years ANSWER GIVEN IN MONTHS__________# (1-24) months Less than 1 month NEVER HAD COVERAGE (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.H8 IF TINS = 0) INSERT you IF Q.S10 = 1 INSERT TARGET IF Q.S10 = 2 (SCRAMBLE ITEMS) H8. Im going to read a list of reasons that people sometimes give for why they dont have health insurance. Please tell me if these are reasons that (you/target) (do/does) not have health insurance? How about (INSERT)? 1 2 D R a. b. c. d. e. g. h. i. k. j. Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

The person in family who had health insurance lost job or changed employers The person in family who had health insurance is no longer part of the family because of divorce, separation or death Family members employer does not offer coverage or not eligible for employers coverage Lost eligibility for Medicaid or the Alliance new verbiage Cost is too high Dont need insurance Dont know how to get insurance Traded health insurance for another benefit or higher pay Insurance company refused coverage new code Some other reason (SPECIFY) __________________

(ASK Q.H9 IF Q.H7 = 01, 02, 00, OR DD) INSERT you IF Q.S10 = 1 INSERT TARGET IF Q.S10 = 2 (ASK ITEMS 02, 05, 12 if TAGE>17) (ASK ITEM 11 if TAGE>15) H9 Thinking back to the last time (you/TARGET) had health insurance, what type of insurance did (you/TARGET) have? (DO NOT READ LIST. UP TO 5 RESPONSES ALLOWED) 01 DELETED 02 Health insurance through (your/TARGETs) work or union 16

03 04 05 06 07 08 09 10 11 12 13 14 15 95 96 97 98 99 DD RR

Health insurance through someone elses work or union Medicare Railroad Retirement Plan Veteran's Affairs, Military Health, TRICARE or CHAMPUS Indian Health Service Medicaid or DC Healthy Families The Alliance or DC Health Care Alliance DELETED Student health plan Health insurance bought directly by you Health insurance bought directly by someone else DELETED Other Non Insurance Payment Source Other Insurance#1 (SPECIFY) __________ Other Insurance#2 (SPECIFY) __________ Other Insurance#3 (SPECIFY) __________ Other Insurance#4 (SPECIFY) __________ Other Insurance#5 (SPECIFY) __________ (DO NOT READ) Dont know (DO NOT READ) Refused

H10 DELETED The next questions concern the health insurance that the other people in your household have at this time. In answering these questions, please exclude any health insurance plans that cover only ONE type of service, like plans for dental care or prescription drugs. (PN: PROCEED DOWN HH ROSTER LIST. ASK Q.S13 THRU Q.I7 IN A SERIES FOR EACH PERSON IN HH EXCEPT TARGET) INSERT Do you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH S13(a-j). Do you/ Does your (RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) currently have health insurance? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused SKIP TO I1 SKIP TO I3 REPEAT FOR NEXT PERSON HH ROSTER REPEAT FOR NEXT PERSON HH ROSTER

(ASK Q.I1 IF Q.S13 = 1) (ASK ITEM c IF I1b < 1 >) (ASK ITEMS b and l if S6>17 or S6b=2) (ASK ITEM l IF I1b AND I1c < 1 >) (ASK ITEM m IF I1l AND I1b AND I1c <1 >) 17

(ASK ITEM k if S6>15 AND </=64 OR I1b AND I1C < 1 >) INSERT are you IF PERSON ASKING ABOUT IS RESPONDENT INSERT is this person IF OTHER PERSON IN HH (UP TO 5 RESPONSES ALLOWED) I1. What type of health insurance (are you/is this person) covered by? Is it (INSERT)? 1 Yes 2 No D (DO NOT READ) Dont know R (DO NOT READ) Refused IF RESPONDENT ASKS TO SKIP THROUGH INSURANCE QUESTIONS, SAY: Im sorry, but I have to read all of the insurance categories. a. b. DELETED Health insurance through (your/his/her) work or union? (IF TAGE<18, ADD PROBE FOR FIRST ADULT: This insurance could also be through COBRA, through a former employer or a retiree benefit.) Health insurance through someone elses work or union (Probe: This insurance could also be through COBRA, a former employer or a retiree benefit.) Medicare (IF TAGE<18, ADD PROBE FOR FIRST ADULT: Medicare is the health insurance for persons 65 years old and over or persons with disabilities. For many people, this will be a red, white and blue card.) Veteran's Affairs, Military Health, TRICARE or CHAMPUS DELETED Medicaid or DC Healthy Families (PROBE: This is a District program for low- and moderate-income families with children, seniors, and people with disabilities. (You/he/she) may have coverage under Medicaid or DC Healthy Families through a health insurance plan such as Health Right, Chartered, or Unison.) The Alliance or DC Health Care Alliance? (PROBE: This is another DC program that provides insurance at no cost for low- and moderate-income residents. (You/he/she) might have coverage under the Alliance through a health insurance plan such as Health Right, Chartered, or Unison.) DELETED A qualifying student health insurance plan (IF TAGE<18, ADD PROBE FOR FIRST HH MEMBER AGE >16: This is a health insurance plan that is sponsored by a college or university.) Health insurance bought directly by (you/him/her) (IF TAGE<18, ADD PROBE FOR FIRST ADULT: For example, bought directly from CareFirst or another company or bought through an insurance broker.) Health insurance bought directly by someone else

c.

d.

f. g. h.

i.

j. k.

l.

m.

18

I1n. DELETED (ASK Q.I1bb IF Q.I1b = 1) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT he IF OTHER PERSON IN HH AND Q.S7 = 1 INSERT she IF OTHER PERSON IN HH AND Q.S7 = 2 I1b(b-j). Is the insurance through work an individual policy or is it a family policy? (READ IF NECESSARY: The health insurance through (your/his/her) work or union?) 1 2 D R Individual policy Family policy (covers more than one person) (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK IF ALL IN I1=2, D, OR R) INSERT Do you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does he IF OTHER PERSON IN HH AND Q.S7 = 1 INSERT Does she IF OTHER PERSON IN HH AND Q.S7 = 2 I2(a-j). (Do you/Does he/she) currently have any other type of insurance? (DO NOT READ, SELECT ANSWERS) 1 2 3 4 5 6 N D R Workers compensation for specific injury/illness Employer pays for bills, but not an insurance policy Family member pays out of pocket for any bills Other Non Insurance Payment Source Indian Health Service (IHS) Other Insurance (SPECIFY)__________ No other insurance (DO NOT READ) Dont know (DO NOT READ) Refused

(IF Q.I2 = 5 ONLY, READ) The Indian Health Service) is not considered comprehensive insurance for the purposes of this survey NOW GO TO Q.I3 (IF Q.I2 = 1-4 ONLY, READ:) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT he IF OTHER PERSON IN HH AND Q.S7 = 1 INSERT she IF OTHER PERSON IN HH AND Q.S7 = 2 For the purposes of this survey, well assume that (you/he/she) does not have health insurance. NOW GO TO Q. I5

19

(ASK Q.I3 IF Q.S13 = 2 OR Q.I2 = N,D,R) INSERT You do IF PERSON ASKING ABOUT IS RESPONDENT INSERT Your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) does IF OTHER PERSON IN HH I3(a-j). Just to be sure I have this right. (You do/RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) does) not have health insurance coverage. Is that correct? 1 2 D R Yes No (DO NOT READ) (DO NOT READ)

(IF Q.I3 = 2 ASK Q.I3a) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT he or him IF OTHER PERSON IN HH AND Q.S7 = 1 INSERT she or her IF OTHER PERSON IN HH AND Q.S7 = 2 I3a(b-j). What insurance do you/they have? (Probe: If you can, it might be helpful to look at your/their insurance card to help identify the type of insurance.) (DO NOT READ, ENTER ONE ONLY) 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 97 DD RR DELETED Health insurance through (your / TARGETs) current work or union Health insurance through someone elses current work or union Medicare Railroad Retirement Plan Veteran's Affairs, Military Health, TRICARE or CHAMPUS Indian Health Service Medicaid or DC Healthy Families The Alliance or DC Health Care Alliance DELETED Student health plan Health insurance bought directly by (you / TARGET) Health insurance bought directly by someone else DELETED Workers compensation for specific injury/illness Employer pays for bills, but not an insurance policy Family member pays out of pocket for any bills Other Non Insurance Payment Source Other Insurance (SPECIFY) __________ (DO NOT READ) Dont know (DO NOT READ) Refused

(IF HH MEMBER IS UNINSURED (I3a = 07,14 15-18), READ: INSERT You do IF PERSON ASKING ABOUT IS RESPONDENT INSERT Your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER 20

FROM Q.S7(b-j) does IF OTHER PERSON IN HH For the purposes of this survey, well assume that (you do/TARGET does) not have health insurance.) IF HH MEMBER IS UNINSURED AND AGE IS UNKNOWN AND LAST PERSON, SKIP TO SETUP3 IF HH MEMBER IS UNINSURED AND AGE IS KNOWN, SKIP TO Q.I5 IF AGE IS UNKNOWN AND LAST PERSON AND INSURED, SKIP TO SETUP 3 IF AGE IS KNOWN AND INSURED GO TO QI4 (MOVED VARIABLE SPEC) Create MINS Variables: MINS A - J (Whether other people are insured) MINS 1 HH member is insured if: (Q.I1b-d,f,m OR h-l = 1 TO ANY OR Q.I2 = 6 OR Q.I3a = 02-06, 08-09, 11-13, 97). 0 HH is NOT insured if: S13=2 AND I3 = 2 AND MINS <> 1 IF MINS=1 SKIP TO NEXT PERSON OR IF LAST PERSON SKIP TO SETUP3 (ASK Q.I4 IF Q.I1b-f OR h-i, k-m = 1 TO ANY OR Q.I2 = 6 OR Q.I3a = 02-06, 08-09, 1113, 97) IF TAGE>=1, USE 1ST VERBIAGE IN PARENS IF TAGE<1 USE 2ND VERBIAGE IN PARENS INSERT Have you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Has your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH VALID AGE RANGE IS <18 OR 18+ I4. [(Have you/Has your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) had insurance coverage for all of the past 12 months?] [Has your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) had insurance coverage for all of the time since he/she was born?] 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK IF Q.I4=2; ELSE SKIP TO NEXT PERSON OR IF LAST PERSON SKIP TO SETUP3) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH I5. How many months during the past 12 months were (you / your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) without 21

health insurance coverage? __________# months (RANGE 1-12) LL Less than 1 month DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK IF MINS=0) (ASK Q.I6 IF Q.I3 =1, D, OR R OR Q.I3a = 07, 14 15-18) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH I6. How long has it been since (you/your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) had any health insurance? (PROBE FOR MONTHS IF LESS THAN 2 YEARS) 01 02 LL NN DD RR ANSWER GIVEN IN YEARS __________# (2-50) years ANSWER GIVEN IN MONTHS __________# (1-24) months LESS THAN ONE MONTH NEVER HAD COVERAGE (DO NOT READ) Dont know (DO NOT READ) Refused

IF Q.I6 = NN SKIP TO NEXT PERSON OR IF LAST PERSON SKIP TO SETUP3 (ASK Q.I7 IF Q.I6 NE NN) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j)/he/she IF OTHER PERSON IN HH (ASK ITEMS 02, 04 if TAGE>17) (ASK ITEM 11 if TAGE>15) I7. Thinking back to the last time (you/your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) had health insurance, what type of insurance did (you/he/she) have? (DO NOT READ, UP TO 5 RESPONSES ALLOWED) 01 02 03 04 05 06 07 08 09 10 11 Cobra - DELETED Health insurance through work or union Health insurance through someone elses work or union Medicare Railroad Retirement Plan Veteran's Affairs, Military Health, TRICARE or CHAMPUS Indian Health Service Medicaid or DC Healthy Families The Alliance or DC Health Care Alliance DELETED Student health plan 22

12 13 14 15 95 96 97 98 99 DD RR I8

Health insurance bought directly by you/him/her Health insurance bought directly by someone else DELETED Other Non Insurance Payment Source Other Insurance #1(SPECIFY) __________ Other Insurance #2 (SPECIFY) __________ Other Insurance # 3(SPECIFY) __________ Other Insurance # 4 (SPECIFY) __________ Other Insurance #5 (SPECIFY) __________ (DO NOT READ) Dont know (DO NOT READ) Refused

DELETED

NOW SKIP TO NEXT PERSON OR IF LAST PERSON SKIP TO SETUP3 SETUP3 CREATE VARIABLES FOR INSURANCE STATUS OF TARGETS SPOUSE/PARTNER AND/OR PARENT/GUARDIAN SESI = 1 IF SPOUSE/PARTNER HAS INSURANCE THROUGH OWN WORK OR UNION (Q.S12b-j = 01 or 02 AND [Q.I1b = 1 OR Q.I3a = 02]) OR (IF TARGET = RESPONDENT AND Q.S8b-j = 01 OR O2 AND [Q.I1b = 1 OR Q.I3a = 02]) PESI = 1 IF ANY PARENT/GUARDIAN HAS INSURANCE THROUGH OWN WORK OR UNION (Q.S12b-j = 04 AND [Q.I1b = 1 OR Q.I3a = 02]) OR (IF TARGET = RESPONDENT AND Q.S8b-j = 04 AND [Q.I1b = 1 OR Q.I3a = 02]) SFAM=1 IF SPOUSE HAS FAMILY COVERAGE THROUGH OWN WORK OR UNION ([IF RESPONDENT = TARGET AND S8b-j = 01] OR S12b-j = 01]) AND [(Q.I1b =1 AND I1bb=2) OR [Q.I3a=02} PFAM=1 IF TARGETS PARENT/GUARDIAN HAS FAMILY COVERAGE THROUGH OWN WORK OR UNION IF (TARGET = 03 IN S8b-j OR S12b-j=4), AND [(Q. I1b = 1 and QI1bb=2) AND (TARGET<26) EMPLOYMENT INSTRUCTIONS FOR SEQUENCE E1 TO E14: IF TAGE<15: PROCEED THROUGH ROSTER FOR TARGETS PARENTS OR GUARDIAN IF TAGE= 15-25: PROCEED THROUGH ROSTER FOR TARGET, TARGETS SPOUSE (IF PRESENT) AND TARGETS PARENTS (IF PRESENT) IF TAGE>25: PROCEED THROUGH ROSTER FOR TARGET & TARGETS SPOUSE (IF PRESENT) My next questions ask about employment.

23

Id like to start by asking about (you/TARGETRELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j). FOR SUBSEQUENT PEOPLE: Now Id like to ask about (you/TARGET/RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j). INSERT Are you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Is TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Is your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E1(a-f). (Are you /Is TARGET/Is your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j) currently 1 2 3 4 5 6 7 D R Self-employed Employed by military Employed by someone else Unpaid worker for a family business Retired Unemployed and looking for work Unemployed and not looking for work (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.E2 IF Q.E1 = 1-3) INSERT Do you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Does your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E2(a-f). (Do you/Does TARGET/Does your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j) have more than one job, including part-time, evening or weekend work? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.E3 IF Q. E2 = 1) INSERT Do you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Does your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E3(a-f). Altogether, how many jobs (do you/does TARGET/ does your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j) have? 24

1 2 3 D R

2 jobs 3 jobs 4 or more jobs (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.E4 IF Q.E1 = 1-3) INSERT Do you/your/you/work IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does TARGET/their/he/works IF PERSON ASKING ABOUT IS TARGET PERSON AND Q.S7 = 1 INSERT Does TARGET/their/she/works IF PERSON ASKING ABOUT IS TARGET PERSON AND Q.S7 = 2 INSERT Does your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j)/their/he/she IF OTHER PERSON IN HH E4(a-f). How many hours per week (do you /does TARGET/(does your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j) usually work at (your/their) (READ IF E2=1: main) job? IF NEEDED: By main job, I mean the one at which (you/he/she) usually (work/works) the most hours. __________ hours (1-100) DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK Q.E5 IF Q.E2 = 1) INSERT Do you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does TARGET/their IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Does your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j)/their IF OTHER PERSON IN HH E5(a-f). How many hours per week (do you /does TARGET/does your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) usually work at (your/their) other jobs? ________ hours (ENTER # OF HOURS 1-100) DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK Q.E6 IF Q.E1=1-3) INSERT have you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT has TARGET/their IF PERSON ASKING ABOUT IS TARGET PERSON INSERT has your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j)/their IF OTHER PERSON IN HH E6(a-f). How long (have you/has TARGET/has your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) worked for at (your/their) 25

(READ IF E2=1: main) job? PROBE FOR MONTHS IF LESS THAN 2 YEARS 01 ANSWER GIVEN IN YEARS (ENTER # OF YEARS 2-60) 02 ANSWER GIVEN IN MONTHS (ENTER # OF MONTHS 1-24) LL Less than 1 month DD (DO NOT READ) Dont know RR (DO NOT READ) Refused NEW (ASK Q.E6a IF Q.E1=6-7 AND((PARENT or GUARDIAN or SPOUSE/UMMARRIED PARTNER) OR( TARGET_AGE > 17))) INSERT have you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT has TARGET/their IF PERSON ASKING ABOUT IS TARGET PERSON INSERT has your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j)/their IF OTHER PERSON IN HH E6a(a-j). How long (have you/has TARGET/has your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) been unemployed? PROBE FOR MONTHS IF LESS THAN 2 YEARS 01 ANSWER GIVEN IN YEARS (ENTER # OF YEARS 2-60) 02 ANSWER GIVEN IN MONTHS (ENTER # OF MONTHS 1-24) LL Less than 1 month DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK Q.E6b IF Q.E1 = 3) NEW QUESTION INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E6b(a-j). (Is your/is TARGETs/(is your (relationship)s [(age) (sex) if multiple members with same relationship code]) (READ IF Q.E2=1: main) job located in the District of Columbia, Maryland, Virginia, or somewhere else? 1 2 3 4 D R District of Columbia Maryland Virginia Somewhere else (SPECIFY)_________ Dont Know Refused

(ASK Q.E7 IF Q.E1 = 3) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT 26

INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E7(a-f). Counting all locations where this employer operates, are there more than 50 people working for (your/TARGETs/your RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) employer? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.E8 IF Q.E7 = 2) INSERT business IF Q.E1 = 1; OTHERWISE, INSERT employer INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E8(a-f). Which category best represents the total number of persons who work for (your/TARGETs your RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) (employer/business)? Would it be? 1 2 3 D R Just one Between 2 and 10 Between 11 and 50 (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.E9 IF Q.E7 = 1) INSERT business IF Q.E1 = 1; OTHERWISE, INSERT employer INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E9(a-f). Which category best represents the total number of persons who work for (your/TARGETs/ your RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) (employer/business)? Would it be? 1 2 3 4 Between 51 and 100 Between 101 and 500 Between 501 and 1000 Over 1000 27

D R

(DO NOT READ) Dont know (DO NOT READ) Refused

NEW QUESTION (ASK Q.E9a IF Q.E1 = 3) INSERT Do you/your/you/work IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does TARGET/their/he/works IF PERSON ASKING ABOUT IS TARGET PERSON AND Q.S7 = 1 INSERT Does TARGET/their/she/works IF PERSON ASKING ABOUT IS TARGET PERSON AND Q.S7 = 2 INSERT Does your RELATIONSHIP FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j)/their/he/she IF OTHER PERSON IN HH E9a(a-f). What type of business is the place where (you / TARGET/( your RELATIONSHIP FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j) usually (work/works) at (your/their) (READ IF E2=1: main) job? IF NEEDED: By main job, I mean the one at which (you/he/she) usually (work/works) the most hours. IF NEEDED: For example, restaurant, hotel, store/sales/retail, construction, hair dresser/barber, architecture/graphic design, law firm, medical office, housekeeping/cleaning, day care, or the like. ________________ (fill in one answer only) DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK Q.E10 IF (TESI = 0 OR IF SESI = 0 OR IF ((S8(b-j)=03 for TARGET person) or (Respondent=TARGET AND TAGE=18-25 AND S8(b-j)=04) or (S12(b-j)=04)) AND I1b(bj)=2,D, or R OR I3a(b-j) NE 2)) AND E1 = 1 3) INSERT you work/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET works/their IF PERSON ASKING ABOUT IS TARGET PERSON INSERT at your RELATIONSHIPs FROM Q.S8(b-j) main job/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) works/at their main job IF OTHER PERSON IN HH if E2=1 E10(a-j). Does the place where (you work/TARGET works/your RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j] works) (at (your main job/at their main job) offer health insurance as a benefit to any of its employees? INTERVIEWER: ASK ABOUT PRIMARY EMPLOYER IF HAVE MULTIPLE EMPLOYERS 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK E11 IF E10=1,D,R) 28

INSERT Are you and in your IF PERSON ASKING ABOUT IS RESPONDENT INSERT Is he/she and in his/her IF PERSON ASKING ABOUT IS TARGET PERSON E11(a-j). (Are you /Is TARGET/Is your RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) eligible to participate in (your/his/her) employers health plan, are (you/he/she) currently in the waiting period to be eligible OR are (you/he/she) not eligible to participate in (your/his/her) employers health plan? 1 2 3 D R Yes, eligible to participate Currently in the waiting period to be eligible Not eligible to participate (DO NOT READ) Dont know (DO NOT READ) Refused

NEW QUESTION (ASK E11a IF NOT RETIRED AND TARGET AND TESI=1 OR SPOUSE AND SESI=1 OR (TARGETS PARENT AND ((S8(b-j)=03 for TARGET person) or (Respondent=TARGET AND TAGE=18-25 AND S8(b-j)=04) or (S12(b-j)=04)) AND I1b(bj)=1 OR I3a(b-j)=2)) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT their IF PERSON ASKING ABOUT IS TARGET PERSON INSERT your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) /their IF OTHER PERSON IN HH E11a(A-F). Was obtaining health insurance through (your/their/your RELATIONSHIPs) job an important factor in (your/their/your RELATIONSHIPs) job choice? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK E12 IF TARGET AND TESI=1 AND H1ba=1,D,R OR SPOUSE AND SESI=1 AND I1bb=1,D,R OR (TARGETS PARENT AND ((S8(b-j)=03 for TARGET person) or (Respondent=TARGET AND TAGE=18-25 AND S8(b-j)=04) or (S12(b-j)=04)) AND I1b(bj)=1) AND I1bb=1,D,R) OR IF E11=1, 2, D, R) (IF TARGET AND TESI=1 AND H1ba=1,D,R OR SPOUSE AND SESI=1 AND I1bb=1,D,R OR TARGETS PARENT AND PESI=1 AND I1bb=1,D,R, INCLUDE Earlier you mentioned that AND INSERT you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET/their IF PERSON ASKING ABOUT IS TARGET PERSON INSERT your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) /their IF OTHER PERSON IN HH) (IF TARGET AND TESI=1 AND H1ba=1,D,R OR SPOUSE AND SESI=1 AND 29

I1bb=1,D,R OR TARGETS PARENT AND PESI=1 AND I1bb=1,D,R OR IF E11=1, 2, D,R, ASK Could dependents be covered under that health insurance.) E12(a-f). Earlier you mentioned that (you / TARGET/ your RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) had health insurance coverage through (your/their) employer. Could dependents be covered under that health insurance? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

ASK IF TAGE OR S6 > 16 or S6a1 = 2 Added INSERT Are you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Is TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Is your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E13(a-j). (Are you/Is TARGET/Is your RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) a veteran of the United States military? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK EVERYONE) INSERT Are you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Is TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Is your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH E14(a-j). (Are you/Is TARGET/Is RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) currently a full-time student? (IF NECESSARY: greater than three-fourths time)? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

ACCESS, USE AND COST (ASK EVERYONE) 30

INSERT your/you/go/are IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs/goes/he/she/is/his/her IF PERSON ASKING ABOUT IS TARGET PERSON A1. My next questions ask about [your/ TARGETs] recent health care experiences. Is there a place where [you/ TARGETs] usually [go/goes] when [you/(he/she)] [are/is] sick or when [you/(he/she)] need advice about [your/(his/her)] health? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.A2 IF Q.A1 = 1) A2. What kind of place is it? Is it.? (READ LIST. ENTER ONE ONLY) 1 2 3 4 5 6 7 D R A doctors office or private clinic A community health center or other public clinic A hospital outpatient department A hospital emergency room An urgent care center that is not part of a community health center Or, some other place (DO NOT READ) Dont go to one place most often (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK EVERYONE) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON A3. The next questions are about the health care [you/ TARGET] received in the past 12 months. In the past 12 months, how many times did [you/ TARGET] receive care in a hospital emergency room? (READ LIST IF NECESSARY. ENTER ONE ONLY) 0 1 2 3 4 D R None 1 time 2 times 3 times More than 3 times (DO NOT READ) Dont know (DO NOT READ) Refused

NEW ASKQ.A3aa if Q.A3>=1 INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON 31

A3aa. Were [you/TARGET] admitted overnight any of these times? 1 Yes 2 No D Dont Know R Refused NEW ASK A3bb if A3aa = 1 (PN: Gen in A3bb=1 if (A3=1 and A3aa=1); Show only relevant responses, based on answer to Q.A3aa) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON A3bb. How many of these times [were you/was Target] admitted overnight? 1 2 3 4 D R 1 time 2 times 3 times More than 3 times Dont Know Refused

(ASK Q.A3a IF Q.A3 = 1-4) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT TARGETs parent IF TAGE<18) A3(a-j). The last time (you/TARGET) went to a hospital emergency room, was it for a condition that (you/TARGET/TARGETS parent) thought could have been treated by a regular doctor if he or she had been available? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.A3b IF Q.A3a = 1) (IF TAGE = <18 INSERT Targets Parent A3b. I'm going to read you a list of reasons why some people go to the emergency room. Please tell me if any of these were important reasons for (your/TARGETs) last visit to a hospital emergency room. (INSERT) Was this an important reason? 1 2 D Yes No (DO NOT READ) Dont know 32

(DO NOT READ) Refused

a. (You were/TARGET was) unable to get an appointment at the doctor's office or clinic as soon as (you/TARGET/TARGETs PARENT) thought one was needed b. (You/TARGET) needed care after normal operating hours at the doctor's office or clinic c. (You/TARGET/TARGETs PARENT) owed money to the doctor's office or clinic d. It was more convenient to go to the hospital emergency room

A4, A5, A5aA6, A7, A8 Deleted (ASK EVERYONE) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Is your RELATIONSHIPs FROM Q.S8(b-j)/AGE FROM Q.S6(b-j)/GENDER FROM Q.S7(b-j) IF OTHER PERSON IN HH FOR ITEM a INSERT for TARGET IF TAGE<18 (SCRAMBLE) A9. Still thinking about the past 12 months, was there any time that [you/ TARGET/ RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]] did (INSERT) because of cost? 1 2 D R a. c. d. Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

e.

Not fill a prescription for medicine (for TARGET) Not get doctor care that (you/TARGET) needed Not get specialist care that (you/TARGET) needed (IF NEEDED: Specialists are doctors like surgeons, heart doctors, allergy doctors, skin doctors and others who specialize in one area of health care.) Not get dental care that (you/TARGET) needed

A9b Deleted A10 Deleted

33

(ASK EVERYONE) INSERT you when respondent is TARGET and TARGETs TFAM_COUNT= 1 INSERT your family when respondent is TARGET and TARGETs TFAM_COUNT> 1 INSERT TARGET/was he/was she/his/her when TARGET is not respondent and TARGETs TFAM_COUNT=1 INSERT TARGETs FAMILY/were they/their when TARGET is not respondent and TARGETs TFAM_COUNT> 1 A11. In the past 12 months, did [you/your family/TARGET/TARGET's family] have any problems paying or [were you/was he/was she/were they] unable to pay any of [your/his/her/their] medical bills? This would include doctor or hospital bills, dentist bills, bills for prescription drugs, nursing home bills, or home care bills. 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

HEALTH STATUS My next questions are about (your/TARGETs) health. (ASK EVERYONE) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON HS1. Would you say (your / TARGETs) health, in general, is excellent, very good, good, fair, or poor? 1 2 3 4 5 D R Excellent Very good Good Fair Poor (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK EVERYONE) INSERT Are you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT Is TARGET/his/her IF PERSON ASKING ABOUT IS TARGET PERSON HS2. (Are you/ Is TARGET) limited in any way in (your/his/her) activities because of a physical, mental, or emotional problem? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused 34

PUBLIC COVERAGE Now I have a few questions about insurance programs available through the District of Columbia for those who are uninsured. P.N. - FOR TARGET<18 ASK ABOUT TARGETS PARENTS OR GUARDIAN (ASK P1 THROUGH P4 IF Q.H3=1 OR Q.H2=1-5 ONLY OR Q.H3c=2,D, OR R OR Q.H4=07 OR15-18) INSERT Have you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Has TARGET IF PERSON ASKING ABOUT IS TARGET PERSON INSERT Has TARGETs parents or guardian IF TAGE<18) P1. New wording (Have you /Has TARGET/Has TARGETs parents or guardian) ever asked for or been given information about any of the programs available in the District for people who are uninsured? This would include Medicaid (or DC Healthy Families) or the Alliance (or DC Health Care Alliance). 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

INSERT you/ you were IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET/TARGET was/he/he was/she/she was IF PERSON ASKING ABOUT IS TARGET PERSON INSERT TARGETs parents or guardian/he/she/TARGET IF TAGE<18) P2. If (you / TARGET/TARGETs parents or guardian) learned (you were/(she/he was) /TARGET was) were eligible for health coverage through one of these programs, would (you / he/she) enroll (//TARGET)? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

35

(ASK P3 IF P2=2, D,R) INSERT you/you were/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET/TARGET was/he/he was/she/she was/her/his IF PERSON ASKING ABOUT IS TARGET PERSON INSERT TARGETs parents or guardian/he was/she was/her/his/TARGET IF TAGE<18) P3. If (you / TARGET/TARGETs parents or guardian) learned (you were/ (she/he) was/ TARGET was) were eligible for health coverage through one of these programs at no cost to (you / TARGET) or (your /her/his) family, would (you /she/he) enroll (//TARGET)? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON ACCEPT ONE RESPONSE ONLY P4. Im going to read you a list of reasons people sometimes give for not enrolling in programs. Please tell me which of these is the most important reason (you are / TARGET is) not enrolled in one of the programs available in the District for people who are uninsured? (INTERVIEWER: READ LIST. ENTER ONE RESPONSE.) 01 02 03 04 05 06 07 08 09 97 DD RR Have other insurance Cost was too high Dont like the benefits package Too much hassle/paperwork Dont need or dont want insurance Not eligible for coverage Dont know about programs Dont know how to enroll in programs Dont want to enroll in a public program Something else (SPECIFY) __________________ (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.P4A IF Q.P4 = 01-09, 97) (PN: IF Q.P4 = any of 01-09, do not show that response in Q.P4A) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON ALLOW MULTIPLE RESPONSES P4A. Please tell me which, if any, of these are also important reasons (you / TARGET) are not enrolled in one of the programs available in the District for people who are uninsured? 36

(INTERVIEWER: READ LIST. ENTER ALL THAT APPLY.) 01 02 03 04 05 06 07 08 09 97 NN DD RR Have other insurance Cost was too high Dont like the benefits package Too much hassle/paperwork Dont need or dont want insurance Not eligible for coverage Dont know about programs Dont know how to enroll in programs Dont want to enroll in a public program Something else (SPECIFY) __________________ No other reasons (DO NOT READ) Dont know (DO NOT READ) Refused

HR1.DELETED HR2. DELETED HR3. DELETED BACKGROUND Now, Id like to ask a few questions to help us describe the people who participated in our survey. (ASK EVERYONE) INSERT Are you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON New wording and list D1. (Are you/TARGET) Columbian, Guatemalan, Honduran, Mexican, Salvadoran, or some other Hispanic or Latino group? 1 2 3 4 5 6 7 D R No, not of Hispanic origin Columbian Guatemalan Honduran Mexican Salvadoran, Other Spanish/Hispanic/Latino please specify _________ (DO NOT READ) Dont know (DO NOT READ) Refused

37

(ASK EVERYONE) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON D2. Which one or more of the following would you say is (your/TARGETs) race? (READ LIST. ENTER ALL THAT APPLY) 1 2 3 4 5 7 D R White Black or African American Asian Native Hawaiian or other Pacific Islander American Indian or Alaska Native Some other race (SPECIFY) ______________ (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.D3 IF Q.S4>1) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON D3. Are all of the other people in this household of the same race and ethnicity as (you/TARGET)? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(PN: IF RACE IS SAME FOR ALL IN HH, GEN IN RESPONSES FOR D4/D5) (ASK IF D3 = 2, D OR R AND ASK Q.D4 AND Q.D5 IN A SERIES FOR EACH MEMBER IN HH)New wording and list D4. Is your (RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j]) Salvadoran, Mexican, Honduran, Columbian, Guatemalan, or another Hispanic or Latino group? 1 2 3 4 5 6 7 D R No, not of Hispanic origin Columbian Guatemalan Honduran Mexican Salvadoran, Other Spanish/Hispanic/Latino please specify _________ (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.D4 AND Q.D5 IN A SERIES FOR EACH MEMBER IN HH) 38

D5. Which one or more of the following would you say is (RELATIONSHIPs FROM Q.S8[b-j] [INSERT AGE/GENDER FROM Q.S6[b-j] AND S7 [b-j])s race? (READ LIST. ENTER ALL THAT APPLY) 1 2 3 4 5 7 D R White Black or African American Asian Native Hawaiian or other Pacific Islander American Indian or Alaska Native Some other race (SPECIFY) ______________ (DO NOT READ) Dont know (DO NOT READ) Refused

ASK D6 D6a FOR OUTBOUND PHONE SURVEY ONLY (Universe outbound phone calls only) D6. Next I have a few questions about the number I dialed to reach you. Is the number (INSERT PHONE NUMBER FROM SAMPLE) a cell phone? 1 2 D R Yes, Cell phone No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.D6a IF Q.D6 = 2) D6a. Is this telephone number? 1 2 D R Listed Unlisted (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK EVERYONE) (PN: IF D6=2, RANGE should be 1-10) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON D7. I would like to ask about any regular, landline telephone numbers in (your/TARGETs) household, that is, all your phone numbers other than cell phone numbers. How many different landline telephone numbers does your household have? ________ Landline phone numbers (RANGE 0-10) 11 11 or more DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK Q.D7a IF RDD SAMPLE AND Q.D7 = 0) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT 39

INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON D7a. Id like to verify the information you just provided. I believe you indicated that (your/TARGETs) household has no landline telephone service for incoming and outgoing calls. Is this correct? 1 2 D R Yes No (RE-ASK Q.D7) (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.D7b IF Q.D7 = 1) (IF Q.D6 = 2, GEN IN RESPONSE FROM Q.D6a) D7b. Is this an unlisted telephone number? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.D7c IF Q.D7 >1) (IF Q.D6a = 1, GEN IN CODE 1 FOR Q.D7c) D7c. Are any of these listed numbers? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.D8 IF Q.D7>1) D8. Excluding any landline numbers used exclusively for fax machines, security systems, the internet, or a professional business, how many of these (INSERT RESPONSE FROM Q.D7) landline telephone numbers are used for incoming calls? ________ Landline phone numbers (RANGE 0-10) DD (DO NOT READ) Dont know RR (DO NOT READ) Refused (ASK EVERYONE) (IF Q.D6 =1, GEN IN CODE 1 FOR Q.D9) INSERT Do you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT Does TARGET/their IF PERSON ASKING ABOUT IS TARGET PERSON D9. Do any members of your household currently have a working cell phone? 1 2 D Yes No (DO NOT READ) Dont know 40

(DO NOT READ) Refused

(ASK Q.D9a IF Q.D9 = 1) (IF Q.D9 =1, AND HH_COUNT=1, GEN IN 1 FOR D9a) (IF >HH_COUNT, READ: YOU SAID EARLIER THAT THERE WERE (HH_COUNT) PEOPLE IN YOUR HOUSEHOLD, HOW MANY HAVE A CELL PHONE?) D9a. How many people in the household have a cell phone? ________ number of people (RANGE 1-10) 11 11 or more DD (DO NOT READ) Dont know RR (DO NOT READ) Refused NEW QUESTION (ASK EVERYONE) D9A2 Do you have access to the internet at your home? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

NEW QUESTION (ASK Q.D9b if D9a=2, D OR R) D9b If you needed to access the internet, where would you most likely go? [DO NOT READ] 1 Workplace 2 Friend/relatives house 3 Library 4 Church 5 Community organization 6 School/college/university 7 Other 8 I never need to access the internet D (DO NOT READ) Dont know R (DO NOT READ) Refused (ASK EVERYONE) D10. Is this residence? 1 2 3 D R Owned by or being bought by you or someone in your household Rented for cash Occupied without payment of rent (DO NOT READ) Dont know (DO NOT READ) Refused 41

(PN: PLEASE CHECK THAT TARGET_AGE >= ANSWER GIVEN HERE, IF NOT GIVE ERROR MESSAGE) INSERT have you IF PERSON ASKING ABOUT IS RESPONDENT INSERT has TARGET IF PERSON ASKING ABOUT IS TARGET PERSON D11. How long (have you/has TARGET) lived in the District? IF NEEDED PROBE FOR MONTHS IF LESS THAN 2 YEARS, IF NEEDED: If (you/TARGET) moved away for a while, how long (have you/has TARGET) lived in the District since moving back to DC? 01 ANSWER GIVEN IN YEARS __________ # (2-100) years 02 ANSWER GIVEN IN MONTHS __________ # (0-24) months 03 Lived in the District for entire life LL (DO NOT READ) Less than one month DD (DO NOT READ) Dont know RR (DO NOT READ) Refused NEW QUESTION (ASK if (Q.D11 = DD, RR) OR (TARGET_AGE < 9 and D11=03) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON AND TARGET_AGE > 8 INSERT TARGETs family IF PERSON ASKING ABOUT IS TARGET PERSON AND TARGET_AGE <9 D11a. What year did (you/ TARGET/TARGETs family) last live outside of the District? RECORD YEAR _____ D Dont Know R Refused NEW QUESTION (ASK if ((Q.D11a = D or R) or # of years in D11 = 9 ) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON AND TARGET_AGE > 8 INSERT TARGETs family IF PERSON ASKING ABOUT IS TARGET PERSON AND TARGET_AGE <9 D11b. Have (you/TARGET/ TARGETs family) lived in the District continuously since the year 2000? 1 2 D R Yes No Dont Know Refused 42

NEW QUESTION (ASK if moved into the District in 2000 or later: (Q.D11a = 2000 - 2009 OR Q.D11b=2 OR response given in Q.D11 is LL or between 0 months and 8 years) INSERT you or your family IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET or his/her IF PERSON ASKING ABOUT IS TARGET PERSON AND D11c. Was getting health insurance for (you or your family/TARGET or his/her family) a consideration in making the decision to live in the District? 1 2 D R Yes No Dont Know Refused

NEW QUESTION (ASK if D11c = 1) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET or his/her family IF PERSON ASKING ABOUT IS TARGET PERSON D11d. 1 2 D R And would (you/Target or his/her family) say that getting health insurance was a primary reason in making the decision to live in the District? Yes No Dont Know Refused

INSERT Were you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Was TARGET IF PERSON ASKING ABOUT IS TARGET PERSON D12. (Were you/Was TARGET) born in the United States? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.D13 IF Q.D12 = 2) INSERT Are you IF PERSON ASKING ABOUT IS RESPONDENT INSERT Is TARGET IF PERSON ASKING ABOUT IS TARGET PERSON D13. (Are you/Is TARGET) a citizen of the United States? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused 43

IN1. My final questions are about income. This information is important because it helps the District understand how to make health care more affordable. (IF TAGE<19 & TMARR=0 & TPAR=0 & RESPONDENT IS PARENT AND RESPONDENT IS MARRIED, READ :) Im interested in your family income, that is your income PLUS the income of your immediate family. (By immediate family I mean your spouse or your unmarried partner and the children or stepchildren under 19 who are living with you). For these questions, Id like you to think back to 2008. During 2008, did you or any members of your immediate family, including your spouse or unmarried partner receive any income from wages or salary? (IF TAGE <19 & TMARR=0 & TPAR=0 & RESPONDENT IS PARENT AND RESPONDENT IS NOT MARRIED, READ :) Im interested in your family income, that is your income PLUS the income of your immediate family. (By immediate family I mean the children or stepchildren under 19 who are living with you). For these questions, Id like you to think back to 2008. During 2008, did you or any members of your immediate family receive any income from wages or salary? (IF TAGE <19 & TMARR=0 & TPAR=0 & RESPONDENT IS NOT PARENT AND RESPONDENT IS NOT THE UNMARRIED PARTNER OF TARGETS PARENT, READ:) Im interested in TARGETs family income, that is the income from his/her parents PLUS the income of any immediate family. (By immediate family I mean parents and siblings under 19 who are living with TARGET). For these questions, Id like you to think back to 2008. During 2008, did any members of TARGETs immediate family receive any income from wages or salary? (IF TAGE <19 & TMARR=0 & TPAR=0 & RESPONDENT IS THE UNMARRIED PARTNER OF TARGETS PARENT, READ:) Im interested in TARGETs family income, including your income, PLUS the income of any immediate family. (By immediate family I mean parents and siblings under 19 who are living with TARGET). For these questions, Id like you to think back to 2008. During 2008, did you or any members of TARGETs immediate familyreceive any income from wages or salary?

44

(IF TMARR=1 & TFAM_COUNT>2, READ :) INSERT your/you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs/his/her/TARGET IF PERSON ASKING ABOUT IS TARGET PERSON Im interested in [your/ TARGETs] family income, that is [your/ TARGETs] income PLUS the income of [your/his/her] immediate family. (By immediate family I mean [your/(his/her)] spouse or unmarried partner and the children or stepchildren under 19 who are living with [you/ TARGET]). For these questions, Id like you to think back to 2008. During 2008, did [you/ TARGET] or any members of [your/his/her] immediate family including your/his/her spouse or unmarried partner receive any income from wages or salary? (IF TMARR=1 & TFAM_COUNT = 2, READ :) INSERT your/you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs/his/her/TARGET IF PERSON ASKING ABOUT IS TARGET PERSON Im interested in the income of (your/TARGETs) immediate family, that is [your/ TARGETs] income PLUS the income of [your/his/her] spouse or your/his/her unmarried partner. For these questions, Id like you to think back to 2008. During 2008, did [you/ TARGET] or [your/his/her] spouse or unmarried partner family members receive any income from wages or salary? (IF TMARR=0 & TPAR=1 & TFAM_COUNT>2, READ :) INSERT your/you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs/his/her/TARGET IF PERSON ASKING ABOUT IS TARGET PERSON Im interested in [your/TARGETs] family income, that is [your/ TARGETs] income PLUS the income of [your/TARGETs] immediate family. By immediate family I mean [your/TARGETs] income plus the income of the children or stepchildren under 19 who are living with [you/ TARGET]. For these questions, Id like you to think back to 2008. During 2008, did [you/ TARGET] or members any of [your/(his/her)] immediate family receive any income from wages or salary? (IF TFAM_COUNT=1, READ :) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET IF PERSON ASKING ABOUT IS TARGET PERSON For these questions, Id like you to think back to 2008. During 2008, did [you/ TARGET] receive any income from wages or salary? 1 2 D R Yes No (DO NOT READ) Dont know (DO NOT READ) Refused 45

(ASK EVERYONE) (DONT SCRAMBLE) INSERT you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET/his/her IF PERSON ASKING ABOUT IS TARGET PERSON INSERT immediate family IF TFAM_COUNT > 1 IN2. During 2008, did [you/ TARGET] (or any members of [your/ his/her] immediate family) receive (INSERT)? 1 2 D R a. b. c. Yes No (DO NOT READ) Dont know (DO NOT READ) Refused

Any dividend income or any interest income from bonds, money market accounts, CDs or other investments Social Security benefits or any type of cash assistance Income from any other sources, such as self-employment, alimony, child support, contributions from family or others, unemployment compensation, workers compensation or veterans payments, pensions, or anything else

INSERT you/your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGET/TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT immediate family IF TFAM_COUNT > 1 IN3. Thinking about all the different sources of income [you/ TARGET] (and [your/ TARGETs] immediate family) received in 2008, what was the combined total income from all sources before taxes and other deductions? Was it under (INSERT AMT3 FOR FAMILY SIZE) or was it (INSERT AMT3 FOR FAMILY SIZE) or more? PROBE: Your best estimate is fine. 1 2 D R Under (INSERT AMT3) (INSERT AMT3) or more (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q,IN4 IF Q.IN3 = 1) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT family IF TFAM_COUNT > 1 IN4. Now, just stop me when I get to the right category. Was [your/ TARGETS] total (family) income ? IF NEEDED: The computer gives me different income values for the question depending on the size of your family. (IF NEEDED, PROBE: Your best estimate is fine) (READ LIST. ENTER ONE ONLY) 46

1 2 3 D R

Less than (INSERT AMT1) (INSERT AMT1) to (INSERT AMT2) (INSERT AMT2) to under (INSERT AMT3) (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.IN5 IF Q.IN3 = 2) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT family IF TFAM_COUNT > 1 IN5. Now, just stop me when I get to the right category. Was [your/ TARGETS] total (family) income ? PROBE: Your best estimate is fine. (READ LIST. ENTER ONE ONLY)

1 2 3 4 5 6 D R

(INSERT AMT3) to under (INSERT AMT4) (INSERT AMT4) to under (INSERT AMT5) (INSERT AMT5) to under (INSERT AMT6) (INSERT AMT6) to under (INSERT AMT7) (INSERT AMT7) to under (INSERT AMT8) (INSERT AMT8) or more (DO NOT READ) Dont know (DO NOT READ) Refused

IF TARGET IS RESPONDENT AND TARGET LIVES WITH UNMARRIED PARTNER (S8=02) OR S12=2) OR TARGET IS NOT RESPONDENT AND TARGET LIVES WITH UNMARRIED PARTNER S12=02 OR (TARGET IS< AGE 19 AND PARENT WHO IS RESPONDENT LIVES WITH UNMARRIED PARTNER S8=02 AND either S8=03 or S12=04 for the unmarried partner BUT NOT BOTH) READ Now I have a few questions just about you/TARGET and your/his/her immediate family, not including your/his/her unmarried partner AND CONTINUE TO Q.IN3A, OTHERWISE, GO TO Q.IN5a IF (TARGET IS< AGE 19 AND RESPONDENT WHO IS NOT THE PARENT HAS A UNMARRIED PARTNER WHO IS THE TARGETS PARENT, S12 FOR THAT PERSON = 04) READ Now I have a few questions just about TARGET and his/her immediate family, not including you AND CONTINUE TO Q.IN3A, OTHERWISE, GO TO Q.IN5a

(PN: INSERT INCOME AMT 3 FOR TFAM_COUNT 1) INSERT you/your IF PERSON ASKING ABOUT IS RESPONDENT AND RESPONDENT IS PARENT OF TARGET INSERT TARGET/his/her/you IF PERSON TALKING ABOUT IS RESPONDENT AND 47

RESPONDENT IS UNMARRIED PARTNER OF TARGETS PARENT INSERT TARGET/TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON IN3A. Thinking about all the different sources of income [you/ TARGET] (and [your/ TARGETs/his/her] immediate family not including your/his/her unmarried partner/you) received in 2008, what was the combined total income from all sources before taxes and other deductions? Was it under (INSERT AMT3 FOR FAMILY SIZE) or was it (INSERT AMT3 FOR FAMILY SIZE) or more? PROBE: Your best estimate is fine. 1 2 D R Under (INSERT AMT3) (INSERT AMT3) or more (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.IN4A IF Q.IN3A = 1) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT family IF TFAM_COUNT - 1 > 1 IN4A. Now, just stop me when I get to the right category. Was [your/ TARGETS] total (family) income ? IF NEEDED: The computer gives me different income values for the question depending on the size of your family. (IF NEEDED, PROBE: Your best estimate is fine) (READ LIST. ENTER ONE ONLY) 1 2 3 D R Less than (INSERT AMT1) (INSERT AMT1) to (INSERT AMT2) (INSERT AMT2) to under (INSERT AMT3) (DO NOT READ) Dont know (DO NOT READ) Refused

(ASK Q.IN5A IF Q.IN3A = 2, D, OR R) INSERT your IF PERSON ASKING ABOUT IS RESPONDENT INSERT TARGETs IF PERSON ASKING ABOUT IS TARGET PERSON INSERT family IF TFAM_COUNT - 1 > 1 IN5A. Now, just stop me when I get to the right category. Was [your/ TARGETS] total (family) income ? PROBE: Your best estimate is fine. (READ LIST. ENTER ONE ONLY) 1 2 3 (INSERT AMT3) to under (INSERT AMT4) (INSERT AMT4) to under (INSERT AMT5) (INSERT AMT5) to under (INSERT AMT6) 48

4 5 6 D R

(INSERT AMT6) to under (INSERT AMT7) (INSERT AMT7) to under (INSERT AMT8) (INSERT AMT8) or more (DO NOT READ) Dont know (DO NOT READ) Refused

New Table 2009 Poverty Guidelines, rounded up to nearest thousand 100% 150% 200% 250% 300% 350% FAMSIZE AMT1 AMT2 AMT3 AMT4 AMT5 AMT6 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
11,000 15,000 19,000 23,000 26,000 30,000 34,000 38,000 41,000 45,000 49,000 52,000 56,000 60,000 64,000 67,000 71,000 17,000 22,000 28,000 34,000 39,000 45,000 50,000 56,000 62,000 67,000 73,000 78,000 84,000 90,000 95,000 101,000 107,000 22,000 30,000 37,000 45,000 52,000 60,000 67,000 75,000 82,000 89,000 97,000 104,000 112,000 119,000 127,000 134,000 142,000 28,000 37,000 46,000 56,000 65,000 74,000 84,000 93,000 102,000 112,000 121,000 130,000 140,000 149,000 158,000 168,000 177,000 33,000 44,000 55,000 67,000 78,000 89,000 100,000 112,000 123,000 134,000 145,000 156,000 168,000 179,000 190,000 201,000 213,000 38,000 51,000 65,000 78,000 91,000 104,000 117,000 130,000 143,000 156,000 169,000 182,000 195,000 209,000 222,000 235,000 248,000

400% AMT7
44,000 59,000 74,000 89,000 104,000 119,000 134,000 149,000 163,000 178,000 193,000 208,000 223,000 238,000 253,000 268,000 283,000

500% AMT8
55,000 73,000 92,000 111,000 129,000 148,000 167,000 186,000 204,000 223,000 242,000 260,000 279,000 298,000 316,000 335,000 354,000

49

NEW QUESTION (ASK EVERYONE) INSERT you IF PERSON ASKING ABOUT IS RESPONDENT and TFAM_COUNT = 1 INSERT TARGET IF TARGET IS NOT RESPONDENT and TFAM_COUNT = 1 INSERT you or your immediate family IF PERSON ASKING ABOUT IS RESPONDENT and TFAM_COUNT > 1 INSERT TARGET or someone in TARGETs immediate family IF PERSON ASKING ABOUT IS TARGET PERSON and TFAM_COUNT > 1 IN5a Did [you/TARGET/you or your immediate family / TARGET or someone in TARGETs immediate family] file an income tax return last year? 1 Yes 2 No D (DO NOT READ) Dont know R (DO NOT READ) Refused

NEW QUESTION (ASK If Q IN5a=2 or D or R ask IN5b) INSERT Have you IF PERSON ASKING ABOUT IS RESPONDENT and TFAM_COUNT = 1 INSERT Has TARGET IF TARGET IS NOT RESPONDENT and TFAM_COUNT = 1 INSERT Have you or anyone in your immediate family IF PERSON ASKING ABOUT IS RESPONDENT and TFAM_COUNT > 1 INSERT Has TARGET or anyone in TARGETs immediate family IF PERSON ASKING ABOUT IS TARGET PERSON and TFAM_COUNT > 1 IN5b [Have you/Has TARGET/ Have you or anyone in your immediate family/ Has TARGET or anyone in TARGETs immediate family] ever filed an income tax return? 1 Yes 2 No D (DO NOT READ) Dont know R (DO NOT READ) Refused IN6. DELETED NEW QUESTION IN7 What Ward do you live in? ______________ (ENTER WARD = 1 8) DD (DO NOT READ) Dont know RR (DO NOT READ) Refused

50

(EVERYONE:) FINAL: Were almost finished now. Do you have any questions for me? IF NEEDED READ For information on health insurance coverage in the District please call the District Ombudsman at 202-724-7491 (INTERVIEWER NOTE: This number can be used for information about both public programs and private coverage.) For questions about the survey, please call Kathy Langdale 1-800-633-1986. To speak with someone from the DC Department of Health Care Finance , please call Jessica Schubel at 202-442-5889.) We would like to thank you for your contribution to this important research. So, the last thing I need is your name and mailing address in order to enter you in the drawing for the $100 prize, and to send you your $20 gift card for either Giant or the Metro, or if you prefer to make a $20 donation to one of the following local charities: Capital Area Community Food Bank, DC Public Education Fund, DC Public Library Foundation, or DC Youth Orchestra]. Which one do you prefer? (IF NEEDED, your name and address will not be given to anyone else and will not be used for any purpose except to mail you your $20 thank you and, if you are the winner of the drawing, the $100.) (IF $20 DONATED SAY: thank you very much for your contribution to [NAME OF CHARITY]. 1 9 1 2 3 4 5 6 Agreed Refused Giant Metro Capital Area Community Food Bank DC Public Education Fund DC Public Library Foundation DC Youth Orchestra

Name: _____________________________ Street: _____________________________ City: __Washington, DC_______________ Zip code: ___________________________

51

(IF respondent gives address for drawing/incentive gift, ask GC.) GC. Is this your home address? 1 Yes 2 No D (DO NOT READ) Dont know R (DO NOT READ) Refused (ASK GC1 if GC = 2, D, R or if dont give us an address for drawing/incentive gift) Wed like to find what general area you live in to get an idea of where the survey respondents reside. GC1. What is your zip code? ____________(RECORD 5 DIGIT NUMBER) RR (DO NOT READ) Refused On what street do you live? ___________(RECORD STREET NAME) R (DO NOT READ) Refused GO TO Q.GC5

GC2.

(ASK GC3 if Street Name given in GC2) GC3. To ensure your confidentiality we do not need your exact address, but are only asking for the nearest 100th. For example if you lived at 127 Main Street you would just tell me you lived on the 100 block. (IF NEEDED: If you lived at 2263 Main Street, you would tell me you lived in the 2200 block) [IF THEY ARE HESITANT REMIND THEM OF CONFIDENTIALITY] ___________ (RECORD BLOCK NUMBER) R (DO NOT READ) Refused [ASK IF GC3=R] GC4. We understand your concern. Well, can you at least tell me what street crosses or meets your street at the nearest intersection? (RECORD VERBATIM) 1 Answer given R Refused [ASK IF GC2=R] GC5. We understand your concern. Well, can you at least give me the names of the two streets that intersect closest to where you live? (RECORD VERBATIM) 1 Answer given R Refused D20. D21. Gender of Interviewer Language of Interview

READ: Those were all of the questions we had for you today. Thank you so much for your time and cooperation. Have a pleasant day/evening.

52