CS-18-1432 Total National and Underserved Survey

Supporting Statement

Approval Request to Conduct Customer Satisfaction (OMB #1545-1432)

Taxpayer Advocate Service (TAS) Total National and Underserved Taxpayers Research

1. JUSTIFICATION

1. Circumstances Making the Collection of Information Necessary

The Taxpayer Advocate Service (TAS) is the Taxpayers’ voice at the IRS. Our job is to ensure that every Taxpayer is treated fairly, and each knows and understands his or her rights and knows where to get tax assistance if it is needed. As an independent organization within the IRS, we help Taxpayers resolve problems with the IRS and recommend changes that will prevent the problems. In order to fulfill our mission, we need insight into Taxpayers’ awareness and use of tax assistance offerings, in particular, how Underserved Taxpayers segment into addressable groups for TAS assistance. This information is not readily available from existing data sources and the research as proposed in this request will help TAS better understand the Underserved Taxpayers and their service and information needs.

The proposed research has an important history within TAS. In 2001-2002, TAS contracted for and conducted benchmark research which identified the size of the Underserved population and segmented them into unique and addressable socio-economic groups so that TAS could reach them with communications. That research showed that the actual TAS Underserved target was much broader than anticipated and included a wide mix of different life situations – ranging from Affluent Families and Empty Nesters on the higher end of the income scale to the Stable Middle Class in the middle to Surviving Spouses, Struggling Young Families and Unmarried Low Income in the lower range. The research also revealed issues requiring clarification/education across all segments, including: TAS qualifying criteria, ability to resolve problems quickly, independence from the IRS, and personal attention provided by TAS case advocates. Based upon those learnings, TAS targeted three of the key Underserved segments specifically where the greatest impact could be made – Surviving Spouses, Struggling Young Families, and Unmarried Low Income.

In 2007, after 4 years of TAS communications to the Underserveds, TAS replicated the 2001-02 research to understand the program’s impact, update its knowledge of the size and shape of its target audience, analyze the effectiveness of its communications, and determine whether further marketing activities were warranted.

2. Purpose and Use of the Information Collection

In 2018, TAS will revisit the Underserved taxpayers in a replication of the earlier research efforts so that – after 10 years without research on the segmentation of this audience – TAS can update its knowledge of the size and shape of the Underserveds target; whether they segment in the same way or not and if not, how they are now shaped; and what they know and expect of TAS and the IRS when they face issues in filing. In addition, TAS wants to use the 2018 research to augment its learning about where the Underserved are living and how this might affect TAS office location decision-making; to learn the extent to which Taxpayers in general and the Underserved have learned about the Taxpayer Bill of Rights; and the role that social media is playing in their lives (this will help modify TAS communications to be more effective in reaching the Underserveds).

The feedback received in these studies will not institute new policy, but will enable TAS to effectively meet Taxpayer needs. Additionally, by using a mixed mode approach for data collection, we will also be able to see how responses vary by data collection methodology. This is extremely important research given that individuals often screen telephone calls and refuse to answer calls from a number they do not recognize, with this trend of non-response increasing dramatically since the previous TAS Underserved taxpayers research – increasing sharply over the past year with the surge in spam telephone calls and scams.

3. Consideration Given to Information Technology

TAS will administer 60 percent of the two studies above by telephone and 40 percent online. In all telephone work, Random Digit Dialing (RDD) will be used and will incorporate cell phone numbers as well as landline telephone numbers in proportion to their current real-world presence (60-65% cell and 35-40% landlines). TAS wants to make sure all pertinent socioeconomic groups and segments of the population are represented. TAS considered other options, but decided telephone surveys are the best option for reaching Taxpayers with lower income and education levels who may not be technologically savvy.

4. Duplication of Information

There is no other research being conducted currently that provides the depth and breadth of the information about the Underserveds that will be secured from this research. In addition, this information is not available from other sources within TAS or the IRS (or elsewhere).

5. Reducing the Burden on Small Entities

Small business or other small entities are not targeted for these surveys and the Taxpayers who are invited to participate will be interviewed only regarding essential information so that burden is limited. In addition, TAS will minimize the burden on Taxpayers by reminding them that participation is voluntary.

6. Consequences of Not Conducting Collection

Without this feedback, TAS will not have timely information from the Underserved Taxpayers that will allow TAS to understand and communicate with these taxpayers in the most appropriate terms and communications channels.

7. Special Circumstances

There are no special circumstances relating to this request. The information collected will be voluntary and will not be used for statistical purposes.

8. Consultations with Persons Outside the Agency

Russell Research, the contractor who conducted both the 2001-02 and 2007 iterations of the Underserved research was consulted for study design and data collection decisions.

9. Payment or Gift

No payments or gifts will be given to participants.

10. Confidentiality

TAS will provide privacy to the extent allowed by law. Personal or sensitive information will not be disclosed. Random digit dialing sampling and online sampling will be used to recruit survey respondents, without TAS providing any lists of Taxpayers or their information. Respondents will never be identified personally, with case ID numbers used instead. In addition, data security and limited use are contractually required of our contractor. All appropriate privacy requirements and background clearances will be obtained and followed. We will also control official access to the information and will not allow public access to the information. TAS will not ask for or maintain any information that could be used to identify the survey respondents.

The data returned to TAS will have no identifying information relating specific records to individual Taxpayers. Information access will be tightly controlled. We will apply fair information and record-keeping practices to ensure protection of all Taxpayers. The criterion for disclosure laid out in the Privacy Act, the Freedom of Information Act, and section 6103 of the Internal Revenue Code, provides for the protection of Taxpayer information as well as its release to authorized recipients.

We will limit and control the amount of information that we collect to those items that are necessary to accomplish the research objectives. We will protect the privacy of the survey respondents to the extent allowed by law by not using names in our report. Data will be maintained according to required record management guidelines.

11. Sensitive Nature

A few questions may be considered sensitive by respondents because they relate to their tax filing behaviors. These questions are used to help establish whether the individual qualifies for the research. We kept sensitive questions to the absolute minimum, and we allow respondents to decline to answer the question if it makes them uncomfortable.

12. Burden of Information Collection

Overall, we expect that 2,000 individuals will complete a survey. For the telephone survey, we expect 1,200 individuals will complete the surveys. Another 800 will complete the survey Online (1,200 + 800 = 2,000).

Burden hours will differ by mode because of the differences in response rates:

• In the Telephone based work involving 1,200 total respondents (600 per study), we will assume that the response rate will be higher than the 9% seen generally now in research (because of the up-to-5-attempts per number) and that it will be approximately 15%. However, the 600 people in the Total Taxpayer Study will qualify at an 80% rate among the 15%, for a net of 12% while the 600 people in the Underserved Taxpayer Study will qualify at about the same level as in 2007, when it was 6% Past-2-Years Underserved, for a net of 1%. The net rate of qualification for two equal size samples is 2x the lowest incidence – so it would be about 2% here.  For screening, the 2% net rate would mean that we need to contact about 60,000 Taxpayers to complete 1,200 surveys. Since screening requires about 2 minutes per contact, we need 60,000 x 2 minutes or 120,000 mins. /60 mins = 2,000 screening burden hours for the telephone surveys.

• In the Online based work involving 800 total respondents (400 per each Quantitative study), there is no standard calculation of response rate.  Instead, respondents enter the survey in representative proportions and then are screened for qualification for the survey.  For the Online portion of the Total Taxpayer Study, the rate of qualification will be about 80% (filed taxes recently) while for the Online portion of the Underserved Taxpayer Study, the rate of qualification will be much lower – likely near/about the same level of Underserveds we found in 2007 (when it was 6% Past-2-Years Underserved).  The net rate of qualification for two equal size samples is 2x the lowest incidence – so it would be about 12% here.  For screening, a 12% response rate means that we would need to screen about 6,667 Taxpayers to complete 800 surveys.  Screening non-qualifieds requires about 2 minutes per contact, resulting in ~223 burden hours for non-respondents. (Calculated as 6,667 x 2 = 13,334 minutes / 60 minutes = 223 burden hours).

 

• Then, for the 2,000 total respondents across both modes who will qualify and participate in one of the surveys, the length of the surveys is about 20 minutes.  This means that there will be (2,000 x 20 = 40,000 minutes / 60 minutes =) 667 burden hours for the 2,000 Quantitative Study participants.

 

• Total Burden for the two Quantitative studies is summarized below:

 

	Estimated Burden
Type of Collection	Number of Cases	Amt Of Time Each	Total Hours
Telephone/Online Survey Screening/Recruiting	66,667	2 minutes	2,223 hrs
Total Ending Sample Respondents Across Both Modes (Telephone + Online)	2,000 Respondents	20 minutes	667 hrs
Total Burden Hours for Telephone/Online surveys			2,890 hrs

 

13.  Costs to Respondents

There is no cost for participants.

14. Costs to Federal Government

The estimated combined cost of both studies in this program is ~$249,000.

15. Reason for Change

No change is being requested. This is a new request for OMB approval

16. Tabulation of Results, Schedule, Analysis Plans

Once OMB approval is received, potential respondents will be contacted for the two surveys. The fieldwork phase of the research is expected to take approximately 2-3 months. In total, it is expected that we will need approximately 5 to 6 months to field, administer, and summarize the learning from both studies (survey studies will be followed-up by focus group interviews). We would like to begin data collection on the two surveys in June 2018, if possible.

This research will consist of 2 different studies:

1. Total Taxpayers Quantitative Study: The overarching purpose of this study will be to determine a new estimate of the Size of the Underserved audience and levels of national Awareness and Usage of TAS.

This would be a mixed mode survey of 1,000 US Taxpayers who have filed a recent Federal tax return, with 600 of them surveyed via a Telephone component and 400 via an Online/Internet component, as follows:

• The Telephone component would replicate the all-telephone Total Taxpayer Studies conducted in 2001-02 and 2007, with survey respondents selected from RDD/random digit dialing. The sample will include Cell phones and Landlines in proportions representative of the general population (now 60-65% Cell and 35-40% Landline).

• The Online component is new and designed to allow TAS to see the impact of the Online mode on its research, as more and more research is moving to the Online platform and as telephone research is proving to be increasingly questionable (because of extremely low, 9%, response rates). Survey respondents in this component will be double-opted-in members of one of the research industry’s premier online panels (SSI/RN), who will be brought to the survey in proportions representative of US Census population data for age, gender, region, and ethnicity. From there, they will be screened for qualification and surveyed.

• We would first conduct all Telephone surveys, assess results and sample characteristics, and then conduct the Online component and, after it is complete, we would compare characteristics from the two components and determine if and how the Online sample needs to be post-weighted to align with the Telephone sample. (Note: TAS will also receive an analysis of how data compared across the two modes.)

Note: in estimating total population figures (e.g., the current size of the Underserved taxpayers target or national awareness of TAS), we would use the Telephone survey results. However, in analyzing key segments, we would use the combined Telephone + Online data in order to increase segment sizes and the stability of analyses of those segments.

2. Underserved Taxpayers Quantitative Study: The purpose of this component of the research program is to explore why TAS assistance was not sought, including whether they are aware of the existence of TAS. Results of this study, together with those of Study #1 above will help determine which marketing channels would make targeted Taxpayers more aware of TAS services.

This would also be a mixed mode survey of 1,000 US Taxpayers who have filed a recent Federal tax return AND are qualified as Past-2-Year Underserveds (i.e., experienced a qualifying tax issue in past 2 years but never used TAS to help with it), with 600 of them surveyed via a Telephone component and 400 via an Online/Internet component, as follows:

• The Telephone component would replicate the all-telephone Underserved Taxpayer Studies conducted in 2001-02 and 2007, with survey respondents selected from RDD/random digit dialing. The sample will include Cells and Landlines in proportions representative of general pop (60-65% Cell, 35-40% Landline).

• Here too, the Online component is new and designed to allow TAS to see the impact of this growing mode on TAS research. Respondents here will also be double-opted-in members of one of the research industry’s premier online panels (SSI/RN), who will be brought to the survey in proportions representative of US Census population data for age, gender, region, and ethnicity and from there, be screened for qualification and surveyed.

• As in the Total Study (#1), we would first conduct all Telephone surveys, assess results and sample characteristics, and then conduct the Online component and, after it is complete, we would compare characteristics from the two components and determine if and how the Online sample needs to be post-weighted to align with the Telephone sample. Note: these are all prior two year Underserveds, so their characteristics will likely differ from those in the Total Taxpayer Study and different weights (if needed) will be developed. (Note: in this study too, TAS will also receive an analysis of how data compared across the two modes.)

Note: in estimating Underserved population figures (e.g., sizes of Underserved segments), we would use the Telephone survey results. However, in analyzing the Underserved segments, we would use the combined Telephone + Online data in order to increase segment sizes and the stability of analyses.

17. Display of OMB Approval Date

IRS is seeking approval to not display the expiration date for OMB approval, as this is a one-time, limited-duration collection.

18. Exceptions to Certification for Paperwork Reduction Act Submissions

These activities comply with the requirements in 5 CFR 1320.9.

19. Data Collection Date

June 2018 through November 2018.

B. STATISTICAL METHODS

The primary purpose of these collections will be for internal management purposes; there are currently no plans to publish the summary reports from this research.

1. Universe and Respondent Selection

The populations of interest in this study vary by research component:

• For the survey of 1,000 US TAXPAYERS, we will interview Individual Taxpayers who have filed a recent tax return – 600 by Telephone, 400 via Online sampling.

• For the survey of 1,000 UNDERSERVED TAXPAYERS, we will interview Individual Taxpayers who in the past 2-3 years have met the TAS qualifying criteria, but have not sought TAS assistance – 600 by Telephone, 400 via Online sampling.

2. Procedures for Collecting Information

All potential respondents in the Telephone components of the research will be telephoned/called by trained telephone interviewers operating from the contractor’s national telephone interviewing center in East Rutherford, NJ. This center operates from 10 a.m. to 12 midnight to allow for 10 a.m. to 9 p.m. calling hours in each US continental time zone. All potential respondents in the Online components of the research will be recruited and allowed into the Online survey in a controlled manner so that their interviewing days and times parallel that of the Telephone surveys, eliminating the risk of differences in data due to differing survey times.

Potential Telephone respondents will be selected from a Random Digit Dialing (RDD) list of computer-generated telephone numbers in working telephone exchanges across the US. The vendor will provide the RDD list. This RDD list will include a representative proportion of cell phones (60-65%) and landlines (35-40%). Cell phone users will be surveyed with different dialing equipment from landline users in compliance with the FCC’s standing regulations regarding calling cell phones vs. landlines. The contractor’s telephone interviewing staff will screen survey candidates and identify potential respondents per screening criteria. The telephone numbers of potential respondents will be dialed up to 5 times in order to establish contact with a potential respondent, with most of these dialings resulting in non-contact (number busy, voice-mail, non-working, etc.). After the first live contact with a potential respondent, interviewers will request their participation in the survey. If they agree, they will be screened for qualification and if qualified, interviewed at that time. If they do not agree/refuse, they will be recontacted by a more senior interviewer who will also try to include them in the survey. If they then refuse again, this person will be recorded in call records as a refusal and not contacted again.

Potential Online respondents will be selected on a random sampling basis from the vendor’s partner panels in Census-rep proportions for age, gender, region, and ethnicity and then screened for the same qualification as the Telephone survey respondents. There are no repeat attempts on Online samples.

Each of the Telephone and Online surveys will include Spanish-speaking respondents, who will be offered the option of being interviewed in either Spanish or English, depending upon their preferences.

3. Methods to Maximize Response

The contractor will strive to maximize response and cooperation rates among Telephone survey respondents who they call by: (1) using up to 5 attempts at live contact; (2) using more senior interviewers to try to convert refusals to completed surveys. Those methods are not available in the Online format. However, other factors increasing response rate for BOTH types of surveys are that: (1) the questionnaire length is being minimized to reduce respondent burden and (2) respondents are assured anonymity of their responses.

4. Testing of Procedures

Because of vendor experience, procedures will not require testing.

5. Contacts for Statistical Aspects and Data Collection

Several contacts will be available for assistance with the statistical requirements. TAS Research will provide analysts with expertise in statistics to assist this effort. The vendor (contractor) will provide experts with knowledge of statistics as well.

Primary Contact: Carol Hatch, TAS Research, 501 396-5954 or carol.m.hatch@irs.gov

Secondary Contact: Jeff Wilson, TAS Research, 317 685-7673 or jeff.a.wilson@irs.gov