CARD Act survey PRA supplement rev 2-2-11

Department of the Treasury, Departmental Offices,
Consumer Financial Protection Bureau Implementation Team

Request for Emergency Processing and Approval

CFPB CARD Act Research

EMERGENCY JUSTIFICATION

The Department of the Treasury (“Treasury”) respectfully requests emergency processing and approval of the collection of information contained in a telephone survey that will be used by the Consumer Financial Protection Bureau (“CFPB”). The collection of information is needed prior to the expiration of the time periods set out in 5 C.F.R. Part 1320 in connection with a conference the CFPB implementation team is sponsoring on the first anniversary of the effective date of the Credit Card Accountability, Responsibility and Disclosure Act (“CARD Act”).

The Dodd-Frank Wall Street Reform and Consumer Protection Act (the “Dodd-Frank Act”), Public Law No. 111-203, Title X, established the Consumer Financial Protection Bureau (“CFPB”). Once fully operational, CFPB will administer, enforce and implement consumer financial protection laws, including the CARD Act.

The bulk of the CARD Act took effect on February 22, 2010. In connection with its first anniversary, the CFPB implementation team is planning a conference to assess, based on the evidence available to date, the impact that the Act has had on consumers and on the card marketplace. The conference will bring together industry representatives, consumer and civil rights groups, government officials, and academics for a fact-based discussion centered on a series of empirical presentations each to be followed by discussants drawn principally from academia and from the Federal Reserve Board. A draft agenda is attached; please note that not every speaker has been confirmed.

In connection with this conference, Treasury’s CFPB implementation team, supported by an outside research firm, has designed a phone survey whose purpose is to assess the extent to which consumers are aware of various of the changes affected by the CARD Act and are utilizing the information made available to them by the CARD Act to affect their behavior. The survey results will be presented at the conference and will contribute to providing a strong empirical grounding for the conference.

The Conference will be held at the Treasury Department on February 22, 2011, which, as noted, is the one-year anniversary of the effective date. Given that timing, Treasury cannot reasonably comply with the normal clearance procedures under 5 C.F.R. Part 1320 with respect to this survey.

Department of the Treasury, Departmental Offices
Consumer Financial Protection Bureau Implementation Team

Request for Emergency Processing and Approval

CFPB Web Survey

1. CIRCUMSTANCES NECESSITATING THE COLLECTION OF INFORMATION
   
   

The CARD Act made substantial changes to the Truth in Lending Act, 15 U.S.C. 1601 et seq. (“TILA”). Among the changes made by the Act were enhanced disclosures in monthly billing statements designed both to increase the transparency to consumers of the cost they are paying for credit and, through a form of “choice architecture,” to guide consumers towards paying down their debt more rapidly to the extent they can afford to do so. The implementing regulations adopted by the Federal Reserve Board also amended some pre-existing disclosure requirements in an effort to assure that the information consumers receiving when applying for a credit card is clearer to them.

Once fully operational, CFPB will be responsible for administering and enforcing TILA and the CARD Act along with a myriad of other consumer financial protection laws. A core principle for CFPB, in exercising its regulatory powers, will be to assure that its decisions are grounded in a deep understanding of the consumer markets over which CFPB will have jurisdiction, including an understanding of how these laws are impacting consumers. The first anniversary of the effective date of the CARD Act provides an appropriate occasion to examine how the credit card marketplace has changed over the past year as well as to reinforce the message contained by the President in his State of the Union address with respect to the important consumer protections brought about by the CARD Act. The proposed survey will further both of these objective.

In particular, the proposed survey will allow the CFPB implementation team to share with participants at the CARD Act Conference data with respect to the extent to which the new information being provided to consumers is actually being absorbed by them, the value they place on that information, and the extent to which they report using the information to change their behavior in ways intended – or, potentially, in ways not intended – by the CARD Act. The survey also will provide valuable insights into how consumers who have applied for a new credit card in the past year went about deciding to do so. This information will complement the other presentations planned for the conference by providing a consumer perspective.

2. USE OF THE DATA
   
   

The aggregate results of the survey (weighted as explained in Part B) – along with cross-tabulations where there are statistically significant differences among relevant sub-populations -- will be presented at the CARD Act Conference. This survey is being conducted to provide formative information for CFPB to share at this conference. Participants in the conference – including representatives of industry and consumer groups -- will have an opportunity to share with the CFPB implementation team and the other government officials in attendance their views as to the significance of these results, further questions the results may raise, and any implications for further examination and research. The results, and the feedback received from the various stakeholders, will contribute to informing CFPB’s views with respect to policy priorities and approaches. The presentation will make clear, orally and on any handout, that this survey is formative in nature.

Aggregate results of the information eventually will be made available as downloadable data. It has not yet been determined whether this will be in a text format (like .csv) or a more instruction-heavy format (like XML).

3. USE OF INFORMATION TECHNOLOGY
   
   

To assure results that are representative of the population as a whole, the survey will be administered via phone rather than using the Internet. Web-based research by definition excludes from the sample those without Internet access. In addition, web-based research is typically limited to an opt-in panel and the results cannot be reliably extrapolated to the population as a whole. Participating in the survey will be entirely voluntary and will not take more than 15 minutes on average.

4. EFFORTS TO IDENTIFY DUPLICATION
   
   

This will be the CFPB implementation team’s first attempt to collect the information sought in this survey. We have identified and obtained results from four surveys with respect to the CARD Act that have been conducted either by advocacy groups or research firms over the past year. For the most part, the results of those surveys are dated, as they were conducted shortly after the effective date of the Act. Additionally, some of those surveys were administered to a non-representative sample while others did not ask questions relevant to assessing the consumer perspective on the law’s impact. Accordingly, the only way to obtain the information we seek is by conducting this survey. We will, however, utilize certain questions asked in prior surveys in order to provide some points of comparison as well as to benchmark the results obtained.

5. IMPACT ON SMALL ENTITIES
   
   

Not applicable. The reported results of the survey will have no direct impact on any small entity. The survey is designed to elicit responses from individuals, not entities. It is assumed that individuals will not answer during time for which they are being compensated, or are responsible to or for someone else, to any extent that it would adversely affect another party.

6. CONSEQUENCES OF LESS FREQUENT COLLECTION AND OBSTACLES TO BURDEN REDUCTION
   
   

If this information is not collected, it will be not be possible to present to the CARD Act conference current data reflecting a consumer perspective with respect to the impact of the Act.

The frequency of collection is not applicable because the survey is a one-time event and participation is entirely voluntary. There is no need to collect information from an individual more than once.

7. CIRCUMSTANCES REQUIRING SPECIAL INFORMATION COLLECTION
   
   

Not applicable. The methodology to be utilized will allow the results to be generalized to the universe of the study.

8. SOLICITATION OF COMMENTS ON INFORMATION COLLECTION
   
   

The CFPB implementation team developed the initial draft of the questionnaire to be used in the study. The survey instrument was reviewed by Dr. Jeanne Hogarth, Program Manager of the Consumer Education and Research Section, Division of Consumer and Community Affairs, at the Federal Reserve Board; and by Dr. Patricia Sahm, Research Director, Auriemma Consulting Group. Following a competitive bidding process, Synovate has been retained to finalize the questionnaire, develop the research methodology to be utilized, administer the survey, and tabulate and present the results. Synovate is the world’s seventh largest research firm, with a more than fifty-year history in market research and a long track record of preparing studies of this type both for governmental agencies and for private sector clients. Synovate’s Demonstration of Expertise is attached to this document.

9. PROVISION OF PAYMENTS TO RESPONDENTS
   
   

The information collection does not provide for making payments or gifts to respondents.

10. ASSURANCE OF CONFIDENTIALITY
    
    

The survey will not capture names, email addresses or other personally identifiable information (PII). Respondents will be assured that their answers will not be attributed to them individually.

11. JUSTIFICATION OF SENSITIVE QUESTIONS
    
    

The most sensitive information on the survey is demographic data about the respondent, such as age range, race, income, and education level and card usage data such as current debt level. This will be collected anonymously and voluntarily. The request is necessary to appropriately balance the sample and also to explore whether different sub-populations report different experiences under the CARD Act. Because no PII is being collected, none of this demographic or card usage information will be tied to any individual.

12. ESTIMATED BURDEN OF INFORMATION COLLECTION
    
    

The full survey will be administered to 800 individuals all of whom indicate that they currently use a accredit card. To obtain 800 completed interviews it is estimated that the initial screening questions to establish eligibility (i.e. current credit card ownership) will be asked of another 350 individuals; ineligible respondents will be asked a few demographic questions as well for weighting purposes between land line and cell phone only segments. The survey will take no more than 15 minutes on average to complete; this will be validated by administering the survey to some employees of Synovate who are not involved in the project before calling begins.

13. ESTIMATED TOTAL ANNUAL COST BURDEN TO RESPONDENTS
    
    

There will be no annualized capital or start-up costs for the respondents to collect and submit this information.

14. ESTIMATED COST TO THE FEDERAL GOVERNMENT
    
    

The cost of the project, including both the administration of the survey and the preparation and presentation of a report to the CARD Act conference, will be $58,879.87. That price has been arrived at through a competitive bidding process.

15. REASONS FOR CHANGE IN BURDEN
    
    

Not applicable. This is a new collection.

16. PLANS FOR TABULATION, STATISTICAL ANALYSIS, AND PUBLICATION
    
    

The responses to each question will be tabulated on an aggregate basis with appropriate weightings as explained in Part B. In addition, cross-tabulations will be prepared for certain key questions showing the results across different age ranges, income ranges, educational levels, debt levels, and payment practices. Standard statistical techniques will be used to highlight variations that are of statistical significance at the 95% confidence level.

Aggregate results of the information eventually will be made available as visible and/or downloadable data. It has not yet been determined whether this will be in a text format (like .csv) or a more instruction-heavy format (like XML). The analysis will be descriptive in nature, including graphical representations, comparisons of different totals for different answers to a question, contextualizing with outside data not gathered from the survey, and other similar descriptions. Publication of information will be subject to our ability to draw findings that would be of interest to a general public audience. Results will be presented at the CARD Act Conference on February 22, 2011.

The survey will be conducted between February 3^rd and February 7^th, 2011.

17. REASONS WHY DISPLAYING THE OMB EXPIRATION DATE IS INAPPROPRIATE
    
    

Because the survey will be conducted by phone, the OMB control number and the OMB expiration date cannot be visibly displayed. Synovate will be notified of and required to adhere to the OMB expiration date. The OMB control number and expiration date will be provided to the interviewers and will be made available to a respondent on request.

18. EXCEPTIONS TO CERTIFICATION REQUIREMENT OF OMB FORM 83-I
    
    

There are no exceptions to the certification statement in item 19 of Form 83-I.

B. COLLECTIONS OF INFORMATION EMPLOYING STATISTICAL METHODS

1, The universe from which the sample for this research will be drawn consists of every adult over the age of 18. Participants in the survey will fall within a subgroup consisting of adults with a credit card and a phone number. It is estimated that approximately 70% of adults, or approximately 160,000,000 Americans, fall within the universe.

Synovate will select random samples of households from the universe using the GENESYS Sampling System developing by Marketing Systems Group to generate random digital dialing lists. Two lists will be procured, one for cell phones and one for land lines. These lists will then be stratified into the four census regions and, within each region, into a cell phone-only segment and a landline segment using the most current data from the National Health Interview Survey; the incidence of credit card ownership within each sub-group is unknown and may vary from the incidence in the population as a whole. Appendix A shows the universe and sample targets for each strata.

Synovate expects a response rate in the 15% to 20% range.

2. Synovate will conduct the calling using the RDD lists procured from GENESYS. Individuals willing to participate will be asked a few screening questions to determine eligibility (i.e. whether the individual has a credit card). Individuals reached on a cell phone but who indicate that they have a landline will be excluded from the survey so that these segments can be balanced to the cell phone-only universe in each region.

Given cost constraints and the relative cost difference in completing a cell versus landline response, Synovate’s protocol is to obtain 66% of the target cell phone only segment, with a floor of 50%. To compensate, Synovate will, as necessary, weight the responses in the cell phone only segments so that it is proportional to their representation in the regional population as a whole.

Synovate will collect basic demographic information from all “willing to participates,” including those who are not eligible because of lack of credit card ownership. Synovate will weight the willing to participates to assure that the willing to participates are representative of the universe with respect to age, income, gender, and race. This balancing will be done on a global basis rather than region by region.

Results will be reported based on the overall sample and not for each of the strata. Cross-tabulations will also be reported for subgroups defined by key demographic variables (e.g. income, education, age) and defined by credit card behavior (e.g. heavy borrowers, light borrowers, convenience users). The size of each sub-group will not be known until the survey is completed. To the extent comparisons are made between subgroups (e.g. heavy borrowers vs. light borrowers), Synovate will first verify that the differences being reported are statistically significant with 95% confidence.

Appendix B provides a detailed explanation of the weighting methodology and the margin of error.

3. To maximize response rates within the time available, calling will begin on a weekday (Thursday) evening and continue through the weekend on both day and evening and conclude Monday evening. Synovate will make four attempts to contact each called number. Where a contact is made, Synovate will ask to speak with the youngest adult male who is at home 75% of the time and the youngest adult female who is at home 25% of the time. While this methodology admittedly is not a random method for selecting an adult in the household, it is known to produce more balanced samples with better representation of the young and males. As a result, it is a methodology applied often for quick turnaround polls (including, e.g., the ABC News/Washington Post poll for which Dr. Lambert managed the sampling and data collection contract prior to joining Synovate). Papers and articles based on these results have been presented at AAPOR conferences and also published in peer-reviewed journals such as the Public Opinion Quarterly).   Use of a methodology such as a birthday method to select respondents, especially in quick turnaround polls, tends to produce samples in which women are overrepresented and younger adults are underrepresented.  Given the time constraints and the formative nature of this research, we believe this is the preferred methodology for this poll.

The questionnaire will be limited to approximately fifteen minutes in length on average.

4. No formal testing is planned. The questions have been reviewed by CFPB staff, by two outside consultants, and by researchers at Synovate. Before commencing the call, Synovate will administer the survey instrument to several employees who are not engaged in the project to test the flow of the survey, evaluate the clarity of the questions, and to validate that the survey is within the desired length.

5. The statistical aspects of the design have been developed by Synovate. The key project staff with respect to statistical design are:

Dr. Alan Roshwalb, Senior Vice President and Senior Statistician, Synovate, (703) 663-7260

Dr. David Lambert, Senior Vice President, Synovate Public Sector Research Group, (610) 897-2059

APPENDIX A: Survey Universe

US Adults (2010 March CPS)	US Adults	% of Region Cell-only	US Cell-only Adults	US Landline Adults
Region 1: Northeast	41,783,671	15.8%	6,599,241	35,168,109
Region 2: Midwest	49,546,291	26.6%	13,174,026	36,352,387
			-	-
Region 3: South	82,914,147	29.3%	24,309,482	58,658,034
			-	-
Region 4: West	52,629,057	23.5%	12,363,793	40,248,094
TOTAL	226,873,166		56,446,543	170,426,623

Synovate Sample Targets

Sample Screens for Cell and Landline by Region (Cell-only at 2/3 proportion)	US Cell-Only Adults	US Landline Adults	Totals
Region 1: Northeast	22	188	210
Region 2: Midwest	45	205	250
Region 3: South	82	336	418
Region 4: West	42	223	265
TOTAL	191	952	1143

APPENDIX B: Weighting in the Consumer Poll

Summary:

The survey protocol requires multiple attempts to reach every phone number in the survey, calling back unanswered numbers on different days at different times of both day and evening across the field period from Thursday through Monday. Respondents are selected in the landline sample using a youngest male rule. Although the respondent selection rule within landline households is a departure from a probabilistic selection process, weights are constructed to account for contacting the household using the number of telephone lines into the residence. Cell phone-only household respondents are weighted according to the number of adults in a region reachable only by cell phone. In addition, the weights are adjusted to account for variation in the sample relating to geographic region, sex, race, age and education.

This note describes the weighting process we will use in the CFPB Consumer Research Poll. The sample design splits the population into landline only households, dual use households (landline and cell phones) and cell-phone only households. Estimates are available for the number of people associations with these groups. For this study, the landline and dual use households will be combined. The steps associated with constructing weights are:

1. Use all screened completes

Create an initial design weight based on Population counts for Landline & Dual Use households and population counts for cell only (see next table).

2. Use raking-ratio, also known as rim weighting or sample balancing to adjust design weights to meet demographic targets for:

• Gender,

• Educational attainment,

• Race/ethnicity, and

• Age.

These adjustments will allow the final sample weights to account for differential cooperation, response, and contact rates that may occur among the different demographic groups.

Weight Construction:

The stratification is reflected in the next table. There are eight strata across the 4 regions and split by land-line/ Dual Use Area Codes and Exchanges versus the Cell-designated area codes and exchanges.

Table: Stratification structure for Cell-Only Screened Design

Landline and dual-use Area Codes and Exchanges stratified over 9 Census Divisions					Cell-designated Area Codes and Exchanges stratified over 4 Census Regions Design
Strata	1	2	3	4	5	6	7	8
	Northeast	Midwest	South	West	Northeast	Mid-west	South	West

The first step is to create a design weight (noted as d_j). The design weight needs to be separated into those for the first four strata versus the last four strata. Telephone numbers in the first four strata are attached to households, while the telephone numbers in the last four strata are attached to individuals. The design weights for the first four strata will take into account the number of telephone lines in the household.

By definition, the design weight is inversely proportional to the chance that a person is selected. The design weight (noted as d_j) will take into account the number of adults living in the household. The equation is basically the same as before, but we have explicitly stated that n = n_a + n_ab = number of landline-only household responses plus the number of dual user household respondents. The formula is:

d_j = DESWT _j = K / (Number of Lines in Household)

.

The next step is to account for the probability of selecting the telephone number. The probability of selection of any telephone number within a stratum is 1 over the number of telephone numbers in the stratum, and it needs to project to the number of adults found in the strata. The formula includes estimates for the N_a,h – the number of adults in landline-only households in stratum h and the N_ab,h -- the number of adults in dual-use households in stratum h are available from the NHIS. These estimates from NHIS are reasonably robust at the region level.

The weights d_j are adjusted such that the weighted sample distribution matches the distribution given in Table 5. The adjustment to the weight has the formula

K_h .

Here, M_h is the number of telephone numbers in the stratum, N_a is the estimated number of adults landline-only households in the U.S., and N_ab is the estimated number of adults in dual use households in the U.S., respectively. Notice that M_h drops out algebraically, so it is not included in the final adjustment. d_j is adjusted by the K_h associated with the response’s stratum. The design weights are normalized so the design weights sum to N_a + N_ab.

The design weights for the last four (i.e., the cell phone-only) strata are based on the number of adults in cell phone-households in the region. The weighting reflects that the protocol screens out dual-use households. The design weight is

d_j = DESWT _j .

Here, N_b is the estimated number of adults cell phone-only households in the U.S, , n_b is the total number selected in the cell-designated strata (strata 5 to 8), and the design weights are normalized so the design weights sum to N_b.

At this stage, the landline-only and dual-use strata weights are projecting to the landline-only and dual-use household totals, N_a + N_ab. The cell phone-only strata weights are projecting to the cell phone-only household totals, N_ab. The combined sample projects to the total population.

Once the design weights are completed, the final adjustment to account for non-response and differential response rates can be made using a raking-ratio adjustment. The adjustments are made on all data, from both the landline-only dual-use strata and the cell phone-only strata, combined. The raking-ratio process adjusts the design weights such the adjusted weights meet a given set of marginal distributions. The process goes by several names such as raking-ratio adjustment, sample balancing, rim weighting, or iterative proportional fitting to known marginal distributions. It is a convenient method because you do not need the joint distribution of the demographic variables. As such, we will use the latest Census Bureau statistics for age, gender, educational attainment, and race/ethnicity.

Margin of error

Number of Groups	Sample Size	Margin of Error (95%)
1	800	3.80
2	400	5.37
3	267	6.57
4	200	7.59

Two Group Comparison
Minimum Detectable Effect Sizes
80% Power and α = .05
One Group's Proportion	Effect Size
10%	6.73%
20%	8.48%
30%	9.42%
40%	9.81%
50%	9.84%
60%	9.44%
70%	8.63%
80%	7.20%
90%	5.10%

APPENDIX C: TENTATIVE CONFERENCE AGENDA

THE CARD ACT: ONE YEAR LATER

8:30 Coffee

9:00 Welcome and Keynote Address

9:30 – 10:45 The Supply of Credit in the Card Market

Moderator: Marla Blow, CFPB

Presenters: Andrew Davidson, Mintel Comperemedia

Michele Ranieri, Experian

Discussants: Prof. Angela Littwin, University of Texas School of Law

Glenn Canner, FRB

11:00 – 12:15 The Demand for and Price of Credit in the Card Market

Moderator: Robert Hunt, Philadelphia Federal Reserve Board

Presenters: Leonard Laufer, Argus Research and Advisory Services*

Ezra Becker, TransUnion

Discussants: Moshe Orenbuch, Credit Suisse

Dr. Larence Ausubel, University of Maryland

12:15 – 1:00 Lunch

1:00 – 2:15 Changes in Consumer Card Behavior

Moderator: Kathleen Johnson, FRB

Presenters: Dennis Campbell, Harvard Business School*

Synovate*

Discussants: Dr. Jeanne Hogarth, FRB

Dr. Patricia Sahm, Auriemma Consulting Group

2:20 – 3:30 Implications for Public Policy

Moderator: David Silberman, CFPB

Panelists: Oliver Ireland, Morrison & Foerster

Josh Franks, Center for Responsible Lending

Professor Katherine Porter, University of Iowa School of Law

David Watson, Merrick Bank*

APPENDIX D – SYNOVATE DEMONSTRATION OF EXPERTISE

Why Synovate

Synovate is a full-service survey and marketing research firm. Established in 1946 as Market Facts, Synovate is one of the longest-tenured companies in the survey and market research industry. Synovate has the knowledge, experience, and resources that will drive success on these surveys. Our firm is an outstanding choice for several reasons:

• Synovate has extensive experience with studies polling and collecting information about attitudes, awareness, opinions and experience with current events, services provided and testing materials from advertising, new services, and general knowledge.

• Our Public Sector Research Group and Financial Services Group together provide the requisite understanding of the CARD Act and its context in consumer finance.

• Synovate has the personnel, resources, and experience necessary to administer complex telephone polls over a short period of time, including experienced programming, field management and data processing staff, a telephone center in Fresno, CA, with more than 100 CATI stations, and protocols and procedures to ensure high quality data collection.

• We also have extensive resources and expertise in conducting online studies, completing over 2.5 million online surveys annually for a wide range of public and private sector clients.

• As the world’s 7^th largest survey research firm, we will be able to devote the necessary resources to conduct the survey.

Demonstration of Expertise

Synovate is a large supplier of research services to the Federal Government. We have conducted or currently conduct studies for numerous Federal agencies including:

The Food and Drug Administration The U.S. Mint The Social Security Administration The Federal Trade Commission National Archives and Records Administration The Coast Guard

The U.S. Postal Service The Department of Defense The Department of the Army The Department of Labor The Department of Veterans’ Affairs The Consumer Product Safety Commission

Many of our studies are custom designed research programs measuring opinions on services provided by the agency. Other studies are general population research projects measuring attitudes, awareness, knowledge and opinions of the U.S. population. These studies are consumer research polls, and they have been conducted for the Food and Drug Administration, the Federal Trade Commission and the Consumer Product Safety Commission.

Synovate has recently conducted general population consumer polls for commercial clients such as U.S. News & World Report. In addition, Synovate sometimes sponsors poll questions on our omnibus vehicles.

Synovate is currently under contract to conduct two modified studies for the Food and Drug Administration – their Food Safety Study and their Health and Diet Survey. These are recurring studies that Synovate has conducted in the past. In addition to data collection, these studies involve complex tests of survey methodology. The projects are in progress with expected fielding in the later part of 2011.

Synovate and Synovate Consumer Opinion Panels comply with the accepted standards for the

marketing research industry and abide by the rules and guidelines set forth by the leading industry organizations:

• European Society for Market Research (ESOMAR)

• The Council for American Survey Research Organizations (CASRO)

• The Market Research Association in the US (MRA)

• Interactive Marketing Research Organization (IMRO)

• The Advertising Research Foundation (ARF)

• The American Marketing Association (AMA)

Key Project Staff

Alan Roshwalb, Ph.D

Senior Statistician and Senior Vice President.

Dr. Roshwalb serves as a senior consultant in the design and execution of sophisticated survey projects. Dr. Roshwalb received his Ph.D. from the University of Michigan’ s School of Business Administration in Statistics and Management Science with an affiliation with the Institute for Social Research. An expert in statistical sampling, he is the principal author of a sophisticated transactional statistical sampling management system currently employed in many key customer satisfaction projects tracking up to 20,000 sampling units with up to 1,000,000 completed interviews per year. Dr. Roshwalb has served as consulting sampling statistician for projects for a variety of government agencies including the Federal Trade Commission, the Social Security Administration, the Food and Drug Administration, National Institutes of Health, the Centers for Disease Control, the Department of Defense, and the U.S. Postal Service. Commercial clients include or have included Coca-Cola, Seagrams, Campbell Soups, American Express, Visa International, ExxonMobil, Kaiser Permanente, CIGNA, IBM, and Capital One among others. These projects have included using in-person door-to-door, telephone, interactive voice response (IVR), internet and mail methodologies. Dr. Roshwalb has authored professional papers on statistical and survey research methodology for social science, marketing, accounting, and management periodicals such as The Accounting Review and The Journal of Business and Economic Statistics. Prior to joining Synovate, he served as Chief Statistician for Audits & Survey Worldwide and on the faculty at Georgetown University’s McDonough School of Business.

David L. Lambert, Ph.D.

Senior Vice President, Synovate Public Sector Research Group

As the Project Executive, Dr. Lambert will oversee the project, establishing the schedule, providing methodological direction and interacting with the Program Staff at the Consumer Financial Protection Bureau. Dr. Lambert is a Senior Vice President with Synovate. Dr. Lambert brings more than 30 years of research experience to the team. For much of his career, he has worked closely with ABC News and The Washington Post conducting their political polls. With his guidance, ABC and The Washington Post have been able to establish sampling and data collection strategies for hard-to-reach populations including “cell only” respondents who are a source of non-coverage in traditional, landline RDD frames. He is well versed in all aspects of survey research including study design, sample design and stratification, sample management, survey instrument development, data collection, weighting and analysis, and reporting. Dr. Lambert has also lent his expertise to a wide variety of non-polling clients and projects, including the American Medical Association, Harvard University, the West Virginia Health Care Authority, the U.S. Army, Independence Blue Cross, Philadelphia Health Management Corporation, the State of Pennsylvania, AARP, and Osaka University