SUPPORTING STATEMENT FOR OMB CLEARANCE
A. Justification
This request is for clearance of data collections for the 2008 and 2009 Insurance Component of the Medical Expenditure Panel Survey (MEPS-IC).
The MEPS-IC is an annual survey of employers to collect information on their health insurance offerings. The sample of employers is derived from a nationally representative sample of employers developed from a list frame. The MEPS-IC first collected information in 1997 for the 1996 calendar year.
This MEPS-IC clearance request is different than those for previous years because it only covers the collection of the above-named list frame sample.
In previous collection years, as part of data collection in AHRQ’s MEPS-Household Component program (OMB Clearance #0935-0118), respondents with employer-sponsored health insurance were identified. A sample consisting of the employers of those respondents was linked to collection of the MEPS-IC. The same information was collected from these linked employers as those selected in the list frame sample, with the goal of providing a unique source of information containing both household and employer information on employer-sponsored health insurance. This linked sample component is not part of the request for clearance.
1. Circumstances Requiring Data Collection
Employer-sponsored health insurance is the source of coverage for approximately 80 million workers and retirees and their families. This insurance costs employers and employees over $500 billion annually.
The MEPS-IC list sample and its predecessors measured the extent, cost, and coverage of employment-based health insurance. The surveys were designed to provide State and national estimates of health insurance spending for the National Health Accounts (NHA) and Gross Domestic Product (GDP); data for evaluating the effects of National and State health care reforms; descriptive data on the current employment-based health insurance system; and detailed data for modeling the differential impacts of proposed health policy initiatives.
There is great need for employer-sponsored health insurance data. The MEPS-IC, along with the National Compensation Survey conducted by the Bureau of Labor Statistics (BLS) which focuses on other questions related to employer-sponsored health insurance and other employer benefits, are the two nationally representative government surveys of employer-sponsored health insurance.
A paper published in Health Care Financing Review (Spring 2002) – Employment-Related Health Insurance: Federal Agencies’ Roles in Meeting Data Needs (Wiatrowski; Harvey; Levit) -- describes in detail the two surveys and the efforts of the OMB-established Interagency Committee on Employment-Related Health Insurance Surveys to coordinate the work of the two programs. (See Attachment A.)
Aside from national level information required by Federal Agencies, State governments and private sector employers and nonprofits also need state level information to assess policies of government at the state level and to help make decisions on health insurance coverage. One of the key design features of the MEPS-IC is its ability to provide State-level and sub-State-level, such as metropolitan area, estimates. Currently, the MEPS-IC is the only survey that provides estimates for all States on an ongoing basis.
AHRQ has received requests for information from diverse groups such as Chambers of Commerce, AFL-CIO, Blue Cross/Blue Shield, and the Robert Wood Johnson Foundation. Tables produced and placed on the Internet are also ideal for the business person who wishes to compare their costs and policies with those of others in the same State or industry.
The rising cost of health insurance, concern about the number of uninsured, and the related proposals by the President now being debated in Congress all increase the demand for employer related health insurance information. The information is vital for good policy making, monitoring and estimation of possible costs of new policies.
The legislation authorizing collection of this information is contained in Title 42, USC, Section 299b-2. (See Attachment B.)
2. Purpose and Use of Information
The primary objective of the MEPS-IC is to collect information on private health insurance available to American workers. Such information is needed in order to provide the tools for Federal, State, and academic researchers to evaluate current and proposed health policies and to support the production of important statistical measures for other Federal agencies.
The MEPS-IC provides annual national and state estimates of aggregate spending on employer-sponsored health insurance (annual premium expenditures) for the National Health Accounts (NHA) that are maintained by the Center for Medicare and Medicaid Services (CMS) and for the Gross Domestic Product (GDP) produced by the Bureau of Economic Analyses (BEA). Both these agencies have been primary users of the MEPS-IC since it inception and have had a part in producing essential changes in the MEPS-IC questionnaire that have allowed AHRQ to make better estimates of expenditures to support their work.
In addition to the regular use of the MEPS-IC data to support production of National Health Accounts by CMS and Gross Domestic Product by BEA, other regular users of previous MEPS-IC data have been:
Department of Treasury;
The House Joint Committee on Taxation;
Department of Labor;
Department of Health and Human Services, including AHRQ & CMS;
General Accounting Office (GAO);
Congressional Budget Office (CBO);
Congressional Research Service (CRS);
Universities; private consulting firms and policy groups;
Government agencies from almost every State.
MEPS-IC staff assists users in both gathering and interpreting published statistics, and also frequently provides special estimates on very quick turnaround for users. Many of these special estimates, while requested through third parties, ultimately inform analyses by the White House or governors’ offices.
Specifically, the MEPS-IC estimates have been used in the following ways:
For large scale analysis of employer health insurance. (GAO)
To answer numerous questions on proposed tax changes set forth by Congress. (Treasury)
To determine how different types employers and employee react to the insurance markets.
To estimate the costs and tax consequences of potential new laws, modeling choices and decisions made by employees concerning their health insurance. (Taxation committee, Treasury, CBO)
For modeling the impact of proposed changes in the tax treatment of employment-related health insurance and potential effects of health care reform initiatives. (CMS, CBO, AHRQ)
To study trends in the supply of health insurance regarding its cost, characteristics and benefit provisions such as: 1) national and state trends in the availability and characteristics of private employer-sponsored health insurance, including shifts from fully-insured to self-insured plans; 2) the impact of national and state health care reform on the availability and cost of coverage and type of plans offered; 3) shifts from fee-for-service to managed care plans; 4) changes in premiums and employer and employee contributions; and 5) changes in enrollment by plan type.
For measuring the annual supply of health insurance provided by the employment-based health insurance system at the national and state levels. These data are needed by health care policymakers for evaluating the health care financing system and for planning purposes.
For identifying the effects of selected state regulatory initiatives in health care markets and to estimate possible costs of new State programs which use employer sponsored health insurance to provide insurance coverage to individuals through State subsidy programs. (various State governments)
To provide data to analyze health care policy issues relating to the supply of private health insurance and the effects of health insurance market regulation such as: 1) What establishment and firm characteristics are correlated with the availability and generosity of health insurance provided to workers? 2) How effective is the health insurance market in pooling risks? and 3) What factors are associated with a firm's decision to self-insure?
The MEPS-IC questionnaire has been updated over the years to improve collection of data and to include new data that users, such as CMS and BEA, require. Major changes since the initial 1996 survey year have included:
use of company-level forms for cases where data collection at the establishment level was not feasible (1997),
expansion of retiree health insurance questions at the establishment level (2000),
conversion of industry categories from SIC to NAICS codes (2000),
collection of costs and enrollments for two-person family coverage rather than just for families of four (2001),
collection of additional information on prescription drug coverage and copays (2003, revised in 2005), and
inclusion of questions on Health Savings Accounts (HSAs) and Health Reimbursement Arrangements (HRAs) (2004, revised and expanded in 2005 & 2006).
These changes help improve estimates of total expenditures, help predict tax income more accurately, and show the response of business to recent health insurance trends.
In addition, minor changes to wording, layout and instructions have been made in response to comments from respondents and data collection staff in order to improve and clarify the questionnaires.
The following questions are different in the current forms compared to those submitted with the last clearance package. Actual question numbers vary for different versions of the form within a year and from year-to-year:
Prescreener forms - No changes
Establishment forms - No changes
Plan forms (listed in order by section) -
1) General Plan Information section: Dropped "Name of Insurance Carrier" and "Group Purchasing Arrangement (MEWA)" questions.
3) Health Savings Account (HSA) section: Added section containing three questions.
4) Payments section: Simplified "Outpatient Drug" question by asking for out-of-pocket payment for only one tier.
5) Health Reimbursement Arrangement (HRA) section: Added section containing one question.
6) Plan Characteristics section: Dropped seven of the ten "Services Covered" categories.
New questions are pretested by the Census Bureau before inclusion on the form. Testing is conducted under the Census Bureau’s general testing clearance plan and burden levels.
To avoid additional burden on individual respondents, increases in questionnaire length are kept to a minimum. Recently, in this regard, questions about Medical Savings Accounts, the name of the insurance carrier, and several questions about specific benefits covered by the health insurance plans were dropped from the survey. These questions were selected for exclusion based on the lack of usefulness and quality of their data relative to other needed questions.
A MEPS-IC Users’ Conference held in November 2005 and frequent meetings with State data users as part of the Health Resource and Services Administration State Planning Grants program have informed many of the recent changes to the survey. As part of the MEPS-IC ongoing efforts to respond to this user community, an important improvement will be made coincident with this clearance. This change will be that the MEPS-IC data reference year will now match the MEPS-IC collection year.
For the 1997 through 2007 data collection years, the MEPS-IC was collected on a one-year, retrospective basis – that is, the data collected during 2007 pertained to employers’ health insurance coverage in reference year 2006. Beginning with the 2008 data collection year, the MEPS-IC will be collecting information pertaining to employers’ current year health insurance coverage.
This change to current year collection is supported by Census Bureau research, through contact with respondents, which shows that more accurate information on health insurance plans can be obtained for the current year than a former year.
Matching the collection year to the reference year has been advocated by numerous users. The reason for the popularity of the change is that the published estimates will be much more current. The private sector estimates will be released no more than eight months after the reference date, instead of about 20 months previously. Government statistics will be published slightly later than private – about 11 months after the reference period – but still ahead of the 20-month lag. In both cases, more current MEPS-IC data means that users’ own estimates and analyses based on the survey will be more accurate and timely.
Operational issues about the switch to matching the collection and reference year have been tested and resolved. One benefit from a collection perspective is that it will reduce the burden on MEPS-IC respondents, whose current year records related to health insurance coverage are much more accessible than those for the prior year.
This change will mean that there will be no MEPS-IC data collected pertaining to employers’ 2007 health insurance coverage (which under the former plan would have been collected in 2008). This will require interpolation of information by users for purposes such as the National Health Accounts and the Gross Domestic Product for 2007, but result in much more timely and useful estimates in 2008 going forward.
3. Use of Improved Technology
CATI is used for prescreening of respondents and telephone follow-up.
4. Efforts to Identify Duplication
There is no survey or study that has been conducted or is currently underway that will meet the objectives of the MEPS-IC. Many federal household surveys, including the MEPS-HC, collect insurance information from households who cannot provide much of the important information provided by the MEPS-IC. Data on employer premiums or about enrollments and offerings by industry are only available through employers.
There are private surveys that produce some of the information collected in the MEPS-IC. Due to the proprietary nature of the surveys and possible sample and universe restrictions on these surveys it is not possible currently to fully compare these with the MEPS-IC. However, no private survey can provide the necessary State level estimates provided by the MEPS-IC.
The Bureau of Labor Statistics’ (BLS) National Compensation Survey (NCS) collects a subset of information similar to the MEPS-IC. Although the two surveys can be used to produce certain overlapping National estimates, the two surveys have very different purposes and samples. See the aforementioned paper in the Health Care Financing Review for an analysis that found that the two surveys each fill special and necessary ‘niches’ in the collection of employer-related health insurance information.
The NCS is designed to produce estimates by occupation while the MEPS-IC is not. To support the collection of occupation information, the NCS must conduct personal visits to implement its occupation sampling processes. Because of this need for personal visits, the survey uses a cluster sampling approach. The cluster sample used does not allow State level estimates and places emphasis on occupational level rather than establishment level data.
The MEPS-IC does not require cluster sampling and can use a sample design which allows for efficient estimation at the State and sub-State level. No occupation sampling is needed within an establishment. These factors allow the MEPS-IC to use less expensive mail and telephone collection.
Another important difference between the surveys is that the NCS offers information about many other benefits beyond health insurance, such as pensions, vacation time, disability insurance, etc. Because of the scope of benefits covered, limited data on each benefit are collected. In contrast, the MEPS-IC only asks for health insurance data, but requests much more detail on coverage and plan specifics than does NCS.
In summary, the MEPS-IC publishes detailed information on health insurance at the State and sub-State level and by establishment characteristics such as industry or firm size, while the NCS provides National estimates for a variety of benefits on an occupational basis.
The Interagency Committee on Employer-Related Health Insurance Surveys (1997-2002) was charged with exploring the similarities and differences between the MEPS-IC and NCS, and making recommendations for changes to the surveys based on its findings. This committee -- which was comprised of staff from OMB, AHRQ, BLS, and other stakeholder agencies -- recommended that the two separate surveys continue due to the reasons discussed above.
Under the terms of clearance (see Attachment S) for its last submission (expiring 04/28/2008), the MEPS-IC was asked to report on recent coordination efforts between the MEPS-IC and the NCS. These coordination efforts have included:
a MEPS-IC sponsored forum of health insurance data users including representatives from the National Compensation Survey and other stakeholders such as the Centers for Medicare and Medicaid Services, the Assistant Secretary for Programming and Evaluation (HHS), the Census Bureau, and government agencies of 15 different States.
a joint NCS/MEPS-IC presentation given at a World at Work conference titled, “Federal Surveys of Employer-Based Benefits.”
a collaboration with NCS staff about the collection of health insurance plan benefits and the ability to abstract benefits from health insurance booklets.
periodic information sharing between NCS and MEPS-IC staff members, such as a MEPS-IC response to a NCS users’ survey.
the use of NCS data to adjust wage parameters on the wage question in the MEPS-IC survey.
the benchmarking of MEPS-IC estimates with National estimates generated by the NCS program.
The aforementioned data users’ forum provided a major impetus for the herein proposed switch to collecting MEPS-IC survey data during their reference year, rather than in the following year. Finally, while many of the concerns regarding survey duplication have been resolved, MEPS-IC staff continues their communication and coordination efforts with NCS staff.
5. Small Business
MEPS-IC respondents include small businesses, and the data show that their health insurance coverage is different than that for large business. The MEPS-IC is designed to minimize respondent burden. Questions have been held to an absolute minimum required for the intended use of the data. Over time, questions have been dropped or changed which were confusing to respondents or which they were unable to answer. During past collection, many small businesses that offer insurance were able to answer most questions without referring to records. Many other small businesses are able to skip a large number of questions because they do not offer health insurance. Because small businesses generally do not offer their retirees health insurance they can skip that entire section of the form. Also, the benefit of doing data collection in the reference year, and thereby reducing burden due to easier access to needed information, is particularly true for small business.
6. Consequences of Less Frequent Data Collection
The MEPS-IC is an annual data collection activity. This clearance is intended to cover collection in 2008 and 2009. Less frequent collection would harm the quality of trend analysis and the ability to analyze and monitor changes caused by new state and federal policies and employer reactions to these policies and the market in general. It would also make data less relevant in a market where large changes in premiums and contributions occur annually. Less frequent list sample data collection would also be harmful to the support of the Gross Domestic Product and National Health Accounts production which is required annually.
7. Special Circumstances Relating to the Guidelines of 5 CFR 1320.6
This submission will fully comply with 5 CFR 1320.6 guidelines.
Federal Register Notice and Consultation Outside the Agency
8a. Federal Register Notice
Notice was published in the Federal Register on June 28, 2007 for 60 Days (see Attachment C). Two comments were received and are shown in Attachment D, along with AHRQ’s response to those comments.
There are no unresolved issues related to the OMB requirement.
8b. Consultations Outside the Agency
Paul Scheible
Office of Compensation and Working Conditions
Bureau of Labor Statistics
Bill Wiatrowski
Associate Commissioner
Office of Compensation and Working Conditions
Bureau of Labor Statistics
Cathy Cowan
Office of the Actuary
Center for Medicare and Medicaid Services
Joseph Piacentini
Office of Policy and Research
Employee Benefits and Security Administration
Patricia Willis
Office of Policy and Research
Employee Benefits and Security Administration
Alexandra Minicozzi
Office of Tax Analysis
US Department of the Treasury
Robert Stewart
Office of Assistant Secretary for Planning and Evaluation
U.S. Department of Health and Human Services
Brent Moulton, Associate Director
Bureau of Economic Analysis
U.S. Department of Commerce
Russell Lusher
Economist, Regional Wage Branch
Regional Economic Measurement Division
Bureau of Economic Analysis
Shirin Ahmed
Chief, Economic Planning and Coordination Division
Bureau of the Census
Kenneth Thorpe, Ph.D.
Robert Woodruffe Professor and Chair
Department of Health Policy and Management
Emory University
Len Nichols
Director, Health Policy Program
The New America Foundation
Ed Neuschler
Senior Program Officer
Institute for Health Policy Solutions
Ron Jeramias
Senior Economist
Joint Committee on Taxation
Stuart Hagan
Health and Human Resources Analyst
Congressional Budget Office
Lynn Blewitt
Principal Investigator
State Health Access Data Assistance Center
Michael Davern
Research Associate
State Health Access Data Assistance Center
Joyce Somsak
Associate Administrator
Healthcare Systems Bureau
Health Resources and Services Administration
P.J. Maddox, Ed.D.
Director, Office of Research
George Mason University
Linda Bartnyska
Chief, Cost and Quality Analysis
Maryland Health Care Commission
Amy Lischko
Commissioner, Division of Health Care Finance and Policy
Executive Office of Health and Human Services
Commonwealth of Massachusetts
9. Payments to Respondents
There are no payments to respondents.
10. Confidentiality Provisions
The list portion of the MEPS-IC is subject to rules and provisions set by the Bureau of the Census since it is drawn from the Census Bureau Business Register, a Census frame. Because of the use of the sampling frame from the Bureau of the Census, the MEPS-IC is bound by the confidentiality standards that apply to the Bureau of the Census. These standards, located in Title 13, Sections 8 and 9 of the United States Code are shown in Attachment E. Because the Census frame is developed using Internal Revenue Service Tax (IRS) information, the data also fall under the review of the IRS which conducts regular audits of the data collection, storage, and use. The statement that will be provided respondents is:
Your response to this voluntary survey is extremely important to assure accurate information is available for policymaker’s to make important decisions about our Nation’s health care system. ..............Sections 8 and 9 of Title 13, United States Code (the U.S. Census Bureau statute), specify that any information that could be used to identify your organization will not be disclosed to anyone other than U.S. Census employees. The information being collected will be used for statistical purposes only.
Respondents are told by the interviewer that confidentiality of their individual response is protected by Federal law prior to answering the questions.
11. Sensitive Questions
The MEPS-IC contains no questions generally considered sensitive.
12. Respondent Burden
MEPS-IC data collection method involves two basic steps. The first step is telephone prescreening. During prescreening, phone contact is made to verify addresses and to collect a small amount of information for the large number of employers who do not have health insurance. For those employers with insurance, a contact name and mailing address are collected for use in the second step in data collection.
The second step in collection involves forms mailing and follow-up. Two separate forms are used in this step – one collects establishment-level data (number of employees, retiree coverage, etc.) and the other collects plan-level data (enrollees, premiums, employee contributions, etc.). Both forms go to respondents known to have insurance, and are mailed in the same package along with an explanatory cover letter.
Respondents receiving the mailed questionnaires complete the establishment level form and one or more plan forms. Response time is similar whether a mail questionnaire is completed or the Census Bureau is required to collect the information by telephone follow-up.
Note that there are several versions (Series 10, 11, 11C, 15) of both the establishment-level and plan-level forms, which are all included in the Attachments. While the same information is collected on all versions of a form, the variations may be formatted differently or use slightly different terminology in order to obtain the requested information from different types of respondents. (For example, questions on forms sent to the private sector refer to the “organization” while those on forms sent to the government sample refer to the “government unit.”) Most importantly, the burden estimates are the same for the different versions of the same form.
The following tables show estimated hours and cost burdens. Burden estimates for all three collection types are based on recent MEPS-IC collections. The burden assumes that the list sample will be collected each year and will have the same sample size annually.
When reviewing Exhibit 1, it is important to understand these points:
There are fewer respondents for the prescreener (31,319) than in total (33,262) because some very large private and government units are not contacted during prescreening.
The total number of respondents is not the sum of the forms’ respondents because more than one form may be completed by any respondent.
“Hours per response” for the total is actually hours per unique respondent.
“Hours per response” per form is so specific because the format of the exhibit had to be revised after the total burden hours (19,032) were published in the 60-Day Federal Register Notice. Due to rounding in the original format, the very specific response hours had to be used in order to derive exactly 19,032 total response hours in the exhibit.
Exhibit 1. Estimated annualized burden hours
Form Name |
Number of respondents |
Number of responses per respondent |
Hours per response |
Total burden hours |
Prescreener |
31,319 |
1 |
.0944 |
2,957 |
Establishment |
25,789 |
1 |
.3758 |
9,692 |
Plan |
22,462 |
1.6 |
.1776 |
6,383 |
Total |
33,262 |
|
.57 |
19,032 |
Exhibit 2. Estimated annualized cost burden
Form Name |
Number of Respondents |
Total Burden hours |
Average Hourly Wage Rate* |
Total Cost Burden |
Prescreener |
31,319 |
2,957 |
25.81 |
76,320 |
Establishment |
25,789 |
9,692 |
25.81 |
250,151 |
Plan |
22,462 |
6,383 |
25.81 |
164,745 |
Total |
|
19,032 |
|
$491,216 |
*Based upon the mean hourly earnings for personnel, training, and labor relations specialists as published in “National Compensation Survey: Occupational Wages in the United States, June 2005, U.S. Department of Labor, U.S. Bureau of Labor Statistics.”
The total burden of the 2008-2009 MEPS-IC is estimated at 19,032 hours per year from 33,262 respondents. Assuming an average wage of $25.81 per hour for persons providing this information, the total estimated annual response cost for the 2008-2009 MEPS-IC survey is $491,216.
Estimates of Annualized Respondent Capital and Maintenance Costs
No costs other than those reported in Items 12 and 14 are expected to result from collection of this information.
14. Estimates of Annualized Cost to the Federal Government
The total annualized cost of the MEPS - MEPS-IC includes costs for the Census Bureau (for sample selection, data processing, data collection, editing, imputation and tabulations), of $9.95 million based on historic costs in the Inter-Agency agreement with Census. Another approximately $350,000 goes to direct staff costs at AHRQ.
15. Changes in Hour Burden
Total burden hours declined due to program changes – both a reduction in sample size for the ongoing list sample, due to budget constraints, and the elimination of the linked sample.
16. Time Schedule, Publication and Analysis Plans
The following is a brief schedule of major milestones for the scheduled project for the year 2008 data collection. The schedule for 2009 data collection would be similar.
Select sample Jan 2008-Mar 2008
Telephone number research
for the sample April 2008-June 2008
Telephone Prescreener June 2008-Aug 2008
First mailout June 2008-Aug 2008
Follow-up mailout Aug 2008-Oct 2008
Telephone follow-up Sept 2008-Dec 2008
Analyst review, edits and Aug 2008-Apr 2009
Callbacks
Private sector
Imputation and reweighting Apr 2009-May 2009
Produce and format
final tables and files June 2009-July 2009
Tables available July 2009
Governments
Imputation and reweighting August 2009-September 2009
Produce and format
final tables and files October 2009-November 2009
Tables available November 2009
As part of the tabulation plans, AHRQ intends to tabulate tables of key estimates. For the 2005 MEPS-IC, approximately 400 tables of estimates were produced for the private sector and a somewhat smaller number for governments. These included a set of important variables, such as, average premiums, average contributions, percent of establishments that offer health insurance, percent of employees eligible, percent of employees enrolled total and percent of self insured establishments each with estimates for a variety of cells. Cells are determined by crossing combinations of industry, size of firm, state and other characteristics. For instance, a set of estimates would be defined as the average premiums for each category defined by the state in which the establishment was located and the size of the firm that owned the establishment.
AHRQ also produces sets of estimates of total expenditures and enrollments for employer health insurance by industry, state and whether a plan is purchased or self-insured. This information is produced by request of CMS and BEA and is also of general interest to others.
AHRQ also produces special requests for other Federal and state agencies. Among these requests in the past have been special tabulations of self insured plans for the Department of Labor and several States, estimates of the numbers of employees who contributed a high percentage of the cost of their plans and estimates of percent of enrolled based upon the average salary distributions within the establishments. These latter types of runs have been produced for both the Department of the Treasury and the House Joint Committee on Taxation in order to assess Congressional proposals.
Based upon user requests, AHRQ has increased the number of tables regularly produced by a factor of 3 since the 1997 survey. Beginning with reference year 2004, tables were added with estimates for metropolitan areas within States (premiums and contributions, and offer, eligibility, and enrollment rates) and for copays and deductibles. Furthermore, as the survey has become more widely known, the numbers of special tabulations done for other government agencies has grown. As with the past survey, as resources and confidentiality restrictions permit, AHRQ also will produce special tabulations upon request.
Sets of estimates produced are placed on the AHRQ Website in a variety of formats (PDF, Excel, HTML, and CSV). The location of this information is publicized at professional meetings and through press releases in order to receive the widest possible use.
The list sample data is also made available to the research community through Census research data centers for microdata modeling and other forms of research. When working with the microdata, analysts must follow strict confidentiality procedures set forth by the Census Bureau and rigorous guidelines are followed concerning tabulations that can be released. This is done to assure that the promise of confidentiality given to survey respondents is kept.
The data center requires that a user submit a research proposal that is thoroughly reviewed before access is given to the micro data. Aside from these researchers, microdata access is limited only to the number of Census and AHRQ employees required to produce the survey.
As with any survey, Census and AHRQ perform various methodological studies to assess the quality of the data and sample design. Among studies done are benchmarking against results from other surveys, such as, the NCS, when similar estimates can be made. Other important research is also taking place to determine methods to improve sample design, weighting and post stratification of the results.
17. Expiration Date Display Exemption
No approval is being sought to conceal the expiration date for OMB approval of this information collection activity.
18. Exceptions to Certification
There are no provisions of this certification with which the co-sponsoring agencies of this information collection cannot comply.
| File Type | application/msword |
| File Title | SUPPORTING STATEMENT FOR OMB CLEARANCE |
| Author | Beth Levin Crimmel |
| Last Modified By | wcarroll |
| File Modified | 2007-10-03 |
| File Created | 2007-10-03 |