



          Supporting Statement for Information Collection Request for
Willingness to Pay Survey for Chesapeake Bay Total Maximum Daily Load: Instrument, Pre-test, and Implementation
                                       
                                       
                                       
                                    part a
                                       
                                       
                                       
                                       
                                       
                                       
                                       

                                       
                               TABLE OF CONTENTS

PART A OF THE SUPPORTING STATEMENT	3
1.	Identification of the Information Collection	3
1(a)	Title of the Information Collection	3
1(b)	Short Characterization (Abstract)	3

2.	Need For and Use of the Collection	5
2(a)	Need/Authority for the Collection	5
2(b)	Practical Utility/Users of the Data	5

3.	Non-duplication, Consultations, and Other Collection Criteria	6
3(a)	Non-duplication	6
3(b)	Public Notice Required Prior to ICR Submission to OMB	9
3(c)	Consultations	9
3(d)	Effects of Less Frequent Collection	12
3(e)	General Guidelines	12
3(f)	Confidentiality	12
3(g)	Sensitive Questions	13

4.	The Respondents and the Information Requested	13
4(a)	Respondents	13
Table A1: Geographic Stratification Design	14
4(b)	Information Requested	15
(I)	Data items, including recordkeeping requirements	15
(II)	Respondent activities	20

5.	The Information Collected - Agency Activities, Collection Methodology, and Information Management	21
5(a)	Agency Activities	21
5(b)	Collection Methodology and Information Management	21
5(c)	Small Entity Flexibility	23
5(d)	Collection Schedule	23
Table A2: Schedule for Survey Implementation	24

6.	Estimating Respondent Burden and Cost of Collection	25
6(a)	Estimating Respondent Burden	25
6(b)	Estimating Respondent Costs	25
6(c)	Estimating Agency Burden and Costs	26
6(d)	Respondent Universe and Total Burden Costs	26
6(e)	Bottom Line Burden Hours and Costs	26
Table A3: Total Estimated Bottom Line Burden and Cost Summary	27
6(f)	Reasons for Change in Burden	27
6(g)	Burden Statement	27


PART A OF THE SUPPORTING STATEMENT

1.	Identification of the Information Collection
1(a)	Title of the Information Collection 

Willingness to Pay Survey for Chesapeake Bay Total Maximum Daily Load:  Instrument, Pre-test, and Implementation

1(b)	Short Characterization (Abstract)

      The Clean Water Act (CWA) directs EPA to coordinate Federal and State efforts to improve water quality in Chesapeake Bay. In 2009, Executive Order (E.O.) 13508 reemphasized this mandate, directing EPA to define the next generation of tools and actions to restore water quality in the Bay and describe the changes to be made to regulations, programs, and policies to implement these actions. In response to the Executive Order, EPA established pollution budgets, called Total Maximum Daily Loads (TMDL), of nitrogen, phosphorus, and sediment for the Chesapeake Bay. These TMDLs called for reductions of 25, 24, and 20%, respectively, of these pollutants (EPA 2011).
      The watershed states of New York, Pennsylvania, Delaware, West Virginia, Virginia, and Maryland, as well as the District of Columbia, have developed Watershed Implementation Plans (WIPs) detailing the steps each will take to meet its obligations under the TMDL.  EPA has begun a new study to estimate costs of compliance with the TMDL.  It is important to put cost estimates in perspective by estimating corresponding benefits.  
       EPA's National Center for Environmental Economics (NCEE) is undertaking a benefits analysis of improvements in Bay water quality under the TMDL, as well as of ancillary benefits that might arise from measures taken to meet the required reductions in the Bay.  Benefits from meeting the TMDL for the Chesapeake Bay will accrue to those who live near the Bay or visit for recreation, those who live near or visit lakes and rivers in the watershed, and those who live further away and may never visit the Bay but have a general concern for the environment.  While benefits from the first two categories can be measured using hedonic property value, recreational demand, and other revealed preference approaches being used by NCEE, only stated preference methods can capture nonuse benefits. 
      To assess the benefits of the TMDL for the Chesapeake Bay, NCEE requests approval from the Office of Management and Budget to conduct a stated preference survey. The survey will collect data on households' prior use of Chesapeake Bay and its watershed, preferences for a variety of water quality improvements likely to follow pollution reduction programs, and demographic information. EPA developed four versions of the survey questionnaire based on input from 10 focus groups and 26 cognitive interviews (OMB Control # 2090-0028).  Each of the four questionnaire versions differs based on assumptions of baseline water quality conditions that would occur without program implementation and on the type of attributes used to represent water quality improvements in the choice questions. One version of the questionnaire will be administered to each sampled household. 
      The target population and sample for this stated preference survey is individuals from east coast US households who are 18 years of age or older.  Section 4, below, contains more information on the target population and sample.  Non-response bias has the potential to occur if there is a systematic difference between characteristics of households returning completed surveys and non-respondents.  EPA will use a combination of telephone and priority mailing to conduct a non-response study. EPA will analyze the characteristics of the completed and non-completed cases from the mail survey and non-response questionnaire to determine whether there is any evidence of significant non-response bias in the completed sample.  This analysis will suggest whether any weighting or statistical adjustment is necessary to minimize the non-response bias in the completed sample.
      EPA will compile survey responses into a database, using responses to estimate willingness to pay for changes related to reductions in nitrogen, phosphorous, and sediment loadings to the Bay. EPA will rely upon standard theoretical tools for non-market welfare analysis to analyze these data, as presented by authors including Freeman (2003) and Just et al. (2004).  Survey results will be used to derive total values following standard practice for choice experiments (Adamowicz et al. 1998). Primary and non-response survey information will be used to assist in the interpretation and validation of survey responses. 
      The total national burden estimate for all components of the survey is 1,040 hours.  The burden estimate is based on 2,400 respondents to the 8,000 mailed questionnaires and 400 respondents to the combined telephone and priority mail non-response survey.  EPA assumes an average burden estimate of 25 minutes per mail survey respondent including the time necessary to complete and mail back the questionnaire and 5 minutes for each participant in the non-response survey.   Given an average wage rate of $23.33 (United States Department of Labor, 2012), the total respondent cost is $24,263.
      

2.	Need For and Use of the Collection
2(a)	Need/Authority for the Collection

      The project is being undertaken pursuant to section 104 of the Clean Water Act dealing with research. Section 104 of the Clean Water Act authorizes and directs the EPA Administrator to conduct research into a number of subject areas related to water quality, water pollution, and water pollution prevention and abatement. This section also authorizes the EPA Administrator to conduct research into methods of analyzing the costs and benefits of programs carried out under the Clean Water Act.
      This project is exploring how public values for ecosystem goods and services are affected by nutrient and sediment loading to the Chesapeake Bay.  Understanding total public values for ecosystem resources, including the more difficult to estimate non-use values, is necessary to determine the full range of benefits associated with reductions in nutrient and sediment loading.  Because non-use values may be substantial, failure to estimate such values may lead to improper inferences regarding benefits and costs. The findings from this study will be used by EPA to estimate the total value of economic benefits of the nutrient and sediment TMDLs designed to meet the requirements of Executive Order 13508.

2(b)	Practical Utility/Users of the Data
 
      EPA plans to use the results of the stated preference survey to estimate the net welfare impacts of the Chesapeake Bay TMDLs.  Specifically, the Agency will use the survey results to estimate values for improvements in Bay and watershed ecosystems under the TMDL, as well as for ancillary benefits that might arise from terrestrial measures taken to improve water quality  In this work, the Agency will follow standard practices outlined in the literature (Freeman 2003; Bennett and Blamey 2001; Louviere et al. 2000; U.S. EPA 2000).  
	The results of the study will be made available to state and local governments which they may use to better understand the preferences of households in their jurisdictions and the benefits they can expect as a result of meeting the TMDL.  Finally, stakeholders and the general public will be able use this information to understand the social benefits of improving water quality in the Chesapeake Bay to accompany the cost information also being developed by EPA.  	

3.	Non-duplication, Consultations, and Other Collection Criteria
3(a)	Non-duplication

      There are many studies in the environmental economics literature that quantify benefits or willingness to pay (WTP) associated with various types of water quality and aquatic ecosystem changes. As an iconic resource and the subject of a long history of restoration efforts, Chesapeake Bay is relatively well-studied but there is no study or set of studies that provides a comprehensive estimate of non-use values or values associated with specific improvements likely to result from the TMDL.  Further, existing studies do not provide a sufficient basis for using benefits transfer to estimate the value of these improvements.  The proposed survey is designed to fill this gap. 
      The most recent review of valuation studies relevant to the Chesapeake Bay TMDL is Cropper and Isaac (2011), which identifies studies that estimate WTP for use and non-use values associated with the Bay.  Cropper and Isaac also consider how studies not associated with Bay may be used in a benefits transfer exercise to value improvements resulting from the TMDL.   While these studies all provide insights into particular categories of benefits associated with the Bay, none provides a comprehensive estimate of non-use values or values associated with specific improvements likely to result from the TMDL.   
      Three of the studies reviewed in Cropper and Isaac examine the residential amenity value of water quality (Leggett & Bockstael, 2000; Poor et al., 2007; Van Houtven, 2009).  Leggett and Bockstael (2000) estimate the impact on housing prices from fecal coliform counts on waterfront properties, finding significant effects.  Poor, et al. (2007) study the effect of dissolved inorganic nitrogen and total suspended solids on property values in St. Mary's County, MD and also find significant effects.  These studies, however, reflect the benefits to homeowners on or near the Bay and not benefits more widely.  
      Recreational benefits in the Chesapeake Bay have been relatively well-studied, and Cropper and Isaac identify six original studies that estimate the recreational activity value of improved water quality. (Bockstael et al., 1988, 1989;, Lipton and Hicks, 1999; Hicks and Strand, 2000;; Lipton and Hicks, 2003; Lipton, 2004; Massey et al., 2006).  Cropper and Isaac also identify three studies that use benefit transfer to estimate recreation benefits (Krupnick, 1988; Morgan and Owens, 2001; Van Houtven, 2009).  The studies focus on the benefits from recreational fishing, swimming or beach visits, and boating.  The limitations of these studies for estimating the benefits of the TMDL are described in Cropper and Isaac.  Key limitations include that only two species (striped bass and flounder) are examined in the recreational fishing studies, and the choice of water quality measure for swimming and boating studies. These limitations make it difficult to use the studies for benefit transfer.  More importantly for this ICR, these studies, even taken together, look at a small subset of the kinds of benefits expected from the TMDL.
       Commercial fishing benefits from improved water quality are the subject of three studies identified in Cropper and Isaac (Kahn and Kemp, 1985; Anderson, 1989; Mistiaen et al., 2003).  While commercial benefits from the TMDL are likely to be important, they are distinct from the other types of benefits the proposed stated preference study will estimate.
      There is a large gap in the literature on the non-use benefits of improved water quality in the Chesapeake Bay.  Cropper and Isaac (2011) identified only two original research and studies that estimate non-use benefits of water quality improvements in the Chesapeake Bay.  The first study, Bockstael, et al. (1988, 1989), estimates willingness to pay to make the Bay "swimmable" for those respondents who considered that it was not acceptable for swimming.  There is no clear way to link "swimmability" to the improvements expected from the TMDL, making it difficult to use this study for valuation.  Further, the study sample was limited to the Washington-Baltimore area.  A second study, Lipton et al. (2004), estimated willingness to pay of non-users for restoring oyster reefs in the Chesapeake Bay. While oyster reefs are likely to benefit from the TMDL, this is only a small part of the expected environmental improvements.  The sample in this study was broader, including most mid-Atlantic states, but did not include Pennsylvania and was not a random sample.
      An alternative to directly estimating Chesapeake Bay non-use benefits is to use existing studies that are similar but focus on other locations.  Van Houtven (2009) suggests this approach, based on a meta-analysis of stated preference studies that value changes in a water quality index.  This approach can produce value estimates for improvements in the index, and it may be feasible to map changes resulting from the TMDL into a comparable index.  However, as noted by Cropper and Isaac (2011), the water bodies used in the meta-analysis differ greatly from the Chesapeake Bay, and the values reported from Van Houtven (2009) are substantially different from those in Bockstael, et al. (1988).  The limitations of existing Bay-specific studies, and the inability to effectively apply benefits transfer, suggest that an original study is needed to estimate non-use benefits from the TMDL.   
      One additional group of researchers, Hicks et al. (2008), used a choice experiment to survey knowledgeable stakeholders (e.g., recreational anglers, charter boat operators) with close connection to the Chesapeake Bay and its restoration about their preferences for a hypothetical Bay restoration package in the context of a prior Chesapeake Bay restoration effort, "Chesapeake 2000." They account for a variety of outcomes related to reduced sediment and nutrient loads: number of seafood consumption advisories, number of beach closures, oyster biomass, blue crab biomass, shad population, and acres of wetlands. While perhaps the most closely related among available studies, the choice experiment setup and scale of the surveyed population prevent its direct use for analysis of benefits of the TMDLs. Specifically, cost was not included as an attribute in the choice experiment, preventing EPA from estimating WTP for restoration activities based on these results. Furthermore, only knowledgeable stakeholders were surveyed; preferences for water quality improvements achieved under the Chesapeake Bay TMDL among this group may substantially differ from that of the general and/or non-user population.
      The proposed stated preference study addresses one additional gap: benefits to the upper watershed that result from TMDL implementation.  Based purely on proximity, residents of the upper watershed are less likely to have direct use value for in-the-Bay benefits of TMDL implementation, yet face substantial implementation costs.  Upper-watershed BMPs designed to reduce nutrient and sediment loadings to the Bay will also improve water quality in the over 150 major rivers and thousands of lakes and reservoirs in the watershed; since upper-watershed residents have relatively more direct access to these amenities, not counting upper-watershed or ancillary benefits is likely to substantially underestimate total benefits of the TMDL.  These effects are not well-considered in the existing economics literature.  For example, Cropper and Isaac (2011) identified only one appropriate study (von Haefen, 2003) related to ancillary benefits from BMP implementation.
      To conclude, while specific aspects of uses of the Bay have been well studied, an original stated preference study is  needed to properly estimate benefits of the Chesapeake Bay TMDL due to the following  reasons: (1) valuation endpoints used in the existing studies tend to be incompatible with endpoints used to measure water quality improvement under the Chesapeake Bay  TMDL (e.g., water clarity and changes in Bay grasses); (2) necessary data to transfer benefits estimates from existing studies to the Chesapeake Bay are lacking, particularly regarding use and non-use data for upper-watershed ecosystems like lakes, reservoirs and rivers; (3) the overall population likely to benefit from Chesapeake Bay TMDL actions is broader than the populations sampled in prior studies, which would likely result in incorrect value estimates; and (4) although studies separating use and non-use components of total value estimate that non-use value is substantial, available studies are based on discrete endpoints (i.e., "swimmability" in the Bay itself, oyster reef acreage) that do not reflect the full range of TMDL-related Bay and watershed improvements (e.g., ancillary improvements in upper watershed water quality, aquatic grasses in the Bay). While these concerns might be ameliorated by using results from environmental economics studies based in other estuaries of the United States, EPA believes that studies based in other estuaries are unlikely to accurately or completely capture willingness to pay for TMDL-related improvements in the Chesapeake Bay watershed given the unique character of this water resource and the goods and services it provides. Cropper and Isaac reach a similar conclusion, writing, "we strongly suggest that a new stated preference study be conducted to elicit willingness to pay for water quality improvements more closely linked to the TMDLs" (Cropper and Isaac, 2011, p. 20).

3(b)	Public Notice Required Prior to ICR Submission to OMB

      In accordance with the Paperwork Reduction Act (44 U.S.C. 3501 et seq.), EPA published a notice in the Federal Register on May 24, 2012, announcing that EPA is planning to submit this request for a new Information Collection Request (ICR) to the Office of Management and Budget (OMB), and soliciting comments on aspects of the information collection request.  A copy of the Federal Register notice (77 FR 31006) is attached at the end of this document (See Attachment 5).  Also see docket # EPA-HQ-2012-0033.  
      
3(c)	Consultations

Consultations with Scholars: Prior to commencing the survey design phase of this project, EPA co-sponsored a workshop on October 31 and November 1, 2011 at Resources for the Future on the costs and benefits of protecting and restoring the Chesapeake Bay.  The main purpose of the workshop was to gather scholars who are working on estimating the benefits and costs of water quality improvements in the Bay to exchange ideas.  The agenda included a wide range of  presentations on the costs and benefits of improving water quality in the Bay and its watershed, the possibility of ecosystem service payments and trading to ameliorate some of the costs, ancillary benefits that may arise, water quality and ecological modeling for the Bay, and policy options for approaching protection and restoration. Although the workshop topics were clearly broader than stated preference techniques, the agenda did include presentations by EPA on the proposed information collection request.  Participants included academics representing academics from major universities within the Chesapeake Bay Watershed (MD, VA, DE, WVA, PA); representatives from several NGOs including the Chesapeake Bay Foundation and the Chesapeake Bay Trust; as well as participants from several government agencies (NOAA, DoI, and USDA).

Consultations with Respondents: As part of the planning and design process for this collection, EPA conducted a series of 8 focus groups and 26 protocol interviews (one-on-one interviews) in venues located inside the tidal portion of the Chesapeake Bay Watershed as well as sites in up-stream locations. An additional two focus groups were conducted in locations outside of the Watershed (for a total of 10 focus groups).  While early focus group sessions were used to narrow the attributes to be highlighted in a survey and the kinds of information respondents would need to answer the questions, later sessions and protocol interviews were employed to test the draft survey materials.  These consultations with potential respondents were critical in identifying sections of the questionnaire that were redundant and lacked clarity and in producing a survey instrument meaningful to respondents.  The later focus group sessions and the protocol interviews were also helpful in estimating the expected amount of time respondents would need to complete the survey instrument.  While completion times varied somewhat, most participants required approximately 20 to 25 minutes to complete the draft surveys.  The focus group session and protocol interviews were conducted under OMB Control  # 2090-0028.

Survey Design Team: Dr. Christopher Moore at the U.S. Environmental Protection Agency serves as the project manager for this study. Dr. Moore is assisted by Dr. Chris Dockins, Dr. Dennis Guignet, Dr. Kelly Maguire and Dr. Nathalie Simon, all with the U.S. EPA's National Center for Environmental Economics. Dr. Alan Krupnick (ph: 202-328-5107), Senior Fellow and Director, Center for Energy Economics and Policy at Resources For the Future, Dr. Maureen Cropper (ph: 202-328-5083), Senior Fellow at Resources for the Future and Professor of Economics at the University of Maryland, and Mr. William Isaacs (ph: 202-328-5082), Research Assistant at Resources for the Future, provide review of the survey development, focus group materials, and one-on-one interview materials.  Dr. Elena Besedin (ph. 617-349-2770), Senior Economist at Abt Associates Inc, and Ryan Stapler (ph. 617-520-3524), M.S., Senior Analyst at Abt Associates Inc. provide contractor support.        
      Dr. Alan Krupnick is Director of the Resources for the Future's Center for Energy Economics and Policy, as well as Director of Research and a Senior Fellow at RFF, and specializes in analyzing environmental and energy issues, in particular, the benefits, costs and design of pollution and energy policies. He was lead author for Toward a New National Energy Policy: Assessing the Options study, examining the costs and cost-effectiveness of a range of federal energy policy choices in both the transportation and electricity sectors. His primary research methodology is in the development and analysis of stated preference surveys. Dr. Krupnick has been a consultant to state governments, federal agencies, private corporations, the Canadian government, the European Union, the World Health Organization, and the World Bank. He is a regular member of expert committees from the National Academy of Sciences and the U.S. EPA. 
      Dr. Maureen Cropper, Professor of Economics at the University of Maryland and Senior Fellow at Resources for the Future, has made major contributions to environmental policy through her research, teaching, and public service. Her research has focused on valuing environmental amenities, estimating consumer preferences for health and longevity improvements, and the tradeoffs implicit in environmental regulations. Previously at the World Bank, her work focused on improving policy choices in developing countries through studies of deforestation, road safety, urban slums, and health valuation.  From 1994 through 2006, she served on the U.S. Environmental Protection Agency's Science Advisory Board, where she chaired the Advisory Council for Clean Air Act Compliance Analysis and the Environmental Economics Advisory Committee. She is a research associate of the National Bureau of Economic Research and a member of the National Academy of Sciences. 
      Dr. Elena Y. Besedin, a senior economist at Abt Associates Inc., specializes in the economic analysis of environmental policy and regulatory programs.  Her work to support EPA has concentrated on analyzing economic benefits from reducing risks to the environment and human health and assessing environmental impacts of regulatory programs for many EPA program offices.  She has worked extensively on valuation of non-market benefits associated with environmental improvements of aquatic resources.  Dr. Besedin's empirical work on non-market valuation includes design and implementation of stated and revealed preference studies and benefit transfer methodologies.
      
Consultations with other Government Agencies: EPA has been working closely with ecosystem modelers in NOAA's Chesapeake Bay Office and National Marine Fisheries Service's Office of Habitat Conservation.  Specifically, NOAA's modelers have provided assistance with the eco-system based fishery models "Ecopath with Ecosim" and "Atlantis."  These consultations have been instrumental in examining the ecological impacts of reducing nutrient and sediment loads to the Bay of the ecosystem-based fishery models, have provided useful background for the survey instrument itself, and will allow EPA to more accurately translate the values people place on the various attributes of the Chesapeake Bay highlighted in the survey to benefits estimates associated with the TMDLs. 
      .
Consultations with Peer Reviewers: The survey instrument has subsequently undergone peer review by three leading scholars specializing in stated preference surveys for estimating benefits associated with environmental improvements:  Dr. Kevin Boyle, Professor, Department of Agricultural and Applied Economics, Virginia Tech University; Dr. John Whitehead, Professor and Chair, Department of Economics, Walker College of Business, Appalachian State University; and Dr. Robert Johnston, Director, George Perkins Marsh Institute, Professor, Department of Economics, Clark University. 
	
3(d)	Effects of Less Frequent Collection

      The survey is a one-time activity.  Therefore, this section does not apply.

3(e)	General Guidelines

      The survey will not violate any of the general guidelines described in 5 CFR 1320.5 or in EPA's ICR handbook.

3(f)	Confidentiality

All responses to the survey will be kept confidential to the extent provided by law.  To ensure that the final survey sample includes a representative and diverse population of individuals, the survey questionnaire will elicit basic demographic information, such as age, number of children under 18, type of employment, and income. However, the detailed survey questionnaire will not ask respondents for personal identifying information, such as names or phone numbers.  Prior to taking the survey, respondents will be informed that their responses will be kept confidential to the extent provided by law.  The survey data will be made public only after it has been thoroughly vetted to ensure that all potentially identifying information has been removed.

3(g)	Sensitive Questions

      The survey questionnaire will not include any sensitive questions pertaining to private or personal information, such as sexual behavior or religious beliefs.

4.	The Respondents and the Information Requested
4(a)	Respondents 
      
The target population for the stated preference survey is individuals who are 18 years of age or older and reside in one of 18 east coast U.S. states:  For the selection of households, the population of households in the 18 states and the District of Columbia will be stratified by three mutually-exclusive study regions: states immediately bordering Chesapeake Bay (Maryland and Virginia) and the District of Columbia, states in the Chesapeake Bay watershed but not immediately bordering the Bay itself (Delaware, New York, Pennsylvania, and West Virginia), and eastern states outside of the watershed(Vermont, New Hampshire, Massachusetts, Connecticut, Rhode Island, Maine, Ohio, North Carolina, South Carolina, Georgia, and Florida).
      Survey participants are selected randomly from the U.S. Postal Service Delivery Sequence File (DSF), which covers over 97% of residences in the U.S. The survey households that will be sampled from the DSF include city‐style addresses and PO boxes, and covers single‐unit, multi‐unit, and other types of housing structures.  EPA will send a copy of the mail survey to a random stratified sample of 8,000 households. Approximately 2,400 of the adults of the 8,000 adults sent a survey are expected to return a completed survey.
      Table A1 shows the stratification design for the geographic regions included in this study.  More detail on planned sampling methods and the statistical design of the survey can be found in Part B of this supporting statement.

Table A1: Geographic Stratification Design
Region
                                States Included
                                    Sample
                                    Size[a]
                             Percentage of Sample
Chesapeake Bay
MD, VA, DC
                                                                            800
                                                                            33%
Chesapeake Bay watershed
DE, NY, PA, WV
                                                                            800
                                                                            33%
Additional East Coast states
CT, FL, GA, MA, ME, NC, NH, OH, RI, SC, VT
                                                                            800
                                                                            33%
Total for Sample Regions

                                                                          2,400
                                                                           100%
[a] Sample sizes presented in this table include total expected completed surveys.

      The total number of households recruited from each region will be divided among four different versions of the survey. Each version of the survey (i.e., box A, B, C, and D, in Figure A1 below) represents a combination of assumed baseline scenarios and indicators chosen to represent ecological and water quality conditions in the Chesapeake Bay watershed used in the choice experiment (e.g., attributes). Multiple survey versions are warranted for several reasons. First, some scientific uncertainty remains regarding the future state of Chesapeake Bay absent the TMDL(e.g., Kemp et al., 2005).  Therefore, EPA is stratifying its sample design to reflect two possible scenarios representing future conditions without TMDL implementation: (1) water quality remains the same in the future as it does today(constant baseline) and (2) water quality in the Bay would decline in the future, relative to today (declining baseline). Given that there will almost certainly be some uncertainty regarding the specifics of the "actual" baselines and improvements, the resulting valuation estimates will allow flexibility in estimating WTP for a wide range of different circumstances.
      Second, there are no standard protocols for defining the effects of ecological changes in stated preference surveys. Boyd and Krupnick (2009) distinguish "inputs" -- environmental features or conditions that, via natural processes, are converted into different features or conditions, from "endpoints" -- biophysical outputs that are more "directly and economically meaningful" to firms and households. In the case of Chesapeake Bay improvements under TMDL implementation, many environmental features are both ecological endpoints in their own right, as well as inputs into the production function for other endpoints. For example, improved water clarity is an aesthetic endpoint for swimmers and boaters but it is also required to restore bay grass habitat which serves as a nursery for commercially valuable fish and shellfish. Because TMDL actions are likely to produce bundles of improvements classified as both endpoints and inputs, and to reduce respondents' cognitive burden in disentangling this complexity, EPA has found the best approach is to administer separate survey versions  based on either input-based or endpoint-based indicators of ecosystem improvements.  The "input" version of the survey describes ecosystem effects of TMDL implementation in terms of ecosystem services and goods such as water quality that are not produced by the Bay but serve as inputs into ecological production function. The "endpoint" version links effects of TMDL implementation to the directly-consumable ecosystem goods such fish population that are produced by the Bay.
      EPA plans to administer a survey version for each combination of the baseline conditions (i.e., declining or constant) and " input" vs. "output" survey attributes (see Figure A1 below)..
      
                                       
                               Constant Baseline
                              Declining Baseline
                             Attributes are Inputs
                                   Version A
                                   Version C
                           Attributes are Endpoints
                                   Version B
                                   Version D
      
Figure A1. Concept map of survey versions, which differ based on choice experiment attributes and information provided about future baseline conditions in the absence of the TMDL. 
      
A sample of 1,800 households would complete versions A and C of the survey. A sample of 600 households would complete versions B and D.  Linking the management practices under the TMDL to forecasts of the "endpoint attributes" requires an additional step of modeling compared with the "input attributes" and one that is based on relatively new science.  For this reason the benefits analysis will rely primarily on the input version of the survey and the endpoint version of the survey will be used for comparison.  Those totals are divided across regional samples, and are allocated to each region in proportion to the total number households in that region.  This is the number required to estimate the main effects under an experimental design model. Part B of this document provides detail on sampling methodology. 

4(b)	Information Requested
(I)	Data items, including recordkeeping requirements
Households randomly selected from the U.S. Postal Service DSF database will be mailed a copy of the survey. The full texts of Versions A (input; constant baseline), B (endpoint; constant baseline), C (input; declining baseline) and D (endpoint, declining baseline) of the mail survey are provided in Attachments 1, 2, 3, and 4.  EPA developed the survey based on the findings of a series of ten focus groups and 26 cognitive interviews conducted as part of survey instrument development (OMB control # 2090-0028).  EPA has determined that all questions in the survey are necessary to achieve the goal of this information collection, i.e., to collect data that can be used to support an analysis of the total benefits of Chesapeake Bay TMDL implementation.
      The following is an outline of the major sections of the input versions of the survey (Versions A and C).
The Chesapeake Bay Watershed.  This section provides a map and narrative description of the Chesapeake Bay and watershed, defines the concept of a watershed, and distinguishes between freshwater and tidal water bodies. The first survey question asks respondents about their recreation in lakes, rivers, and tidal waters of the Chesapeake Bay and watershed during the last five years. The second survey question asks respondents about how frequently they visited these water bodies to participate in recreational activities. Information and questions in this section are designed to elicit the respondent's status as either a recreational user or non-user of the Chesapeake Bay and watershed. This section begins to prepare respondents to answer the stated preference questions by motivating respondents to consider how they use or do not use the Chesapeake Bay watershed.  Responses to questions 1 and 2 can also be used to test if certain respondent characteristics influence responses to the referendum questions.  
Water Quality in the Chesapeake Bay Watershed.  This section provides respondents information about nutrient and sediment pollution in the Chesapeake Bay, thereby promoting understanding of the context for subsequent stated preference questions. Question 3 checks the respondent's prior knowledge about the effect of these pollutants on wildlife habitat, and question 4 checks awareness of the sources of nutrient and sediment pollution in the Chesapeake Bay watershed.
Reducing excess nutrients and sediment in the Chesapeake Bay Watershed. This section begins by presenting a general list of practices to improve water quality, a list of water quality improvements resulting from changes in the way land, storm water and waste water are managed, and resulting improvements in ecosystems and recreation. Information in this section is consistent with the TMDLs, and is designed to promote understanding of the type of actions taken to achieve reductions in nutrient and sediment pollution. It also informs respondents of the major benefits of such actions. Following this information, question 5 asks respondents about their perceptions of how environmental conditions in lakes and rivers close to home, and in Chesapeake Bay, have changed compared to 10 years ago. Question 5 also solicits respondents' level of agreement or disagreement with statements about their concern about current pollution levels in (a) lakes and rivers and (b) Chesapeake Bay, and with a statement that some decrease in environmental quality is inevitable given economic growth. The latter question is designed to assess respondents' subjective perception of water quality in the Bay. 
Environmental Quality in the Chesapeake Bay.  Information and questions in this section are designed to promote understanding of the water quality attributes used in subsequent choice experiments. This section encourages respondents to consider the meaning of each attribute and consider their general preferences for these qualities, prior to considering specific policy options.  The section begins by describing three input-based indicators of Chesapeake Bay environmental quality: dissolved oxygen, water clarity, and amount of aquatic grass habitat, and tells respondents that state governments and federal agencies work with scientists to develop goals for Chesapeake Bay water quality based on these measurements. Then, Question 6 asks if, before the survey, respondents had heard about environmental goals for dissolved oxygen, water clarity, or aquatic grasses. 
Next, a short narrative and chart describe baseline environmental quality of Chesapeake Bay tidal waters and establish predicted conditions for the year 2025 if no further action is taken to reduce nutrients and sediment. 
Environmental Quality in Lakes. This section describes the how different amounts of nutrients entering lakes can affect aesthetic and ecological conditions.  The information is based on the trophic state index (Carlsson and Simpson 1996) and links it to measurements of visibility, water color, and most common types of fish. To place respondents in a mindset where they are cognizant of baseline environmental quality in lakes, question 7 asks respondents to rate the lake condition in lakes closest to their home. Ratings are based on a lake condition scale used earlier in the section to illustrate water quality.  
Lakes in the Chesapeake Bay Watershed. This section encourages respondents to synthesize information about the upstream effects of achieving nutrient and sediment reductions in Chesapeake Bay. A short narrative and chart describe baseline environmental quality of lakes in the Chesapeake Bay watershed portion of the respondent's state of residence and establish predicted conditions in these lakes for the year 2025 if no further action is taken to reduce nutrients and sediment.  No questions are asked in this section.
Pollution Reduction Programs for the Chesapeake Bay Watershed.  This section informs respondents that federal and state agencies are developing programs to meet water quality and wildlife goals for Chesapeake Bay, including the time frame of implementation and information about how progress towards goals would depend on program design. This information sets up the following choice experiment, putting respondents in the frame of mind that different programs will have different effects on the Bay. Question 9 asks respondents if they have ever heard of nutrient and sediment reduction programs in the Chesapeake Bay watershed. 
Paying for Pollution Reduction Programs.  This information is designed to remind respondents that there are other items on which their household could spend their income, rather than spending these funds to pay for changes required under pollution reduction programs.  Such questions are commonly used in stated preference surveys (e.g., Mitchell and Carson 1984) to remind respondents that, given scarce household budgets, paying for pollution reduction programs reduces their ability to purchase substitute goods.  
Deciding Future Actions. The questions in this section are the key part of the survey.  Questions 10, 11, and 12 are "choice experiment"  or "choice modeling" questions (Adamowicz et al. 1998; Bennett and Blamey 2001), and ask respondents to choose how they would vote, if presented with two hypothetical regulatory options (and a third "status quo" choice to reject both options) for waters within Chesapeake Bay tidal waters and lakes in the Chesapeake Bay watershed.  Each of the multi-attribute options is characterized by (a) percent of Bay area meeting dissolved oxygen goals, (b)  feet of visbility in Bay tidal waters (c) acres of aquatic grass habitat, (d) average condition of lakes in the respondent's state (based on water color, visibility, and fish population composition), and (e) an unavoidable, permanent cost of living increase for the respondent's household.  Environmental attribute levels for all three scenarios are described both in numerical terms and in percentage increase or decrease from present day conditions. The cost attribute is defined in terms of both annual cost and monthly cost. In the "declining baseline" versions of the survey (e.g., Versions C and D), the "status quo" option references attribute levels that are lower than the present day. In the "constant baseline" versions of the survey (e.g., Versions A and B), the "status quo" option references attribute levels that are the same as present day (e.g., 0% change).
      Following standard choice experiment methods, respondents choose the regulatory options that they prefer, based on their preferences.  Respondents always have the option to vote for neither option -- providing the status quo option is necessary for appropriate welfare estimation (Adamowicz et al. 1998).  Advantages of choice experiments, and the many examples of the use of such approaches in the literature, are discussed in later sections of this ICR.  Following standard approaches (Opaluch et al. 1993, 1999; Johnston et al. 2002a; 2002b, 2003b), respondents are instructed to answer each of the three choice questions independently, and not to add up or compare programs across different pages.  This is included to avoid biases associated with sequence aggregation effects (Mitchell and Carson 1989).  
      Questions 13, 14, 15 and 16 ask respondents to rate the effect of factors on their choices, and why they voted for or against the regulatory programs.  Factors include the relative consideration given to lakes versus Bay waters (Question 13), the ranked importance of choice experiment attributes (Question 14), the most important consideration aside from cost (Question 16), and agreement or disagreement with a series of statements that describe respondent voting in the prior questions. Responses to questions such as these have been used in the literature to successfully control for hypothetical bias. 
      Question 18 is a follow-up to the prior voting questions, and asks respondents to identify the primary reason for voting no, if they always voted for "no policy" in questions 10-12.  It is designed to identify respondents whose "no policy" responses are based on their budget constraint, respondents who do not consider Bay and freshwater lake improvements important enough to vote for a policy, or respondents who ignored information presented in the survey and answered questions based on their general convictions and principles. In general, this question provides information that will be used by EPA to identify protest responses. Question 18 is a follow-up question to the prior voting, and is asked only of respondents who voted for one of the regulatory programs in the prior questions. Responses to question 19 are important to successfully control for hypothetical bias.   
Demographics.  Questions 19-26 ask respondents to provide basic demographic information, including age, gender, marital status, highest level of education, household composition (number of children under 18), dwelling type and ownership status, and employment industry.  This information will be used in the analysis of survey results, as well as in the non-response analysis. Responses to these questions will be used to estimate the influence of demographic variables on respondents' voting choices, and ultimately, their WTP for water quality improvements.
Comments.  The survey offers respondents a chance to comment on the survey. This section helps identify protest responses, i.e. responses from individuals who rejected the format of the survey or the way the questions were phrased.
      
      Differences in the survey instrument for the endpoint version relative to the input survey are as follows:
      
Environmental Quality in the Chesapeake Bay. The section replaces two of the three input-based indicators of Chesapeake Bay environmental quality with endpoint-based environmental goals: number of blue crabs and tons of oysters. Water clarity  is retained as one of the environmental indicators. These replacements are carried throughout the remainder of the questionnaire.
Deciding Future Actions. Choice experiment attributes in questions 10-12 include (a) Bay tidal water average water clarity, (b) blue crab population, (c), oysters, (d) average condition of lakes in the respondent's state (water color, visibility, and fish population composition), and (e) an unavoidable, permanent cost of living increase for the respondent's household. One follow-up question to the choice experiments is modified to accordingly reference these attributes (question 13).  

(II)	Respondent activities
EPA expects individuals to engage in the following activities during their participation in the valuation survey:  
   * Review the background information provided in the beginning of the survey document.  
   * Complete the survey questionnaire and return it by mail.
A typical subject participating in the mail survey is expected to take 25 minutes to complete the survey.  These estimates are derived from focus groups and cognitive interviews in which respondents were asked to complete a survey of similar length and detail to the current survey.


5.	The Information Collected - Agency Activities, Collection Methodology, and Information Management
5(a)	Agency Activities

The survey is being developed, conducted, and analyzed by EPA's National Center of Environment Economics and Abt Associates Inc. and is funded by EPA contract No. EP-W-11-003, which provides funds for the purpose of analyzing the economic benefits of the Chesapeake Bay TMDL. Agency activities associated with the survey consist of the following: 
   * Developing the survey questionnaire and sampling design.
   * Randomly selecting survey participants from the U.S. Postal Service DSF database. 
   * Printing of survey.
   * Mailing of preview letter to notify the household that it has been selected.
   * Mailing of surveys.
   * Mailing of postcard reminders.
   * Resending the survey to households not responding to the first survey mailing.
   * Mailing the follow-up letter reminding households to complete the second survey mailing.
   * Conducting a follow-up study of non-respondents to the mail survey using a combination of telephone and priority mailing to reach nonrespondents.
   * Data entry and cleaning.
   * Analyzing survey results.  
   * Analyzing the non-response study results.
   * If necessary, EPA will use results of the non-response study to adjust weights of respondents to account for non-response and minimize the bias.
Although not covered under this ICR, EPA will primarily use the survey results to estimate the social value of changes in ecosystem quality, as part of the Agency's analysis of the benefits of the TMDL.  

5(b)	Collection Methodology and Information Management

      To develop the survey questionnaire, EPA conducted a series of 10 focus groups and 26 cognitive interviews (OMB control # 2090-0028).  Focus groups provided valuable feedback which allowed EPA to iteratively edit and refine the questionnaire, and eliminate or improve imprecise, confusing, and redundant questions.  Focus groups and cognitive interviews were conducted following standard approaches in the literature, as outlined by Desvousges et al. (1984), Desvousges and Smith (1988), Johnston et al. (1995), Schkade and Payne (1994), Kaplowicz et al. (2004), and Opaluch et al. (1993).  
      EPA plans to implement the proposed survey as a mailed choice experiment questionnaire.  First, EPA will use the U.S. Postal Service DSF database to identify households which will receive the mail questionnaire.  Prior to mailing the survey, EPA will send the selected households a preview letter notifying them that they have been selected to participate in the survey and briefly describing the purpose of this study.  The mail survey will be mailed one to two weeks after the preview letter accompanied by a cover letter explaining the purpose of the survey. The preview and cover letters are included as Attachments 7 and 8, respectively.
      EPA will take multiple steps to promote response. All households will receive a reminder postcard approximately one week after the initial questionnaire mailing. The postcard reminder is included as Attachment 9. Approximately three weeks after the first round of survey mailing, all households that have not responded will receive a second copy of the questionnaire with a revised cover letter (see Attachment 10).  A week after the second survey is mailed, a letter will be sent to remind households to complete the survey.  The letter reminder is included as Attachment 11.  Based on this approach to mail data collection, it is anticipated that approximately 30 percent of the selected households will return the completed mail survey.  Since the desired number of completed surveys is 2,400, it will be necessary to mail surveys to 8,000 households (Dillman 2008).
      Data quality will be monitored by checking submitted surveys for completeness and consistency, and by asking respondents to assess their own responses to the survey.  Question 16 asks respondents to rate their understanding of the choice experiment descriptions.  Questions 13, 14, and 15 are designed to assess the presence or absence of potential response biases by asking respondents to indicate their reasoning and rate the influence of various factors on their responses to the choice experiment questions.  Responses to the survey will be stored in an electronic database.  This database will be used to generate a data set for a regression model of total values for ecosystem improvements achieved under the Chesapeake Bay TMDL.
      To protect the confidentiality of survey respondents, the survey data will be released only after it has been thoroughly vetted to ensure that all potentially identifying information has been removed.  

5(c)	Small Entity Flexibility

      This survey will be administered to individuals, not businesses.  Thus, no small entities will be affected by this information collection.

5(d)	Collection Schedule

The schedule for implementation of the survey is shown in Table A2.  
Table A2: Schedule for Survey Implementation
                              Pretest Activities 
                           Duration of Each Activity
Printing of questionnaires 
                                                                  Weeks 1 to 3 
Mailing of Preview Letters
                                                                         Week 4
Mailing of survey
                                                                         Week 5
Postcard reminder (one week after initial survey mailing)
                                                                         Week 6
Initial Data Entry and Pilot Tests
                                                                          Week7
Mailing of 2[nd] survey to non-respondents
                                                                         Week 9
Letter reminder (one week after 2[nd] survey mailing)
                                                                        Week 10
Telephone non-response interviews
                                                                 Weeks 12 to 13
Ship priority mail non-response survey
                                                                        Week 12
Data entry
                                                                  Weeks 5 to 15
Cleaning of data file
                                                                        Week 16
Delivery of data
                                                                        Week 17
                          Full Survey Implementation
                                                                               
Printing of questionnaires 
                                                                Weeks 22 to 24 
Mailing of Preview Letters
                                                                        Week 25
Mailing of survey
                                                                        Week 26
Postcard reminder (one week after initial survey mailing)
                                                                        Week 27
Initial Data Entry and Pilot Tests
                                                                         Week28
Mailing of 2nd survey to non-respondents
                                                                        Week 29
Letter reminder (one week after 2nd survey mailing)
                                                                        Week 30
Telephone non-response interviews
                                                                 Weeks 32 to 33
Ship priority mail non-response survey
                                                                        Week 32
Data entry
                                                                 Weeks 26 to 36
Cleaning of data file
                                                                        Week 37
Delivery of data
                                                                        Week 38




6.	Estimating Respondent Burden and Cost of Collection
6(a)	Estimating Respondent Burden

      Subjects who participate in the survey and follow-up interviews will expend time on several activities.  Based on the administration of the mail survey to 8,000 households, the national burden estimate for all respondents is 1,008 hours assuming that 2,400 respondents will complete and return the survey. Based on pretests conducted in focus groups, EPA estimates that on average each respondent mailed the survey will spend 25 minutes reviewing the introductory materials and completing the survey questionnaire.  Thus, the average burden per respondent is 25 minutes (0.42 hours) for these 1,500 respondents to the mail survey.  
      EPA plans to conduct a non-response follow-up study that uses a short questionnaire and a combination of telephone and priority mailing. The short version of the questionnaire is included in Attachment 14. The short questionnaire will be administered by phone to 200 nonrespondents and by priority mail to 200 nonrespondents. EPA estimates that telephone non-response interviews will take 5 minutes (0.08 hours) per interview for each of the 400 households completing interviews.  EPA estimates that each of the 400 households completing the mail version of the short questionnaire will take 5 minutes to do so (0.08 hours). Thus the national burden estimate for all non-response survey respondents is 32 hours, or an average burden per respondent of 5 minutes (0.08 hours) for the 400 total participants in the non-response survey.
      These burden estimates reflect a one-time expenditure in a single year.

6(b)	Estimating Respondent Costs

According to the Bureau of Labor Statistics, the average hourly wage for private sector workers in the United States is $23.33 (2012$) (U.S. Department of Labor, 2012).  Assuming an average per-respondent burden of 0.42 hours (25 minutes) for individuals mailed the survey and an average hourly wage of $23.33, the average cost per respondent is $9.80.  Of the 8,000 individuals receiving the mail survey, 2,400 are expected to return their completed survey.  The total cost for all individuals that return surveys would be $23,517.
Assuming an average per-respondent burden of 0.08 hours for each of the 400 total participants in the non-response study and an average hourly wage of $23.33, the average cost per non-response study participant is $1.87.  Therefore the total cost to participants in the non-response study phase would be $747.  
      EPA does not anticipate any capital or operation and maintenance costs for respondents.

6(c)	Estimating Agency Burden and Costs

      This project will be undertaken by Abt Associates Inc. with funding of $398,779 from EPA contract EP-W-11-003, which provides funds for the purpose of analyzing the economic benefits of the Chesapeake Bay TMDL.  Abt Associates Inc. staff is expected to spend 2,295 hours pre-testing the survey questionnaire and sampling methodology, conducting the mail survey, conducting the non-response survey, and tabulating and analyzing the survey results. The cost of this contractor time is $398,779.  In addition to the effort expended by EPA's contractors, EPA staff is expected to spend 4,500 hours managing and reviewing this project and contributing to the analysis at a cost of $336,000.  Agency and contractor burden is 6,995 hours, with a total cost of $734,779 excluding the costs of survey printing
       and mailing.  Printing of the survey is expected to take 226 hours and cost $170,899.  Thus, the total Agency and contractor burden would be 7221 hours and would cost $905,678.

6(d)	Respondent Universe and Total Burden Costs

      EPA expects the total cost for survey respondents to be $15,445 (2012$), based on a total burden estimate of 662 hours (across both primary survey administration and non-response follow up surveys) at an hourly wage of $23.33.

6(e)	Bottom Line Burden Hours and Costs

The following table presents EPA's estimate of the total burden and costs of this information collection:

Table A3: Total Estimated Bottom Line Burden and Cost Summary
                             Affected Individuals
                                Burden (hours)
                                 Cost (2012$)
Mail Survey Respondents
                                     1,008
                                    23,517
Total for Survey Respondents
                                     1,040
                                    24,263
Priority Mail Non-Response Subsample
                                      16
                                      374
Telephone Non-Response Subsample
                                      16
                                      373
EPA Staff
                                     4,500
                                    336,000
Survey Printing and Mailing
                                      226
                                    170,899
EPA's Contractors for the Mail Survey
                                     6,995
                                    734,779
Total Burden and Cost
                                 13,801 hours
                                  $1,290,205

6(f)	Reasons for Change in Burden

The survey is a one-time data collection activity.

6(g)	Burden Statement

EPA estimates that the public reporting and record keeping burden associated with the mail survey will average 0.42 hours per respondent (i.e., a total of 1,008 hours of burden divided among 2,400 survey respondents).  Households included in the non-response study are expected to average 0.08 hours per screening interview participant (i.e., a total of 32 hours of burden divided among 400 non-response study participants).  This results in a total burden estimate of 1,040 hours including both the mail survey and non-response study.  Burden means the total time, effort, or financial resources expended by persons to generate, maintain, retain, or disclose or provide information to or for a Federal agency.  This includes the time needed to review instructions; develop, acquire, install, and utilize technology and systems for the purposes of collecting, validating, and verifying information, processing and maintaining information, and disclosing and providing information; adjust the existing ways to comply with any previously applicable instructions and requirements; train personnel to be able to respond to a collection of information; search data sources; complete and review the collection of information; and transmit or otherwise disclose the information.  An agency may not conduct or sponsor, and a person is not required to respond to, a collection of information unless it displays a currently valid OMB control number.  The OMB control numbers for EPA's regulations are listed in 40 CFR part 9 and 48 CFR chapter 15. 
      To comment on the Agency's need for this information, the accuracy of the provided burden estimates, and any suggested methods for minimizing respondent burden, including the use of automated collection techniques, EPA has established a public docket for this ICR under Docket ID No. EPA-HQ-OA-2012-0033, which is available for online viewing at www.regulations.gov, or in person viewing at the Office of Water Docket in the EPA Docket Center (EPA/DC), EPA West, Room 3334, 1301 Constitution Ave., NW, Washington, DC.  The EPA/DC Public Reading Room is open from 8:30 a.m. to 4:30 p.m., Monday through Friday, excluding legal holidays. The telephone number for the Reading Room is 202-566-1744, and the telephone number for the Office of Water Docket is 202-566-1752.  
      Use www.regulations.gov to obtain a copy of the draft collection of information, submit or view public comments, access the index listing of the contents of the docket, and to access those documents in the public docket that are available electronically.  Once in the system, select "search," then key in the docket ID number, EPA-HQ-OA-2012-0033. 

