Looking for clear and concise—yet sophisticated—editing? Need to comply with plain language requirements?

The American Community Survey (ACS) is the largest continuous household survey in the United States, providing a wealth of information about the economic, social, and demographic characteristics characteristics features of personspersonspeople, as well as housing characteristicscharacteristicsfeatures. One goal of the ACS is to serve as a cost-effective vehicle for collecting information required by law, regulation, or executive order on behalf of Federal agencies or at the direction of Congress. In conducting the ACS, the U.S. Census Bureau leverages its existing infrastructure and data collection expertise expertise knowledge on behalf of the Federal Government. Like the decennial census long form before it, the ACS provides comprehensive and representative statistical information about large and small communities on a variety of important topics-all while protecting respondents' confidentiality. In addition, because the ACS is a continuous survey, the data is published annuallyannuallyyearly, providing more timely information than the once-per-decade long form.

The value of the ACS lies in its ability to provide statistical information representing large and small communities and demographic groups across the Nation that can be used by governments, businesses, and individuals to make better decisions. This ability depends upon depends upon depends on sufficiently large samples from these communities, which in turn in turn depends on the mandatory mandatory required nature of the survey. About 3.5 million households are sampled every year for the American Community Survey, and it takes respondents a total of a total of about 2.4 million hours per year to answer the survey. In exchange, the Nation acquires uniquely comprehensive, representative, topical, and timely data about its communities that serves as an essential component component part of the information infrastructure relied upon relied upon relied on by decision makers through­ out the country.

What would happen to the quality of decision making if the ACS were less comprehensive, representative, topical, or timely? The answer depends upon depends upon depends on the specific use to which the data would be put and require speculation about what, if any, , if any, data would be used instead. Still, Canada's recent experience in moving from a mandatory to a voluntary survey provides some insight into the value of a comprehensive, representative, topical, and timely survey, and what happens when those qualities are compromised. After it became voluntary, the Canadian survey yielded much lower response rates, resulting in information that less comprehensively and reliably covered the small areas of Canada-and at a higher cost. Canadian decision makers were left to use out-of-date data that did not necessarily represent the specific geographic areas in question.

ACS data is uniquely tailored to help guide myriad specific decisions across myriad specific decisions across many specific decisions across the country, making government more intelligent, businesses more competitive, and the public more informed. As a tool for smarter government, ACS data offers evidence to help government operate betteroperate betterrun better. Federal agencies rely on the ACS to help them make operational decisions, including managing and evaluating programs, allocating evaluating programs, allocating evaluating programs , assigning over $400 billion in funds, determining eligibility for pro­ grams, and benchmarking other statistics. For example, the Department of Veterans Affairs (VA) uses ACS data to estimate the number and geographic distribution of veterans with service-connected disabilities who are eligible but not registered for VA health services.

State and local communities use the wealth of information provided by information provided by information from the ACS for a wide variety of purposesa wide variety of purposesmany purposes, including comprehensive planning, economic development, and research on local issues and conditions. For example, a researcher generated maps of vulnerable populations, such as students, households in poverty, and persons persons people with disabilities in the path of a lava flow on the island of Hawaii, using small-area population and demographic estimates from the ACS. And the Greater Houston Partnership, which works with companies to retain retain keep and create jobs, make investments to expand the tax base, and increase general business activity and local incomes, uses ACS data to answer companies' questions about such things as commuting times and the availability of science and engineering workers.

The ACS provides immense value beyond government. Often, intermediaries combine ACS data with other data, analyze it, tabulate it, use it in models, and display it in charts, maps, and tables to help decision makers make better use of make better use of use it. As a tool for more competitive businesses, ACS data provides reliable information to help businesses make a wide variety of decisionsa wide variety of decisionsmany decisions, from where to locate a store to what to stock on the shelves. Locating a business is a critical decision, with more than $55 billion spent on commercial construction last year. Businesses, such as Kroger, rely on third-party proprietary models that use ACS small area estimates to project sales for potential new grocery stores sites or remodeling of existing stores. Likewise, with retail inventories of over $563 billion last year, businesses need solid evidence about how to stock their shelves. Retailers, such as Target, use the ACS for neighborhood-level demographic data such as population density, owner-occupancy, and house­ hold size to determine the optimal the optimal the best mix of goods with which to stock its stores throughout the country. This report provides other examples of how ACS data has served as an important component component part of the information that businesses use to stay competitive-helping them run efficiently, hire wisely, and serve their customers' needs. And as a tool for a more informed public, ACS data is used by researchers and the media to advance the Nation's understanding of a broad range of issues. This report discusses examples of these uses as well.

No other source of data offers what the ACS can: timely information about important topics, comprehensively covering communities throughout the Nation with the statistical precision to represent them accurately and to compare them. Without the ACS, decision makers throughout the country would be left to make decisions based make decisions based decide based on information that is, relative to the ACS, stale, incomplete, unreliable, or not as relevantnot as relevantnot.

The American Community Survey (ACS) collects information on a large number of economic, social, and demographic characteristics characteristics features of personspersonspeople, as well as housing characteristicscharacteristicsfeatures. The survey is conducted by collecting data from a random sample of housing unit addresses every month, for a total of 3.5 million housing units contacted each year. The large sample size affords comprehensive coverage of the Nation and permits statistically reliable estimation for small and large geographic areas. The ACS program is conducted by the U.S. Census Bureau and provides statistics about thousands of communities across the Nation on a wide range of subjects.

To ensure that ensure that make sure all communities are adequately rep­ resented, respondent participation is mandatorymandatoryrequired. The confidentiality of responses is protected by criminal and civil penalties for illegal disclosure. Together these aspects aspects parts of the ACS help to ensure the data's quality and credibility.

A Brief History of the American Community Survey

The history of the ACS is rooted in the decennial census. The first census, conducted in 1790, had six demographic questions that went beyond the bare enumeration of the Nation's population. Over time, and often at the behest of Congress, more questions about social, economic, demographic, and housing characteristics characteristics features were added. The development of statistical sampling theory allowed the Census Bureau to collect this detailed data from only a random sample of the population, and from the 1940 Census through the 2000 Census, most households were asked to pro­ vide responses to a short set of questions (the "short form"), while a random sample was asked a longer set of questions (the "long form"). In the 2000 Census, the long form was delivered to about 17 percent of addresses, or roughly 20 million households.

After the 1990 Census, the Census Bureau began to explore alternatives to the long form, with the goals of simplifying the decennial census, containing costs, and producing timelier information to inform policy decisions and legislative actions. In 2005, after nearly a decade of rigorous testing and experimentation, the Census Bureau fully fully implemented the alternative to the long form, the American Community Survey.

Figure 1 shows the timeline for the addition of for the addition of for adding various topics to the decennial census starting in 1790 through 2000 and to the ACS since it was first fully fully implemented in 2005. While the first several decennial censuses asked a small number of a small number of a few basic demographic questions (about age, race, and sex), as the Nation expanded geographically and as its population and economy grew in both size and complexity, more questions were added.

This report aims to illustrate the value of the information provided the information provided the information given by the ACS by examining how it supports decision making throughout the Nation. The remainder remainder rest of this report is organized as follows. The first section discusses the ACS's role in providing a broad array of a broad array of an array of information to represent comprehensively the Nation's communities. The collection of this information has been driven by the Federal Government's need for information, and the value of the ACS lies in the fact that it provides a broad range of current and topical information about large and small communities. The section includes discussion about the types of types of questions asked, confidentiality protections, limitations of the survey, and how the ACS content can be revised to meet the Nation's changing information needs.

The second section of the report focuses on what the Nation gets in return for its investment in the ACS. It provides examples of how ACS data is used to inform decisions across the government and the economy. Although the information needs of the Federal Government are paramount to the mission of the ACS, the section also discusses the importance of the ACS in informing a wide range of state and local community decisions, and how businesses rely on the ACS to improve their operations. Finally, this section discusses the value of the ACS in research and in informing citizens about their communities.

The value of the ACS is that it provides statistically precise information that is topical, comprehensive, representative, and timely.

The ACS provides topical information, driven by the Federal Government's need for information to operate effectively. The ACS questionnaire2 contains contains has 48 questions about each person in a given given residence and 24 questions about the housing of about the housing of about housing the respondent. The information it collects includes individual characteristics characteristics features such as age, sex, race, ethnicity, ancestry, citizenship, educational attainment, veteran status, disabilities, and language; household structure, including family relationships; employment status, industry of employment, occupation, job location, and commuting time; earnings and other income; health insurance; housing characteristicscharacteristicsfeatures, such as number of rooms and renter/owner-occupied, and expenditures expenditures spending on certain goods and services, such as rent, utilities, and homeowner costs.

The ACS is comprehensive, providing a broad array of a broad array of an array of information covering the entire Nation. The ACS is also representative geographically, providing statistically valid information for states, counties, and other large and small areas, such as cities, townships, and villages, congressional and state legislative districts, American Indian and Alaska Native areas and Hawaiian home lands, zip code tabulation areas, and school districts. The ACS is also representative of is also representative of represents many small distinct populations, for example, reporting 2009-2013 estimates of per capita income by race and Hispanic origin for Crenshaw County, Alabama (2014 estimated population: 13,977). And ACS estimates are timely-more current than information based on the once-per-decade Census long form. Because ACS information is based on samples collected continuously each year, new estimates are released annuallyannuallyyearly.

The Census Bureau annually publishes estimates based on data covering different time time periods for different geographic areas, based on the population size of the areas. Estimates for areas with populations of at least 65,000 are available based on 12 months of collected data {1-year estimates). In contrast, estiIn contrast, estiEsti­ mates for areas with smaller populations are released annually annually yearly but are based on 60 months of collected data. Likewise, estimates are released annually annually yearly for small subpopulations in low-population areas, such as American Indian areas, Alaska Native areas, Hawaiian home lands, and racial or ethnic groups, based on 5 years of accumulated accumulated collected data.4 Although the use of pooled multiyear data for smaller areas and groups introduces some some uncertainty into the estimates, annual releases of data allow users to com­ pare data more frequently frequently often than the decennial census would allow. Also, because the ACS uses a consistent methodology a consistent methodology a consistent method to produce estimates, data users can make direct comparisons for a wide range of groups and communities. Such comparisons would be more difficult if communities or other entities independently collected their own data using different methods, defini­ tions, sources, and standards.

American community survey costs and Burdens

To provide statistically valid information about both small and large areas and groups, a sufficient number of a sufficient number of enough completed questionnaires are required, and this drives the cost of the survey. The cost of the ACS includes both the costs to taxpayers, as reflected in the ACS budget, and the time that individuals spend responding to the survey.

The budget for the ACS was $242 million in 2013, representing an annual cost of about 76 cents per per­ son in the United States. In addition, each respondent spends an estimated average of 40 minutes answering the survey, for a total burden of about 2.4 million hours per year.5 Based on average hourly earnings in the private sector of roughly $24 in 2013,6 the estimated value of respondent time spent on the ACS in 2013 was approximately approximately about $58 million.

Another perceived burden of the ACS is concerns about privacy and confidentiality. Some respondents have concerns about divulging information required by some of the survey questions. Therefore, Therefore, So informa­ tion provided by respondents is kept confidential by law.7 The information on individuals and households is aggregated to create statistics about the Nation's com­ munities, devoid of personally identifiable information. Individual-level data is available for certain additional statistical tabulations, but without any without any with no information that would allow identification of respondents. The Census Bureau cannot share respondents' answers-including with other government agencies. Violators are subject to fines and incarceration; disclosure of information that could identify respondents can result in up to 5 years in prison, or $250,000 in fines, or both.

The ACS is a mandatory mandatory required collection of information for statistical purposes. The mandatory mandatory required nature of the survey ensures that ensures that makes sure the sample sizes for all areas are sufficient are sufficient are enough for reliable statistics. It also helps to achieve this reliability while minimizing minimizing reducing the cost to taxpayers.

To maintain a sufficiently large sample under a voluntary survey scenario, the Census Bureau would need to increase efforts at following up with respondents who choose not to respond initiallyrespond initiallyrespond, and would need to administer additional additional more questionnaires to compensate for the increase in nonrespondents. This would take more time and more taxpayer money.

Prior to Prior to Before the full implementation of the ACS, and at the request of Congress, in 2003 and 2004 the Census Bureau studied the impact of a voluntary, rather than mandatorymandatoryrequired, ACS. The Census Bureau found that a voluntary ACS would decrease the response rate by about 20 percentage points and negatively affect the reliability of the survey estimates. As Table 1 shows, if in 2013 the ACS had been voluntary, and no additional measures were taken, there would be a notable loss in reliability of the survey results (53 percent increase in statistical variance); increasing the initial sample size in order in order to maintain the reliability of the survey in 2013 would have cost an additional $89 million (52 percent more than the actual cost). Simply maintaining the current ACS sample size would have cost an extra $36 million due to additional respondent follow-up and would have also resulted in a less reliable survey.8 Further, additional additional more nonresponse as a result of as a result of because of vol­ untary methods may be unevenly distributed across demographic, social, and economic subgroups, so that the representativeness and completeness of the data from the survey would be at risk. Figure 2 depicts the projected detrimental detrimental harmful effect that voluntary meth­ ods would have on the percentage of census tracts within a county with acceptable quality data, based on the 2006-2010 5-year American Community Survey results.9 In both panels darker-shaded counties have a higher share of census tracts with acceptable qual­ ity data than lighter-shaded counties. The top panel depicts the share of census tracts with various levels of acceptable data quality under a mandatory mandatory required survey. It shows that only about 5 percent of counties have the lowest level of quality (i.e., the share of tracts with acceptable quality data is less than 20 percent) cov­ ering only about 2.5 million people. The top panel also shows that about 42 percent of counties covering 55 million persons persons people have more than 80 more than 80 over 80 percent of tracts with acceptable data quality. In contrast, the In contrast, the The bottom panel shows that under a voluntary survey, 24 percent of counties covering 61 million persons persons people would have less than 20 percent of tracts with acceptable data quality, while 17 percent of counties covering about 9 million persons persons people have 80 percent or more of their tracts with acceptable data quality. Finally, a less reliable ACS would also be detrimental to be detrimental to harm other Census Bureau products, since the ACS is used as used as a benchmark for surveys and other populations and statistical reporting purposes.

In addition to the Census Bureau research, the expe­ rience of Canada shows that response rates decline when shifting from a mandatory to a voluntary survey. In Canada, the response rate to a survey similar to the ACS fell from 94 percent in the last year it was mandatory to 64 percent in the first year it was voluntary, a 25 percentage point drop.10 (See Box 5. ) Given the lower response rates, the mandatory mandatory required nature of the survey is fundamental to its cost-effectiveness as a comprehen­ sive and representative source of data.