User’s Guide to the 2020 Arts Basic Survey

About the 2020 Arts Basic Survey

The Arts Basic Survey (ABS) is a nationally representative survey of American arts participation. It is conducted by the U.S. Census Bureau, on behalf of the National Endowment for the Arts, as a supplement to the Current Population Survey (CPS). For a more detailed description of the survey, see the Arts Basic Survey 2020 study homepage. Please note that while this guide specifically addresses the 2020 survey, the information is generally applicable to prior years as well.

The 2020 ABS can be used to estimate a wide variety of arts participation. A sample of the activities covered by the survey include:

From Part A of the ABS:

  • Working with pottery or ceramics;
  • Doing weaving, crocheting, or creating other textile arts;
  • Playing a musical instrument;
  • Performing or practicing any acting, singing, or dance;
  • Taking photographs as an artistic activity;
  • Doing creative writing.

From Part B of the ABS:

  • Attending live performing arts events;
  • Going to art exhibits;
  • Going to movie theaters;
  • Reading literature (e.g., novels and short stories, poetry, and plays);
  • Using devices (e.g., TV, radio, laptop computers, mobile devices) to watch, listen, or download performances or information about the arts;
  • Taking arts classes such as classes in drawing, painting, and fashion design or in film making.

Survey Design

Respondents to the 2020 ABS were divided into two groups.

The first group answered questions (part A) about their personal performance and creation of art, including whether they shared the art they created, and the second group of respondents answered questions (part B) about live arts attendance, literary reading, consumption via media, and whether they had taken art classes.

No ABS respondent answered questions from both sections of the 2020 survey. Therefore, the ABS cannot be used to compare variables in one part of the survey with those from the other. For example, the ABS cannot be used to estimate the share of people who attend live performing arts events (part B of the survey), given that they personally play a musical instrument (part A of the survey).

Additionally, the ABS uses spouse proxies, which means the respondents were asked not only about their own participation, but also about whether their spouses/partners (if applicable) participated.

Including spouse proxies, each part of the 2020 ABS contains nearly 17,300 cases.

Using ABS Survey Weights and Calculating the Number of People Participating in the Arts

The ABS is a probability survey that requires the application of weights in order to obtain unbiased estimates.

The 2020 ABS weight variable is "PWABSWGT."

Notably, estimated counts generated using the ABS weight result in four implied decimal places. For example, the weighted estimate of the number of people who play a musical instrument in the 12 months leading up to February 2020 is 236,580,000,000.

This estimate, however, requires division by 10,000 in order to adjust for the four implied decimal places.

Consequently, the survey shows that 23.7 million people (or 9.6 percent of adults aged 18 and older) reported playing a music instrument during the period studied.

The practice of weighted counts containing implied decimal places is common in data put out by the U.S. Census Bureau and data providers. The origin of the practice has to do with some versions of statistical software packages not accurately producing estimates when the weight values themselves are fractional. Please read documentation carefully when using weighted data.

Provided the survey questions are worded and placed similarly within the survey, the ABS—and, to some extent, its long-form counterpart, the Survey of Public Participation in the Arts—can be used to examine trends in the share of adults participating in the arts.

It is not, however, recommended that the ABS or SPPA be used to estimate trends in the number of people participating in the arts.

Prior to 2020, the most recent ABS had been conducted in 2016. According to that survey, 9.0 percent of adults played a musical instrument in the 12 months leading up to February 2016. By comparing rates, we see that the percentage of U.S. adults who play a musical instrument was virtually unchanged between 2016 and 2020.

But, due to factors such as population change, the number of U.S. adults playing musical instruments in 2020 should not be compared with the number who played in 2016.

Illustration Using 2020 ABS Data

Using 2020 ABS data, and the SDA tool available in NADAC, this example will illustrate the calculation of the share and number of adults who play a musical instrument for the following age groups: 18-24; 25-34; 35-44; 45-54; 55-64; and 65 and older.

The two variables from the 2020 ABS we will use are: "INSTRUMENT" (a composite variable of the respondents' reply and his or her spouse/partner proxy reply, if applicable) and "PRTAGE" (person’s age, top-coded at 85).

Steps to Generate Table Estimates

  1. Select "SDA Frequencies/Crosstabulation Program"
  2. In "ROW" enter: PRTAGE(r: 18-24; 25-34; 35-44; 45-54; 55-64; 65-85)
    (Note: This groups the PRTAGE values into categories, which enables us to calculate participation rates and counts for specific age groups.)
  3. In "COLUMN" enter: INSTRUMENT
    (Note: This is the composite variable that relates to the number of respondents who played a music instrument in the last 12 months.)
  4. In "WEIGHT" enter: PWABSWGT
    (Note: This is the weight needed to produce unbiased estimates from the 2020 ABS. It has four implied decimal places.)
  5. In TABLE OPTIONS, "Percentaging," check the box for Row.
    (Note: Select chart options to suit your preferences. Here we’ve chosen row percentages because age is more likely to be an independent variable – that is, you want to know whether the percentage of people who play an instrument is different for various age groups. Choosing column percentages for this table would tell you what percent of people (not) playing instruments falls into a particular age group. You can also add a title to your table, for example, Percent of U.S. adults who play a musical instrument: 2020)

The table resulting from the SDA tool is shown below. It shows, for example, that the share of young adults aged 18-24 who played a musical instrument in 2020 was 11.6 percent.

The number playing a musical instrument is 33,239,711,974. However, because the ABS weight contains four implied decimal places, this figure should be divided by 10,000.

Division, in turn, yields 3,323,971, the number of adults aged 18-24 who reported playing a musical instrument in 2020.

The last row in the table shows that 9.6 percent of all U.S. adults (23.7 million people) played a musical instrument in the 12 months ending February 20.

Illustration Using 2020 ABS Data

Time Period Covered

The 2020 Arts Basic Survey covers the period during the 12 months ending in February 2020, before the COVID-19 pandemic.

Resources

Technical documentation for the 2020 ABS is available from the U.S. Census Bureau.

In addition, the Office of Research & Analysis, National Endowment for the Arts, publishes Arts Data Profiles, collections of statistics, graphics, and summary results from data mining about the arts.

Free and easy access to data on the arts and on the arts' value and impact for individuals and communities