Killing the National Endowment for the Arts: Race, Income, and Arts Participation

President Trump recently announced plans to cut funding for the National Endowment for the Arts (NEA). The NEA is currently the largest single source of funding for the arts in America, but makes up a small fraction of total funds made available to arts nationwide. The pie chart below represents all of the money allocated to the arts in a given year. The red silver represents money from the NEA. Tiny!

Duplication of a graph from the publication: “How the United States Funds the Arts” (NEA, 2012)

Despite being overshadowed by large pools of money from other sources, advocates of the NEA argue that the funds it makes available have a huge impact on how Americans interact with the arts. The benefits of the NEA are said to be especially important for Americans who live in areas that otherwise would not have enough money to afford arts programs. The Metropolitan Museum of Art in NYC, for example, has a multi-billion dollar endowment and will never be in danger of closing its doors even if the NEA is completely destroyed. In contrast, the relatively small grants the NEA provides seem to be the biggest single source of money for many smaller institutions around the country.

Let’s take a look at the NEA’s funding over time, starting in 1980. The most dramatic changes here happened in 1996 and 2009, when Republican and Democratic actors cut and increased funding, respectively.

In 1994, then-Speaker of the House Newt Gingrich tried to eliminate the NEA altogether. Funding in this graph has been adjusted for inflation to 2017-dollars.

To go with the NEA budget data, I went through interviews conducted by the National Archive of Data on Arts & Culture (NADAC). Every five years, NADAC collects data on large groups of Americans on their art habits and preferences. And then they make the data available to the public! In all, I used data on 97,295 Americans from six interview periods taken over 30 years: in 1982, 1985, 1992, 2002, 2008, and 2012.1 Comparing the data from these interviews to the NEA budget over time should reveal some relationship between NEA funding and arts participation (if there is one).2

The NADAC interviews have information about seeing live jazz, classical music, opera, musical theater, non-musical theater, ballet, and art museums. I’m not going to analyze the whole interview, but will focus on the questions that asked “did you go to [art activity] during the last 12 months?” I’ll also include some of the demographic information recorded about the respondents.

I’ll be the first to admit that this is an incredibly complicated question. Many factors other than federal funding can affect whether or not a person engages with the arts. One that seems too important to ignore is household income. When you’re struggling to pay rent, it’s hard to justify spending a bunch of money on a concert or at a theater. This is especially true if you have kids. As you can see in the plots below, there’s a pretty wide variety of household sizes and incomes in the NADAC interviews.


Another variable that might change the way federal funding affects arts participation is race. A substantial portion of NEA money each year is awarded through grants to traditionally underserved racial groups, so stopping the flow of this money might have a disproportionate impact there (one of the “focus areas” for NEA grant awards is racial justice). For this blog post, I will be comparing White and Black Americans. Race is a complicated issue that honestly I don’t feel very qualified to comment on, but I’ll say that one reason I chose to focus on these two groups in particular was because “White” and “Black” were the only races recorded consistently across the entire 30-year period of NADAC interviews. What I mean is, the first two surveys—1982/1985—only asked if you were White or if you were Black. The 1992 survey added two more categories, Asian and Native American, and the 2002 survey added 16 more.3

To analyze the relationship between NEA funding, race, income, household size, and participation in the arts, I’ll use a procedure called mixed-effects regression. Basically, it accounts for the fact that people responding within the same survey year might be more similar to each other than they are to people from different years. Yadda yadda yadda, right? If you’re curious to learn more, check out this lecture on YouTube from my undergraduate thesis advisor, John Nezlek, he does a good job of explaining the logic behind it.

Anyway, let’s get to the data. In each interview, people were asked seven separate questions about seeing art: if they had seen live jazz, classical music, opera, musical theater, non-musical theater, ballet, or been to an art museum in the past 12 months. “Yes” responses were coded as 1 and “No” responses were coded as 0. First, I’m going to reduce the seven arts questions into a single question by giving people a score of “1” if they reported seeing at least one type of art asked about in the survey. So, the question I’ll be graphing first is essentially: “did you go to any kind of art activity during the last 12 months?”

Let’s ignore income and racial groups for now and just figure out the overall effect of NEA funding. From the lowest funding level to the highest we see a 6 percentage point increase in people who reported seeing some kind of art in the past year, from 33% to 39%.

The effect of funding on arts participation controlling for art type, income, and race. Error bars represent standard errors.

Although the change from 33% to 39% represents a 17% increase in overall arts participation, it was not a “statistically significant” increase (p=.129). This means that although on average participation increased when NEA funding went up, there were many people for whom more funding did not result in more arts participation. The fact that NEA funding helped some Americans but not others suggests that there might be meaningful sub-groups in the effect (i.e., an interaction). Next we’ll see if the effect of NEA funding depends on: 1) income, and 2) race.

Let’s delve deeper and make this graph more complicated. I drew separate lines for White and Black Americans, and split the graphs for the 25th and 75th percentiles of household income ($27,766/year and $92,838/year). I included household size as a control variable in all of the analyses, so what you see below is the effect of household income independent of household size.


This plot shows us a few things:

  1. Rich Americans are more involved in the arts than poor Americans are (p<.001);
  2. White Americans are more involved in the arts than Black Americans are (p<.001);
  3. NEA funding increases arts participation more for Black Americans than it does for White Americans: you can see that both of the Black lines have steeper slopes than the White lines (p=.004).

In my mind, this is pretty good evidence that even a tiny amount of funding (relatively speaking) can make a difference for Americans who might not otherwise have the opportunity to become involved in the arts.

I’ve also included the graphs split by art type in a slideshow below if you want to see how the effect might differ for jazz, classical music, opera, musical theater, non-musical theater, ballet, and art museums. Interestingly, Jazz is the only type of art (on the survey) that Black Americans are more likely to participate in than White Americans.

This slideshow requires JavaScript.

Thanks for reading!

Want more data visualization? Check out my other posts at:


  1. NADAC also collected data in 1997, but this dataset was excluded per their recommendation, because “the survey design for the 1997 [interview] was quite different from the other six SPPA studies.”
  2. I calculated the three-year NEA budget average before each NADAC survey to examine budget vs. art participation for each interview period (e.g., 1980, 1981, and 1982 budget years were averaged so they could be lined up with the 1982 NADAC interview). I did this because my assumption is that it takes longer than a single year for arts funding to “get the ball rolling,” in other words, that there’s a lag-time between funding and arts participation.
  3. The 16 race options added in 2002 accommodated multi-racial individuals, and were: White-Black, White-Al, White-Asian, White-Hawaiian, Black-Al, Black-Asian, Black-HP, Al-Asian, Asian-HP, W-B-Al, W-B-A, W-Al-A, W-A-HP, W-B-Al-A, 2 or 3 Races, and 4 or 5 Races.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s