NBC News Online Survey: Public Opinion on Taxes for April 15

Embargoed for release after 5:00 AM Wednesday, April 15, 2015

Where Do We Want Our Tax Dollars to Go?

Education Tops the List
The latest NBC News Online Survey conducted by SurveyMonkey asked Americans a number of
questions about their taxes with the April 15 deadline for filing taxes just a week away,
including where they would like to direct their tax dollars to go.
Respondents were asked to choose up to three areas where they would like to direct the
federal government to spend their taxes. Education was named as a priority by 51 percent of
Americans, followed by health care at 37 percent and national defense at 30 percent
(Reminder: These numbers do not add up to 100 percent because respondents could select up to
three categories of spending).

Education

All

18 - 29

30 - 44

45 - 59

60+

Rep

Dem

Ind

51%

65%

63%

44%

37%

39%

60%

50%

Health care

37

35

37

40

36

23

52

33

National defense

30

13

23

39

42

57

18

25

27

39

31

22

19

15

29

32

Natural resources, energy, and

environment
Veterans benefits

27

19

23

31

33

38

20

27

Reducing the deficit

26

22

27

27

30

36

14

32

Assistance for low-income,

unemployed, and disabled

19

22

18

18

18

12

29

15

Immigration, law enforcement, and

administration of justice

16

15

16

23

30

16

14

21

15

11

11

11

11

19

13

12

12

12

16

19

11

10

12

10

10

11

Science, space, and technology

programs
Job training
Community and regional
development
Response to natural disasters

11

11

Agriculture

International affairs

Other (please specify)

Some notable differences in spending priorities:

Support for education was strongest among younger Americans nearly two-thirds of
those under 45 years old prioritized education, compared to less than half of those
between 45 and 59, and fewer than 4 in 10 of those over the age of 60. Among
Democrats and independents, education also rose to the top.

National defense was a higher priority for Republicans than education, with 57 percent
choosing it, compared to 39 percent who chose education.
Those over 45 years old were about 3 times more likely to prioritize national defense
than those under 30. Men were 10 points higher than women on national defense.
About 4 in 10 Americans had health care among their choices the second highest
overall category. And while age didnt make much of a difference, party identification
did: half of Democrats listed health care as a top-three priority, compared to a third of
independents and a quarter of Republicans.
Twice as many under the age of 30 prioritized natural resources, energy and
environment as those over 60; women were 10 points higher on spending on the
environment than men.
About a third of Republicans and independents wanted their tax dollars used to reduce
the deficit, compared to just 14 percent of Democrats.
Support for spending tax dollars on science, space, and technology was highest among
those under 30. Those over 60 were the most supportive of spending on immigration,
law enforcement, and administration of justice.

Although education ranked highest for Americans overall, education spending comes in 10th
place in terms of where actual federal taxpayer dollars go, according to the White House
website that shows how 2013 tax revenues were spent
(https://www.whitehouse.gov/2013-taxreceipt). Health care and national defense are the
largest two pieces of the federal budget pie -- nearly matching how Americans say they
would appropriate their taxes. However, the third largest source of government spending is
the so-called safety net programs, such as aid for low-income, unemployed, and those
with disabilities. That area ranked seventh among Americans choices for how their taxes
should be spent.
Americans Are Split on Tax Burden
Americans are divided as to whether or not they pay too much in taxes: 48 percent say they pay
too much in taxes and 47 percent said they pay their fair share. Just 4 percent say they pay
too little in taxes. Party identification plays a role, with a majority of Republicans and
independents saying they pay too much and a majority of Democrats saying they pay their fair
share.
How Many Procrastinators Are Out There?
The NBC News Online Survey was in the field the week before Tax Day; with a week left to file,
only about 3 in 10 Americans said they had not yet filed their taxes. Sixty percent said they
were already filed, while 20 percent said they expected to get them done before the 15 th. But 1
in 10 Americans were procrastinating as long as possible 4 percent said they would get them
done on Tax Day itself, and 6 percent said they would apply for an extension.

TOPLINES
Questions 1-7, 9-17, 20-23, 27-30 already released. Questions 8, 18-19, 31-35 held for future
release.
24. Income taxes for 2014 are due on Wednesday, April 15th, 2015. Have you already filed your
income taxes, or will you get them done before the 15th, on the 15th, or will you apply for an
extension?

4.6-8.15

Already
filed
60%

Will get them

done before
the 15th
20%

Will get
them done
on the
15th
4%

Will apply for

an extension
6%

Dont
file
10%

DK/NA
0%

25. In general, do you think you pay your fair share in federal taxes, pay too little, or pay too
much?
Pay my fair share Pay too little Pay too much DK/NA
4.6-8.15
47%
4%
48%
1%
26. Imagine that it was possible to direct the federal government as to how you wanted your
taxes spent. Among the choices below, select UP TO THREE areas where you most want to
see your tax dollars spent. [Randomize responses 1-13]

National defense
Health care
Assistance for low-income,
unemployed, and disabled
Education
Job training
Veterans benefits
Natural resources, energy, and
environment
International affairs
Science, space, and technology
programs
Immigration, law enforcement,
and administration of justice
Agriculture
Community and regional
development
Response to natural disasters
Reducing the deficit
Other (Please specify)
DK/NA

4.6-8.15
30%
37%
19%
51%
13%
27%
27%
2%
14%
16%
7%
10%
8%
26%
5%
-

METHODOLOGY
The NBC News-SurveyMonkey Poll was conducted online April 6-8, 2015 among a national
sample of 2,052 adults aged 18 and over. Respondents for this non-probability survey were
selected from among those who have volunteered to participate in the SurveyMonkey
Audience panel. This SurveyMonkey Audience project was run using a balanced sample. The
process of sample balancing starts by setting targets for desired numbers of completed
responses among certain groups (in this case: by gender, age, and race groupings). After that, a
specified number of potential respondents are allocated to the project based on expected
completion rates. Panelists are either linked directly to the survey from an email invitation, or
routed to the survey after agreeing to take an additional survey after completing a prior one.
Once the survey is put into the field, the system calculates actual completion rates by group,
and uses that information to re-adjust the flow of new panelists to the survey.
SurveyMonkey panelists are emailed no more than once every three days, and on average
panelists receive one email every two weeks. SurveyMonkey also imposes a daily limit on the
number of surveys a panelist can take.
Data for this survey have been weighted for age, race, sex, education and region using the
Census Bureaus American Community Survey, along with data from the Kaiser Health Tracking
Poll on evangelicalism to reflect the demographic composition of the U.S. Because the sample
is based on those who initially self-selected for participation rather than a probability sample,
no estimates of sampling error can be calculated. All surveys and polls may be subject to
multiple sources of error, including, but not limited to sampling error, coverage error, and
measurement error.
To assess the variability in the estimates and account for design effects, we create a bootstrap
confidence interval, meaning we use the weighted data to generate 1000 independent samples
and calculate the standard deviation of the weighted average using those samples, producing
an error estimate.
When analyzing the survey results and their accuracy, this error estimate should be taken into
consideration in much the same way that analysis of probability polls takes into account the
margin of sampling error. For example, if 47 percent of voters say they support Candidate A and
43 percent of voters support Candidate B, and the error estimate is plus or minus 3.5
percentage points, Candidate A could be supported by as low as 44 percent of voters and
Candidate B could have as high as 47 percent of support. Therefore, Candidate A does not have
a "lead."

The following table provides the unweighted sample sizes and the error estimate that has been
calculated in place of the margin of sampling error for a variable that is expected to have close
to an even split in most groupings (gender):
Group
Unweighted N
Total sample
2052

Plus or minus
3.0 percentage points

Republican
Democrat
Independent

424
798
796

6.3 percentage points

4.7 percentage points
4.6 percentage points

18-29
30-44
45-59
60+

315
376
754
607

7.5 percentage points

6.2 percentage points
4.5 percentage points
5.3 percentage points

White
Black
Hispanic
Other

1383
204
317
148

3.8 percentage points

9.0 percentage points
7.2 percentage points
12.3 percentage points