The first weekly wave of the Democracy Fund + UCLA Nationscape survey, fielded July 18–24, 2019. Approximately 6,250 completed online interviews drawn from the Lucid respondent exchange platform using a non-probability quota design, with raking weights calibrated to ACS demographic targets and 2016 presidential vote choice.
ns_wave1A data frame with approximately 6,250 rows and 171 variables
(170 survey variables plus wave_id added by the prepare script).
Unique respondent ID (integer).
Interview date (character, "YYYY-MM-DD" format).
Wave identifier: "ns20190718" for all rows in this
dataset.
Raking weight calibrated to ACS demographic targets and 2016 presidential vote choice. Use for all population-level estimates.
Country direction: 1 = Right direction,
2 = Wrong track, 3 = Not sure.
Economy outlook: 1 = Better, 2 = Worse,
3 = Same, 4 = Not sure.
Political interest (4-pt): 1 = Very interested,
4 = Not at all interested.
Voter registration: 1 = Registered,
2 = Not registered, 3 = Not eligible.
Trump presidential approval: 1 = Strongly approve,
2 = Somewhat approve, 3 = Somewhat disapprove,
4 = Strongly disapprove.
2020 vote intention: 1 = Trump,
2 = Democratic candidate, 3 = Other, 4 = Don't plan to vote,
5 = Not sure.
2016 presidential vote. See labels.
Write-in for vote_2016 "other" choice.
Would consider voting for Trump: 1 = Yes,
2 = No.
Reason for not considering Trump (open text).
Primary vote party: 1 = Democratic,
2 = Republican, 3 = Other.
Democratic primary vote intention. See labels.
Write-in for dem_vote_intent "other".
Top-ranked Democratic presidential candidate. See labels.
Second-ranked Democratic candidate. See labels.
Third-ranked Democratic candidate. See labels.
Wants non-Trump Republican nominee: 1 = Yes,
2 = No, 3 = Not sure.
U.S. House vote intention: 1 = Democrat,
2 = Republican, 3 = Other, 4 = Won't vote, 5 = Not sure.
U.S. Senate vote intention. Same codes as
house_intent.
Governor vote intention. Same codes as
house_intent.
Used social media for political news in past
week: 1 = Selected, 2 = Not selected. See "question_preface"
attribute for shared question stem. Same coding for all news_sources_*
variables.
Used CNN for political news.
Used MSNBC for political news.
Used Fox News for political news.
Used network news (ABC/CBS/NBC/PBS).
Used local TV news.
Used Telemundo or Univision.
Used NPR.
Used AM talk radio.
Used a national newspaper.
Used a local newspaper.
Used another news source: 1 = Selected,
2 = Not selected.
Write-in for news_sources_other.
Favorability toward Whites: 1 = Very
favorable, 2 = Somewhat favorable, 3 = Somewhat unfavorable,
4 = Very unfavorable, 5 = Not sure. Same coding for all
group_favorability_* variables.
Favorability toward Blacks.
Favorability toward Latinos.
Favorability toward Asians.
Favorability toward Christians.
Favorability toward Socialists.
Favorability toward Muslims.
Favorability toward labor unions.
Favorability toward the police.
Favorability toward undocumented immigrants.
Favorability toward gays and lesbians.
Favorability toward Republicans.
Favorability toward Democrats.
Favorability toward Donald Trump. Same
5-point scale as group_favorability_* variables.
Favorability toward Barack Obama.
Favorability toward Alexandria Ocasio-Cortez.
Favorability toward Joe Biden.
Favorability toward Kamala Harris.
Favorability toward Pete Buttigieg.
Favorability toward Elizabeth Warren.
Favorability toward Bernie Sanders.
Favorability toward Mike Pence.
Trump vs. Biden head-to-head: 1 = Trump, 2 = Biden,
3 = Not sure. Same coding for all trump_* matchup variables.
Trump vs. Sanders.
Trump vs. Harris.
Trump vs. Warren.
Trump vs. Buttigieg.
Trump vs. Cory Booker.
Trump vs. Julian Castro.
Trump vs. Tulsi Gabbard.
Trump vs. Kirsten Gillibrand.
Trump vs. Beto O'Rourke.
Pence vs. Biden head-to-head: 1 = Pence, 2 = Biden,
3 = Not sure. Same coding for all pence_* matchup variables.
Pence vs. Buttigieg.
Pence vs. Harris.
Pence vs. Sanders.
Pence vs. Warren.
Whether Donald Trump cares about telling
the truth: 1 = Yes, 2 = No, 3 = Not sure. Same coding for all
cand_truth_* variables.
Whether Elizabeth Warren cares about the truth.
Whether Joe Biden cares about the truth.
Whether Bernie Sanders cares about the truth.
Whether Pete Buttigieg cares about the truth.
Whether Kamala Harris cares about the truth.
Whether Donald Trump relies on facts vs.
hunches: 1 = Facts and evidence, 2 = Hunches, 3 = Not sure. Same
coding for all cand_facts_* variables.
Whether Elizabeth Warren relies on facts.
Whether Joe Biden relies on facts.
Whether Bernie Sanders relies on facts.
Whether Pete Buttigieg relies on facts.
Whether Kamala Harris relies on facts.
Agree/disagree: minorities should work
their way up without special favors. 1 = Strongly agree, 2 = Agree,
3 = Neither, 4 = Disagree, 5 = Strongly disagree. Same scale for
all racial_attitudes_* and gender_attitudes_* variables.
Agree/disagree: generations of slavery make it difficult for Blacks to work out of the lower class.
Agree/disagree: I prefer close relatives marry someone from the same race.
Agree/disagree: it's alright for Blacks and Whites to date.
Agree/disagree: more comfortable with a male boss than female boss.
Agree/disagree: women are just as capable of thinking logically as men.
Agree/disagree: increased opportunities for women have improved quality of life.
Agree/disagree: women who complain about harassment cause more problems than they solve.
Perceived discrimination against Blacks:
1 = A great deal, 2 = A lot, 3 = A little, 4 = None at all,
5 = Not sure. Same scale for all discrimination_* variables.
Perceived discrimination against Whites.
Perceived discrimination against Muslims.
Perceived discrimination against Christians.
Perceived discrimination against Women.
Perceived discrimination against Men.
U.S. Senate knowledge question. See labels.
U.S. Supreme Court knowledge question. See labels.
3-category party ID: 1 = Democrat, 2 = Republican,
3 = Independent, 4 = Something else.
7-point party ID (legacy coding). See labels.
Strength of Democratic ID (conditional on
pid3 == 1). See labels.
Strength of Republican ID (conditional on
pid3 == 2). See labels.
Partisan lean of Independents (conditional on
pid3 == 3). See labels.
5-point ideological self-placement: 1 = Very liberal,
5 = Very conservative.
Employment status (selected choice). See labels.
Write-in for employment "other".
Born outside the U.S.: 1 = Yes, 2 = No.
Primary language at home. See labels.
Religious affiliation (selected choice). See labels.
Write-in for religion "other".
Born-again or evangelical Christian: 1 = Yes,
2 = No.
Sexual orientation. See labels.
Labor union membership: 1 = Yes, 2 = No,
3 = Non-union household, 4 = Not sure.
Household gun ownership: 1 = Yes, 2 = No,
3 = Not sure.
Support building a wall on the southern U.S. border:
1 = Strongly support, 2 = Somewhat support, 3 = Somewhat oppose,
4 = Strongly oppose, 5 = Not sure. Same scale for all policy items
through limit_magazines. See "question_preface" attribute on each
variable for the exact shared question stem.
Support capping carbon emissions.
Support large-scale government investment in environmental technology.
Support requiring background checks for all gun purchases.
Support cutting taxes for families making < $100K/year.
Support eliminating the estate tax.
Support raising taxes on families making > $600K.
Support ensuring all students can graduate from state colleges debt-free.
Support requiring a waiting period and ultrasound before an abortion.
Support never permitting abortion.
Support permitting abortion in cases other than rape/incest/life at risk.
Support permitting late-term abortion.
Support allowing employers to decline abortion coverage.
Support guaranteeing jobs for all Americans.
Support enacting a Green New Deal.
Support creating a public registry of gun ownership.
Support separating children from parents prosecuted for illegal border crossing.
Support shifting to a merit-based immigration system.
Support requiring proof of citizenship to wire money internationally.
Support impeaching President Trump.
Support withdrawing military support for Israel.
Support legalizing marijuana.
Support requiring 12 weeks of paid maternity leave.
Support Medicare-for-All.
Support reducing the size of the U.S. military.
Support raising the minimum wage to $15/hour.
Support banning people from predominantly Muslim countries.
Support removing barriers to domestic oil and gas drilling.
Support granting reparations to descendants of slaves.
Support allowing people to work in unionized workplaces without paying union dues.
Support displaying the Ten Commandments in public schools and courthouses.
Support limiting trade with other countries.
Support allowing transgender people to serve in the military.
Support raising taxes on families making > $250K.
Support providing tax-funded vouchers for private or religious schools.
Support providing government-run health insurance to all Americans.
Support providing the option to purchase government-run insurance.
Support subsidizing health insurance for lower income people not on Medicaid.
Support creating a path to citizenship for all undocumented immigrants.
Support a path to citizenship for DREAMers.
Support deporting all undocumented immigrants.
Support banning all guns.
Support banning assault rifles.
Support limiting gun magazines to 10 bullets.
Respondent age in years.
Gender: 1 = Male, 2 = Female, 3 = Other.
Census region: 1 = Northeast, 2 = Midwest,
3 = South, 4 = West.
Hispanic or Latino origin: 1 = Yes, 2 = No.
Race/ethnicity (6 categories). See labels.
Household income (7 brackets). See labels.
Educational attainment (6 categories). See labels.
U.S. state of residence (2-letter abbreviation).
Congressional district.
This dataset is the first of 77 weekly waves collected from July 2019 through January 2021. The full survey ran in three phases:
| Phase | Weeks | Dates | Approx. N |
| Phase 1 | 1–24 | Jul 18, 2019 – Dec 26, 2019 | 150,000 |
| Phase 2 | 25–50 | Jan 2, 2020 – Jun 25, 2020 | 162,500 |
| Phase 3 | 51–77 | Jul 2, 2020 – Jan 12, 2021 | 168,750 |
Only Wave 1 is bundled in the package because 77 waves × ~6,250 rows
would be prohibitively large. To obtain the full dataset by phase, use the
prepare scripts in data-raw/ (see the Source section).
Survey design:
The Nationscape is a calibrated non-probability sample (quota design with
raking weights). Use as_survey_nonprob() — it is designed specifically
for this use case and will gain bootstrap re-calibration variance in Phase
2.5:
svy <- as_survey_nonprob(ns_wave1, weights = weight)
Metadata:
All substantive columns carry variable labels ("label" attribute) set
during data preparation. Battery items additionally carry a
"question_preface" attribute with the shared question stem. Value
labels ("labels" attribute) are present for all coded response items.
Battery structure:
Most multi-item question groups follow a {battery}_{item} naming
convention. All items within a battery share an identical
"question_preface" attribute:
| Battery prefix | Preface summary | N items |
news_sources_* | News sources used in past week | 13 |
group_favorability_* | Favorability toward named groups | 13 |
cand_favorability_* | Favorability toward named candidates | 9 |
trump_* | Trump head-to-head matchups | 10 |
pence_* | Pence head-to-head matchups | 5 |
cand_truth_* | Whether each candidate tells the truth | 6 |
cand_facts_* | Whether each candidate relies on facts | 6 |
racial_attitudes_* | Agree/disagree racial attitude items | 4 |
gender_attitudes_* | Agree/disagree gender attitude items | 4 |
discrimination_* | Perceived discrimination by group | 6 |
Three policy batteries share the same Agree/Disagree/Neither scale:
wall, cap_carbon, environment, guns_bg, mctaxes, estate_tax,
raise_upper_tax, college, abortion_waiting, abortion_never,
abortion_conditions, late_term_abortion, abortion_insurance,
guaranteed_jobs, green_new_deal, gun_registry,
immigration_separation, immigration_system, immigration_wire,
impeach_trump, israel, marijuana, maternityleave,
medicare_for_all, military_size, minwage, muslimban,
oil_and_gas, reparations, right_to_work, ten_commandments,
trade, trans_military, uctaxes2, vouchers, gov_insurance,
public_option, health_subsidies, path_to_citizenship, dreamers,
deportation, ban_guns, ban_assault_rifles, limit_magazines.
Tausanovitch, Chris and Lynn Vavreck. 2021. Democracy Fund + UCLA Nationscape, October 10–17, 2019 (version 20210301). Retrieved from voterstudygroup.org/data/nationscape.
Rivers, Douglas and Delia Bailey. 2009. "Inference from matched samples in the 2008 U.S. national elections." Proceedings of the Joint Statistical Meetings, Social Statistics Section.
# Design variables
head(ns_wave1[, c("response_id", "weight", "age", "gender")])
# Inspect a battery item's metadata
attr(ns_wave1$group_favorability_blacks, "label")
attr(ns_wave1$group_favorability_blacks, "question_preface")
attr(ns_wave1$news_sources_cnn, "labels")
# Create a calibrated survey design (correct approach for raked
# non-prob samples)
svy <- as_survey_nonprob(ns_wave1, weights = weight)
get_freqs(svy, pres_approval)
# Party identification distribution
table(ns_wave1$pid3)
Run the code above in your browser using DataLab