Effect of Partisanship on Family Ties

The Effect of Partisanship and Political Advertising on Close Family Ties

Study Focus: Political polarization's impact on private life, specifically family ties, often overlooked in favor of studying public institutions and political processes.
Data Used: Anonymized smartphone-location data combined with precinct-level voting data from the 2016 election.
Key Finding: Thanksgiving dinners were shorter by 30 to 50 minutes when attended by people from opposing-party precincts.
- The average dinner duration was 257 minutes.
- This reduction persisted even after applying spatial and demographic controls.
Impact of Political Advertising: The duration decrease tripled in 2016 for travelers from media markets with heavy political advertising compared to 2015.
Asymmetric Effects: Republican-precinct residents shortened dinners by a greater margin than Democratic-precinct residents due to political differences.
Overall Time Loss: An estimated 34 million hours of cross-partisan Thanksgiving dinner discourse were lost nationally in 2016 because of partisan effects.

Background on Political Partisanship

Increased Partisanship: Over the last 25 years, political partisanship in America has significantly increased.
Negative Feelings: In 2016, over 55% of both Democrats and Republicans had “very unfavorable” feelings toward the opposing party, a large increase from the mid-1990s (17-21%).
Spatial Sorting: Spatial partisan sorting is creating increasingly homogeneous electoral “bubbles”.
Cross-Party Intolerance: Intolerance extends into personal relationships, with parents disapproving of cross-party dating and marriage.
Broader Impact: Political polarization influences decisions about shopping and work at rates higher than race, ethnicity, or religion.

Study on Family Gathering Durations

Focus: Study examines if politics strains close family ties by measuring the duration of family gatherings.
Thanksgiving as a Focal Point: Thanksgiving was chosen because it often brings together family members with differing political views shortly after elections.
Anecdotal Evidence: Many families canceled or shortened Thanksgiving plans with politically problematic relatives after the 2016 election.

Cognitive Biases and Political Interactions

Partisan Selective Exposure: Individuals avoid counter-attitudinal political information to prevent cognitive dissonance or relationship harm.
Belief Polarization: Individuals move toward more extreme positions during political discussions.
Incorrect Expectations: People incorrectly expect others to respond similarly to debates, anticipating belief convergence instead of polarization.
Attribution of Bias: Individuals attribute a lack of convergence to the bias and irrationality of others.

Methodology and Data Analysis

Data Integration: Smartphone location data from over 10 million Americans was combined with precinct-level election data from 2016 to determine the relationship between disagreement and time spent together.
Isolating Partisanship Effects: Comparison groups of smartphone users sharing the same home-destination pairs were constructed.
Political Advertising Consideration: The study compared users on opposite sides of media-market boundaries to account for the influence of political advertising.
Data Sources: Precinct-level election results, political advertising data from Kantar Media’s Campaign Media Analysis Group, and demographic controls from the 2010 Decennial Census and the Census Bureau’s 2012–2015 American Community Survey were used.
Location Data: Location data was aggregated by SafeGraph, using pings from over 10 million U.S. smartphones.
Home and Thanksgiving Locations: Home locations were inferred from pings between 1:00 a.m. and 4:00 a.m. over three weeks before Thanksgiving. Thanksgiving locations were based on modal location between 1:00 p.m. and 5:00 p.m. on Thanksgiving Day.
Sample Representativeness: The sample represents the 77% of Americans who own smartphones.
Vote Share Imputation: Each resident was assigned a vote ratio proportional to their home precinct’s two-party vote share to test political representativeness.
- Example: A resident of a precinct with 150 Clinton and 50 Trump votes would be assigned 0.75 Clinton and 0.25 Trump votes.
Accuracy: Imputed votes were accurate to within 1 percentage point in 33 states and within 5 percentage points in all states.

Analysis of Thanksgiving Dinner Duration

Sample Restriction: The sample was restricted to residents who were home in the morning and night of Thanksgiving but traveled for dinner.
Equation: The following equation was used to estimate the effects:
- $duration{ij} = a + b \cdot mismatch{ij} + gF{ij} + e{ij}$
- Where $mismatch{ij} = Pi(1 - Pj) + (1 - Pi)P_j$
- $duration_{ij}$ is the number of minutes traveler i spent with host j on Thanksgiving.
- $F_{ij}$ is a set of fixed effects that form groups of people defined by pairs of home (i) and destination (j) locations.
- $b$ is the coefficient of interest.
- $Pi$ and $Pj$ are the two-party vote shares associated with home precincts for i and j, where $Pi = \frac{democratici}{democratici + republicani}$ .
- $mismatch_{ij}$ is the imputed probability that persons i and j voted for different candidates in 2016.
Fixed Effects: Regressions controlled for progressively finer (i, j) location pairs, culminating in five-digit geohash (geohash-5) boxes (3 miles by 3 miles).
Controls: Regressions compared Thanksgiving dinner durations between travelers with the same home and destination areas to control for demographics, distance, and travel time.
Results: Families likely to have voted for different presidential candidates spent about 30 to 50 fewer minutes together. The estimate remained stable with finer spatial controls (56.3 ± 14.6 min under geohash-5 controls).

Asymmetric Effects and Political Advertising

Components of Mismatch: The two components of $mismatch{ij}$ , $Pi(1 - Pj)$ and $(1 - Pi)P_j$ , were examined to measure the effect of voting disagreement among Democratic-precinct residents (DPRs) visiting Republican-precinct residents (RPRs) and vice versa.
Differential Impact: DPRs shortened visits to RPR hosts by about 20 to 40 minutes, while RPRs shortened visits to DPRs by about 50 to 70 minutes.
Statistical Difference: F-test results indicated estimates were statistically different (P < 0.0001 in four of five specifications), with RPRs shortening their cross-party stays more than DPRs.
Effect of Advertising: Cross-partisan Thanksgiving dinners were further shortened by around 2.6 minutes on average for every 1000 political advertisements aired in the traveler’s home media market.
- For example, vote-mismatched families in Orlando saw a 69-minute-shorter Thanksgiving dinner compared to those in markets without advertising due to the high number of ads (26,000+).

Placebo Test and Destination Choices

Placebo Test: A placebo test concerning whether advertisements in 2016 affected Thanksgiving dinner behavior the year before airing found no evidence of preexisting partisan effects in regions with high advertising levels.
Destination Choices: Travelers did not significantly change plans to reduce political divisions from 2015 to 2016.
Duration Adjustment: Travelers were more likely to change the duration of Thanksgiving gatherings than the destination.

Travel Likelihood and Overall Time Loss

Reduced Travel: DPRs reduced their likelihood of travel between 2015 and 2016 by 2 percentage points more than comparable RPRs, especially in areas with heavy political advertising.
Matched Resident Comparison: When matched residents within 1.5 miles were compared, results were consistent.
Overall Time Loss: Partisan differences cost Americans 73.6 million hours of Thanksgiving time in 2016.
- Political advertising–related partisanship comprised 15.9 million of those lost person-hours.
- An estimated 33.9 million person-hours of cross-partisan discourse were eliminated.

Implications and Conclusions

Corroboration of Media Reports: Study corroborates anecdotal media reports after the 2016 election about avoiding personal confrontations over political issues among Democratic voters.
Republican Sensitivity: RPRs were more sensitive to partisan differences at Thanksgiving dinners, supporting findings of greater partisan-selective exposure among Republicans in news-media consumption.
Extension of Polarization: Partisan polarization extends to close family settings in quantitatively meaningful ways.
Impact of Political Advertising: Political advertising and related campaign efforts can exacerbate these fissures.
Policy Implications: Policies designed to shorten campaigns may reduce the private costs of political polarization, especially as abbreviated Thanksgiving gatherings tend to accumulate in regions with greater campaign activity.