Election Got You Feeling Down? Good News: It Isn’t Just You

When it comes to maintaining the mental health and wellness of American citizens, the 2016 presidential election cycle has been less than ideal. In a recent article in Politico, therapists and their clients described how the last several months of political turmoil have thrown them for a loop. One psychologist even commissioned a poll of 1,000 voter-age respondents to probe the emotional impact of the election. The results: Nearly 30 percent of respondents reported emotional distress due to Hillary Clinton’s campaign, while over 40 percent reported emotional distress due to Donald Trump’s campaign.

Perhaps most tellingly, a whopping 90 percent of those who reported emotional distress felt that the toll of this election was worse than any other election in the past, offering perspective into the general American psyche right about now. But most public health issues—including mental wellbeing—impact different people differently. Women are more likely to seek care for mental health issues than men, for example, while younger individuals are less likely to seek care than older men and women. Similarly, we might expect the mental health and wellness of certain groups of Americans to be more susceptible to the impact of the 2016 presidential election cycle. The effects could vary across state and party lines.

Maimuna Majumder

About

Maimuna Majumder is an Engineering Systems PhD candidate at MIT and computational epidemiology research fellow at HealthMap. Her research interests involve probabilistic modeling, Bayesian statistics, and “systems epidemiology” in the context of emerging infectious diseases.


So we tried to test this hypothesis. By looking at Google search traffic across the United States, we attempted to determine if anxiety- and depression-related terms correlated with searches for election-related terms—and if so, in which states and in what direction?

Over the last several years, Google search traffic has been validated as an effective proxy for a variety of public health issues, including not only infectious diseases like Zika, dengue, and influenza, but mental health as well. When data on the incidence or prevalence of health conditions isn’t readily available—or is expected to be under-reported due to lack of care seeking—Google search traffic can help public health practitioners estimate trends.

While searches for anxiety- and depression-related terms are likely a reasonable indicator for mental wellbeing, correlation against election-related searches doesn’t mean that the election caused changes in mental health. This kind of analysis merely allows us to explore the possibility that trends in mental health—as associated with the 2016 presidential election cycle (but not necessarily because of it)—may vary from one state to the next and, more generally, between red, blue, and swing categories. Insights like these may shed light on which groups, if any, should be surveyed to look for causal relationships. And if those relationships exist, steps can be taken to address election-induced emotional distress in the future.

To assess trends in Google search traffic over time, we collected state-level data on eight election-related terms and two mental-health-related terms (“anxiety + anxious” and “depressed + depression”) from November 2015 to present in weekly increments. We then developed two multivariate models for each state to deduce which election-related terms correlated against anxiety and depression.

Multivariate models allow us to assess the individual impact of each input variable (i.e. election-related search terms in a given state) on the outcome of interest (i.e. anxiety or depression in a given state), after controlling for the others. This is particularly important when the input variables are associated with each other, which is likely true here given that the search terms used are all conceptually related. Moreover, by building models that use data collected at multiple points in time, we’re able to consider expected week-to-week fluctuations in search behavior, such as upticks in election-related Google searches during campaign trail events, for any given state. And perhaps most importantly, by modeling each state separately, we’re able to determine if trends vary across borders.

Seeing Red and Feeling Blue

As it turns out, the relationship between election-related Google search traffic and mental wellbeing isn’t entirely uniform across the United States. For instance, searches for “vote + voting” and “voter registration” are more likely to be correlated with depression and anxiety in swing states than in red and blue ones. Meanwhile, when compared against blue and swing states, increased searches for both “Hillary Clinton” and “Donald Trump” are more likely to be correlated with depression and anxiety in red ones. This indicates that certain brands of susceptibility to election-associated emotional distress differ across groups; though voter-related issues seem to have most influence on mental health and wellbeing among swing state voters, red state voters appear to be more vulnerable to candidate-related issues.

Though some generalizations can be made across party lines, state-by-state results are considerably heterogeneous, suggesting that additional time- and space-varying factors—such as local (i.e. state) government and finances—may also be important. Furthermore, future analysis at the county- or district-level may shed further light upon these relationships, which might be particularly valuable for swing states, where political affiliations are more fragmented.

Increased searches for presidential election seem to be associated with searches for depression and anxiety.

However, some findings are consistent across red, blue, and swing states. Among them is the fact that increased searches explicitly for “presidential election” seem to be associated with searches for depression and anxiety, which suggests that the 2016 presidential election cycle may have been a source of emotional distress for Americans, irrespective of political leaning.

Paired with existing evidence, the results from this analysis add a dimension of complexity to an already compelling argument regarding the impact of the 2016 presidential election cycle on mental health and wellness across the United States. If correlations between election-related Google search trends and mental-health related Google trends are taken as causal, our findings indicate that—though susceptibility varies across groups—the ramp-up to today’s election has likely had adverse effects on the mental health and wellbeing of American citizens.

No matter your state or party, the 2016 presidential election cycle hasn’t been easy. Here’s hoping that today brings our mental health and wellness some much-needed peace.

Go Back to Top. Skip To: Start of Article.

Leave a Reply