Potential ‘super-spreaders’ –people with many social interactions, extraverts–, the people most likely to widely spread COVID-19, are the most willing to take a free test, despite the increased probability of being quarantined

Testing for COVID-19: willful ignorance or selfless behavior? Linda Thunstroem et al. Behavioural Public Policy , Volume 5, Issue 2, April 2021, pp 135-152. https://doi.org/10.1017/bpp.2020.15

Rolf Degen's take: Potential "super-spreaders," who would suffer the most from being quarantined, are nevertheless those who would most willingly submit to free COVID-19 testing

Abstract: Widespread testing is key to controlling the spread of COVID-19. But should we worry about self-selection bias in the testing? The recent literature on willful ignorance says we should – people often avoid health information. In the context of COVID-19, such willful ignorance can bias testing data. Furthermore, willful ignorance often arises when selfish wants conflict with social benefits, which might be particularly likely for potential ‘super-spreaders’ – people with many social interactions – given people who test positive are urged to self-isolate for two weeks. We design a survey in which participants (n = 897) choose whether to take a costless COVID-19 test. We find that 70% would take a test. Surprisingly, the people most likely to widely spread COVID-19 – the extraverts, others who meet more people in their daily lives and younger people – are the most willing to take a test. People's ability to financially or emotionally sustain self-isolation does not matter to their decision. We conclude that people are selfless in their decision to test for COVID-19. Our results are encouraging – they imply that COVID-19 testing may succeed in targeting those who generate the largest social benefits from self-isolation if infected, which strengthens the case for widespread testing.

Discussion

Widespread testing is one of the most important actions that US governments at any level can undertake to help slow down the spread of COVID-19. Given budget and testing supply constraints, it is likely that random, but voluntary, testing will be the most effective policy. We design a survey to examine the risks from self-selection into taking a COVID-19 test.

Overall, we observe that around 70% of people would agree to a costless COVID-19 test. We find that people who are more worried about their own health due to COVID-19 are more likely to test, as are young healthy people, relative to older healthy people. Ability to afford self-isolation for 14 days does not seem to affect the decision to test. Furthermore, people who worry more about their health, and people with health insurance or health coverage through Medicare or Medicaid, are more likely to take the test, as are people identifying as Democrats compared to Republicans.

Contrary to our expectation, we also find that potential ‘super-spreaders’ are more likely than other individuals to agree to a costless COVID-19. It could be that extroverts are more willing than expected to take a COVID-19 test because their private cost of doing so is unusually low due to the broadly implemented social distancing at the time of data collection for this study. If extroverts are already relatively isolated (i.e., due to a stay-at-home order and mandated closures by the state governor of public spaces, such as gyms, restaurants and bars), the personal cost of testing might be low. Furthermore, extroverts might be more likely to get infected if they socialize more, which could be a ‘selfish’ motivation to get tested. However, we control for the current level of compliance with social distancing, which should address both of these private motivations for increased probability of testing, and we find that people who comply more are less motivated to take the test. We also control for their worry about own health due to COVID-19. Even so, the positive effect on willingness to test from being an extrovert persists. We therefore conclude that the positive effect of being an extrovert on willingness to test for COVID-19 is likely due to social health benefits weighing more heavily in their decision than their private costs from potential self-isolation for 14 days, should the test come back positive. The importance of the prosocial motive in determining COVID-19 testing is consistent with the results of the study by Jordan et al. (2020), who find that prosocial messages are more effective than self-interested messages in promoting behavior that prevent the spread of COVID-19 (e.g., hand washing, hand shaking, hugging).

Our results suggest that the risks of adverse selection (in terms of failing to target the people most likely to spread the virus) in testing for COVID-19 might be fairly low. This underscores the value of widespread testing, even if it cannot be truly random, and the importance of making such testing available nationwide in the USA as soon as possible.

An important shortcoming of our analysis is that it builds on hypothetical survey data. It is well documented that survey answers may be affected by a ‘hypothetical bias’, meaning that people answer one way in a survey and behave in a different way when faced with real, incentivized decisions. This risk pertains to our study as well, and the hypothetical bias might be particularly pronounced if the choice to test for COVID-19 is regarded as prosocial. Several studies suggest that a hypothetical bias is particularly likely when measuring prosocial behavior – people often exaggerate the extent to which they engage in such behavior (e.g., Murphy et al., 2005; Vossler et al., 2012; Jacquemet et al., 2013). Furthermore, it is possible that personal costs to the testing decision are less salient in a hypothetical context. Once testing is more widespread in the USA, it will be important to examine who actually chooses to get tested, and the extent to which they deviate from the general population. That said, hypothetical and incentivized behavior generally correlate, such that an analysis like ours can provide important insights into the potential pitfalls of voluntary testing, prior to the actual testing. This is useful information to have on hand when designing an efficient and cost-effective testing strategy.