Polarized Pandemic Response and Covid-19 Connectedness Across US States
Yves here. This study is useful because it tests and confirms some widespread intuitions: that more aggressive Covid containment measures reduce contagion levels, and that the degree of connection between high infection areas (which skew Republican) increases infection among them and also generates net spread to lower infection states.
By Umut Akovali, Post-doc Researcher, Department of Economics, Koç University and Kamil Yilmaz, Professor of Economics, Koç University. Originally published at VoxEU
Balancing the trade-off between strict public health measures and economic activity has been the key concern for governments since the Covid-19 outbreak. This column studies Covid-19 infections and their connectedness across US states. It finds that states with lax government policy and community mobility response had higher case growth trajectories and generated connectedness of Covid-19 cases to other states. Further, states with Republican governors tend to have higher connectedness of Covid-19 cases among themselves and generate net connectedness to states with Democratic governors.
Balancing the trade-off between strict public health measures and economic activity has been the key concern for governments since the Covid-19 outbreak. While strict measures impose high economic costs, they may play an important role in containing the spread of the virus, and consequently in preventing a more prolonged and costly reduction in economic activity (Furman 2020, Odendahl and Springford 2020). In the case of the US, the federal government showed a lax response to the pandemic, focusing on the economic impact rather than the rapidly increasing cases of Covid-19 and the death toll. The resulting public debate politicised the federal and state-level public health policies and led to the formation of demarcation lines across the party lines in a year of presidential elections.
In a new paper (Akovali and Yilmaz 2020), we study the substantial differences in state-level responses leading to a wide variation in the new Covid-19 case trajectories across the country. Our work adds to the recent literature by demonstrating the close association between government policy and community mobility responses to the pandemic and the trajectories of new Covid-19 cases as well as their connectedness/spillovers across the US.
Politicised State-Level Response and Pandemic Momentum
We first focus on the relative performance of the states in terms of new Covid-19 cases. Figure 1 shows the average per-capita number of daily new cases of Covid-19 in states with Democratic and Republican governors. Many coastal states that have Democratic governors were harshly hit at the earlier stages of the outbreak, but they managed to limit the pandemic’s early growth with strict measures. The states with Republican governors outstripped their Democratic counterparts as of the first week of June. Since then, the number of new cases per capita has been 30-40% higher in states with Republican governors compared to states with Democratic governors.
Figure 1 Average per-capita new cases in Democratic and Republican states
In a next step, we use panel regression analysis and find that the strict public health policy implementation by state governments was critical for containing the virus. Panel regression results also reveal that the changes in community mobility patterns played a crucial role in the containment of Covid-19 cases. States with Democratic governors (blue bars) overwhelmingly followed more stringent (above the national average) policies to contain the virus. In contrast, the majority of the states with Republican governors (red bars) followed less stringent (below the average) policies (Figure 2).
Figure 2 Average state-level government stringency index
Figure 3 shows the time-series of the unweighted average government policy response in the states with Democratic and Republican governors. While panel a shows that, on average, states with Democratic governors followed more stringent policies than those with Republican governors, panel b displays a clear positive trend of the policy stringency difference between the two groups over time. Both panels of Figure 3 provide further evidence for the government policy difference between the states with Democrat and Republican governors.
Figure 3 Stringency index in Democratic and Republican states
Even though the community’s use of public spaces can be influenced by strict public policy measures, this also depends on the local community’s awareness of the risks involved. Therefore, it is worth focusing on the changes in the state-level community mobility patterns in response to the pandemic. The average mobility indices for the two groups of states in Figure 4 (panel a) depicts a picture similar to the one for the policy stringency index in Figure 2. Communities in states with Democratic governors decreased their use of public spaces more than communities in states with Republican governors. This difference persisted until the end of the sample on 23 October.
Figure 4 Community mobility and travel intensity: Democratic vs Republican states
In line with the lax government policies, states with Republican governors also did not impose air travel restrictions across states. Consequently, travel intensity between states with Republican governors has been higher than the travel intensity between states with Democratic governors. Furthermore, travel intensity from Republican to Democratic states has been higher than travel intensity from Democratic to Republican states (see Figure 4, panel b).
Within and Between Connectedness Dynamics across Democratic and Republican States
The absence of leadership from the federal government resulted in relatively high rates of Covid-19 cases in many states. Moreover, the substantial variation in public health policy measures across states fostered the transmission of Covid-19. In a time-series context, a connectedness analysis (Diebold and Yilmaz 2014) is a useful tool to estimate the transmissions when structural models are not readily available. The connectedness of case growth provides further empirical evidence on the dissociation between the pandemic’s course in states along the party lines.
In our paper, we show the dynamic behaviour of system-wide connectedness over time. Connectedness was much more volatile in the early phases of the Covid-19 pandemic, but it became relatively more stable over time. Interesting results are obtained when the state-level connectedness measures are aggregated along the political party lines. As expected, the Covid-19 connectedness between Democratic states (Figure 5a) and from Democratic to Republican states (Figure 5b) is higher in the early phases of the pandemic. Starting in May, however, there was a reversal in the direction of Covid-19 connectedness. The directional connectedness from Republican to Democratic states and between Republican states peaked when the number of new Covid-19 cases in Republican states increased from the middle of May up to the first week of July. As Americans started travelling for the summer holidays, Covid-19 cases increased in both Democratic and Republican states from the 4th of July up to the Labor Day weekend. Consequently, over this period, the directional connectedness within and between Democratic and Republican states fluctuated around the same level. After the summer vacation, the directional connectedness from Republican to Democratic states and within Republican states increased again.
Figure 5 Within and between group connectedness across republican and democratic states
Finally, using secondary regression analysis, we focus on the impact of government and community responses and the travel intensities on the directional connectedness between pairs of states. The states with lax government and community responses to the pandemic generated connectedness of Covid-19 cases to those states that followed stricter policy and community response. We also found that Covid-19 connectedness between pairs of states was quite strongly related to the travel intensity across states, as measured by the mobile phone location exposure index (LEX). Finally, the Republican party affiliation of the governors turned out to be a significant determinant of the pairwise Covid-19 connectedness across states.
We show that politicised responses to the coronavirus pandemic contributed to the spread of the virus across US states. We further find that strong government and community responses were effective in reducing the number of Covid-19 infections and lower the connectedness of Covid-19 cases across states. Finally, we conclude that unless the federal government starts implementing stringent public health policies at the national level and coordinating state-level policies, controlling both Covid-19 cases and their connectedness across states will remain challenging.