"Early, rapid, aggressive": when strategic interactions between governments, opposition, and lobbies can hinder effective responses to epidemics

Abstract

Background: Two critical factors in the success of the response to a threatening epidemic outbreak are the degree of responsibility of the main political actors involved in the response and the population compliance to the proposed measures. The Behavioural epidemiology literature has focused on the latter factor but largely disregarded the former. The multiple failures in COVID-19 control and the lack of consensus that still surround the main response options (i.e., the elimination-suppression-mitigation trichotomy) highlight the importance of considering the political layer in preparedness activities.

Methods: We integrate a simple transmission model into a game-theoretic framework for the interaction between the main political actors involved in the response, namely a government, its opposition and lobbies. The aim is to provide a conceptual framework allowing one to identify the political factors promoting a timely and effective response.

Results: Low degrees of responsibility (i.e., prioritizing consensus over health protection) of political agents can delay or de-potentiate the response until when epidemic growth eventually overtakes the agents' payoffs, thereby forcing them to switch towards the higher degree of responsibility needed to promote an adequate response. When both the government and the opposition are only "partly" responsible, a stall in the response decision-making process likely arises, further delaying the response. Policy and epidemiological parameters amplifying the response delay are ranked by a sensitivity analysis.

Conclusions: Promoting a high degree of responsibility of political actors and lobbies during emergency situations should be a key target of preparedness. Therefore, future pandemic plans should also include, beyond technical indications, ethical statements "guiding" political entities to cooperation.

Keywords: effective responses to epidemics; game theory; political consensus vs. health protection; political responsibility; preparedness; social distancing; transmission models.

Figure 1

The case of partly responsible government vs. an irresponsible opposition initially entrapping the political game on the M policy and causing a delayed switch to policy H. The different subgraphs report, along the two different policy options (M vs. H), the temporal trends of (A) epidemic prevalence $Y(t)$, (B) incidence of severe disease cases, (C) direct epidemic costs, (D) indirect costs, (E) government payoffs; and (F) $\mathrm{\Delta }Payof{f}_{MH}$. The thick blue traits in each subgraphs denote trends during the pre-intervention phase. All parameter values are set at their baseline level (Table 3).

Figure 2

The case of a partly responsible government vs. an irresponsible opposition during the response epoch: one-parameter sensitivity analyses of the time at switch $(t_S)$ from policy M to policy H and related main epidemiological outputs. Top row: sensitivity to the duration of the pre-intervention phase $D_{pre}$. Central row: sensitivity to $R_M$. Bottom row sensitivity to the governmental preference towards protection of direct costs, $w_{Gov}$. Each row reports from left to right: (A) time to switch $(t_S)$, (B) epidemic prevalence at $t_S$, $Y(t_S)$, (C) cumulative number of severe cases at $t_S$, $Z(t_S)$, and (D) temporal trend of the $\mathrm{\Delta }Payof{f}_{MH}$. Other parameter values are set at their baseline level (Table 3). The interpretation of trends of $\mathrm{\Delta }Payof{f}_{MH}$ is provided in the main text.

Figure 3

The response epoch for the case of a partly responsible government vs. an irresponsible opposition: sensitivity analysis based on partial rank correlation coefficients (PRCC) of the time at switch from M to H policy with respect to critical policy parameters, namely, (1) the duration of the pre-intervention phase $(D_{pre})$, (2) the intensity of transmission under policy $\mathrm{M}(R_M)$, (3) the delay of appearance of serious disease with respect to infection $(T_A)$, (4) the scale of pre-intervention payoffs (ABCD), (5) the governmental preference $(w_{Gov})$ for direct costs, and (6) the factor of indirect cost relative to direct ones. Parameter ranges are those in Table 3.