Sage Journals: Discover world-class research

Abstract

This paper explores the role of political institutions, government effectiveness and, social trust in determining government financial responses to the coronavirus pandemic. The results from a globally representative sample of eighty countries over the Q3 2020 - Q3 2021 period suggest that politics did play a role in determining financial response to the coronavirus pandemic but the effects were more driven by current factors (electoral competition and executive fragmentation) and the effectiveness of, and trust in, government, than by institutional factors such as regime type or dynamic factors such as the political business cycle. Ineffective governments outspent effective ones with the magnitude of financial responses diverging as trust in government increases. These findings suggest that highly trusted effective governments were able to capitalize on trust premiums to minimize covid spending while highly trusted ineffective governments took advantage of trust premiums to maximize the size of their spending packages.

Keywords

coronavirus coronavirus financing public finance political economy trust government effectiveness

Introduction

At its core, the coronavirus pandemic was an unanticipated public health issue that necessitated scientific solutions ideally under a responsible economic decision-making constraint. In practice solutions became entangled with politics and political discourse in an environment of widespread distrust in politicians and the effectiveness of political institutions. This, in turn, may have influenced public financial responses to the pandemic. The political economy literature suggests that we should expect systematic differences in financial decision making across i) regime types, dependent on their ability to make use of repression at lower marginal costs, ii) electoral competition and the political business cycle and, iii) the size and fractionalization of meaningful actors in the budget process. Moreover, because policymakers were reacting to the pandemic at a time of diminishing levels of trust in government since the Great Financial Crisis (GFC) of 2008, we may expect this to influence the size of their financial rescue packages. The purpose of this paper is to examine the effects of political institutions, government effectiveness and, social trust on government financial responses to the coronavirus pandemic. This is accomplished by drawing on a global sample of eighty countries covering the Q3 2020–Q3 2021 period from a collection of unique datasets with source data ranging from IMF's Database of Fiscal Policy Responses to Covid 19 to Wellcome Trust micro-level survey data.

The results suggest that politics did play a role in determining financial response, but these effects were more driven by the effectiveness of, and trust in, government than from institutional factors such as regime type. Given the urgency of response packages current circumstances (whether it was an election year, government ideology and executive fragmentation) played a role in financial decision making, but forward-looking strategic measures did not (the political business cycle). The effect of social trust and government effectiveness suggests that ineffective governments outspent effective ones with the magnitude of these financial responses diverge as trust in government increases. These findings suggest that highly trusted effective governments were able to capitalize on trust premiums to minimize covid spending while highly trusted ineffective governments took advantage of trust premiums to maximize the size of their spending packages.

The remainder of this article is organized as follows. Section 2 discusses the political and financial background and context leading up to government covid financing. Section 3 considers relevant theoretical contributions from the political economy literature and Section 4 reviews the data and measurement issues. The fifth section presents the statistical specification and results. In Section 6 we consider the robustness of the results in the context of automatic stabilizers. Section 7 reviews some of the limitations of the empirical approach and Section 8 concludes.

Background: Political Baggage and the Great Financial Crisis

The GFC put policymakers across the globe in a difficult economic, financial, and political predicament. Not only were they ‘flying blind’ when it came to designing an effective and responsible rescue package (Ramey 2019), but the most practical and necessary financial solutions came with political ramifications. Voters in advanced economies perceived the bailout of financial institutions as rewarding private firms who willingly took on risk with the goal of generating short run profits. From this perspective, when these risks did pay off the financial benefits were seen as rewarding investors and executives but, when they did not pay off there was an implicit need for taxpayers to rescue risk seeking institutions who had genuinely become too big to fail. This created a catch twenty-two for policymakers in a post 2008 environment – do nothing and the financial systems spiral out of control leading to the wrath of angry voters, intervene and face the wrath of angry voters who may attempt to disrupt financial systems. Movements like ‘Occupy Wallstreet’ started taking to the streets in 2011 advocating against governments perceived cozy relationship with ‘the one percent’ and trust in government and financial institutions fell to new lows. The year 2012 was labeled as the “fall of government” with global trust in government falling to all-time lows of 38%, below all other sectors including NGOs, media and business (Edelman 2012).

During the post GFC period, trust in politicians not only continued to fall but disseminated into other parts of the public sector including accommodating central bankers and, more recently, public health officials. Studies on trust in the European Central Bank (ECB) using Eurobarometer data, have found that the GFC had a significant downward effect, or structural break, on trust in central bankers which was partly associated with their implicit relationship with government (Gros and Roth 2010, Gros, Nowak-Lewhmann and Roth 2012, Wälti 2012). These findings were validated by Bursian and Furth (2015) who found that a wide range of socioeconomic, macroeconomic, fiscal, and political factors played a role in determining levels of trust in the ECB over the 1999–2010 period, with a structural break after 2008. In the case of the US, social distrust expanded beyond politicians and central bankers to include public health authorities and regulators in the context of an opioid epidemic that was “created or exacerbated by webs of influence woven by several pharmaceutical companies” involving “health professionals, patient advocacy groups, medical professional societies, research universities, teaching hospitals, public health agencies, policymakers, and legislators” (Marks 2020). The aftermath of the GFC also saw politics becoming more polarized, leading to more divisive societies and a rise of populist politicians, rallying against ‘big government’ and ‘big business.’ Political discourse often fuelled these divided bases and politicians across the ideological spectrum engaged in expansionary fiscal policy to finance their populist rhetoric. Fortunately for politicians in advanced economies, there was a slow abandonment of restrictive orthodox economic policymaking that had been broadly accepted under the umbrella of the Washington Consensus, which promoted free market principles with limited government intervention in private firms, the maintenance of low government deficits and debt, and limited central bank purchases of government debt (Armingeon 2012). In fact, the latter was restricted by law in Article 123 of the ECB Treaty at its inception which prevented any immediate or direct monetary financing of public spending by the central bank.

This put policymakers in a precarious position when the coronavirus swept across the globe in 2019: not only was “[d]oing science in real-time always a difficult matter” (Boettke and Powell 2021), trust in public institutions was low, the public was politically polarized and, governments/central banks were already managing large balance sheets and exposure to off-balance sheet risks in the form of contingent liabilities. Effectively, the least trusted and already highly leveraged part of society was responsible for designing and successfully implementing complicated and ambitious financial rescue packages for the divided societies they were trying to govern. These challenges were further exacerbated in the US where highly publicized failures from health authorities during the opioid epidemic meant that public distrust had already proliferated beyond pharmaceutical companies and expanded to a wider realm which included public health experts, scientists, and science itself. Vaccine acceptance rates differed significantly based on political affiliation and the advice or endorsement of political elites (Kreps et al. 2020) leading to a “diminished confidence in the government's ability to adequately oversee and regulate the COVID-19 vaccine development and testing process” (Boodram et al. 2021). In the U.S. (April 2020), only 52% of Democrats and 27% of Republicans reported confidence in their belief that ‘scientists act in the best interest of the public’ (Funk et al. 2020; Boodram et al. 2021). Further evidence from McLaughlin, Mewhirter and Sanders (2021), who examined survey data for 12,037 US residents, found that a large portion of the population believed that politics motivated the way scientific studies were conducted. A global survey on potential acceptance of covid vaccinations in June of 2020 found that countries with high vaccine acceptance were nations with high levels of trust in the central government with 88.6 percent of respondents from China (and over 80 percent in Singapore and South Korea) saying that they would take a ‘proven, safe and effective vaccine’ (Lazarus et al. 2021). The results from Lazarus et al. suggest that “trust is an intrinsic and potentially modifiable component of successful uptake of a COVID-19 vaccine” with general findings showing that trust in government plays a significant role in determining vaccine acceptance rates and public compliance with government recommendations.

In reaction to a growing distrust of ‘mainstream’ outlets, some political leaders leaned into scientific distrust, adopting “destructive denialist approaches” which questioned the existence of the virus or argued that it would “work itself out” (Greer et al. 2021). Then president of Tanzania, John Magufuli (who subsequently died from coronavirus), argued that “Coronavirus, which is a devil, cannot survive in the body of Christ … It will burn instantly,” and was “certain that many Tanzanians believe, that the corona disease has been eliminated by God” (BBC). Fringe organizations formed in the late 2010s went as far as to argue that the pandemic was a conspiracy led by a team of global elites who wanted to control populations. Vaccines and facemasks became politically divisive symbols and created social disruptions often fuelled by political discourse and widespread distrust of a vaguely defined group of elites. Media outlets arguably exacerbated the problem where balanced and nuanced coverage considering the difficult trade-offs and policy responses gave way to more dramatic twenty-four-hour catastrophe reporting that encouraged mass hysteria and political division (Boettke and Powell 2021; Rosi et al. 2021; Furton 2023). A common narrative of ‘following the science’ failed to incorporate the longer-term financial constraints and need for well-designed policy responses that did not impose undue future burdens on the taxpaying population. This one-dimensional reporting left those who already distrusted politicians and political institutions feeling further abandoned from what they saw as legitimate concerns about long term economic burdens and an enduring ratchet effect created by expensive short term lockdown measures (see Higgs 1987; Furton 2023). In democratic countries there was an additional need to consider the will of the people whose economic and physical freedom was becoming increasingly restricted. The real time speed at which responses were required to prevent spreading not only left some segments of the public feeling ignored but also opened the door for interest groups and politically connected elites to mobilize and benefit from the chaos (Furton et al. 2023; Miozzi and Powell 2023). While lockdown speed and measures were partly determined by ideology and economic freedom (McCannon 2021; McCannon and Hall 2021; Miozzi and Powell 2023; Hallerberg and Wehner 2025), the exposure of politically affiliated individuals profiting from lucrative pandemic contracts led entire populations to feel as if there were taken advantage of by the same governments they distrusted before covid (for example the high profile case of Baroness Michelle Mone's PPE Medpro in the UK).

In short, polarized and distrustful societies with strongly held beliefs about a pandemic that was, at its core, a public health issue were further fuelled by clickbait catastrophe reporting that led to further polarization and distrust. On one extreme, no public officials could be trusted, and mainstream media reporting was ‘fake news’ while, on the other extreme, public official's sole focused was on ‘saving lives’ and media reporting was gospel. In short, there was “no way to understand the different responses to COVID-19 and their effects without understanding policy and politics” (Greer et al. 2020; Furton et al. 2023).

Theory

The widespread nature of the pandemic meant that it affected countries across a wide range of political regimes, institutional frameworks, and social structures. If political institutions did play a role in financial decision-making, what should we expect from theoretical contributions in previous literature?

Regime Type

One of the unique features of the coronavirus pandemic was a need to compel the general population to comply with restrictions on their individual freedoms to ensure collective public safety and slow down the spread of the virus while a vaccine was being developed. In this context, non-democratic regimes may have had some advantages over their democratic counterparts. In the formal framework introduced by Wintrobe (1990), nondemocratic regimes ensure their long-term existing using a convex combination of loyalty and repression based on the marginal cost of each option – they will make use of whichever option is cheaper at the margin. In practice, nondemocratic regimes will have already established some degree of effective repression by 2019 that could be enforced at comparatively lower marginal costs than democratic regimes. This allows nondemocracies to benefit from a lower reliance on spending money to compel loyalty (which may be expensive with high levels of social distrust). It has also been argued that authoritarian regimes may be able to react more quickly to national crises given their more streamlined decision-making process (Nelson 2021; Hallerberg and Wehner 2025). Contrasting this with democratic regimes where repression was expensive, undemocratic, and unpopular, policymakers may have opted instead to purchase more expensive marginal compliance from their citizens or well-organized interest groups (Olson 1965; Boettke and Powell 2021). While nondemocratic regimes may have an advantage when it comes to quick and efficient enforcement of lockdowns, it has been argued that democratic regimes may be more financially responsive (or generous). In Olson (1993) and McGuire and Olson (1996) they argue that nondemocratic regimes will provide only the required level of public goods to remain in power indefinitely keeping any residual wealth for themselves. We should therefore expect that, while democratic governments may take longer to formulate policy responses to the coronavirus pandemic, over time they will be more generous than nondemocratic governments especially for broad based spending that will appeal wide swaths of the voting population (see Elgin and Yasar 2024). It has also been argued that a more informed and politically active electorate creates stronger incentives for governments to respond to citizen needs (Besley and Burgess 2002). These theoretical expectations were validated using government responses to natural disasters in sixteen major Indian states over the 1958–1992 period with the caveat that “a number of other factors, including turnout, political competitiveness, polarization and the timing of elections affect how governments respond” (Besley and Burgess 2002). More recently, Elgin and Yasar (2024) found that fiscal responses increased over time in democratic countries who also put a greater emphasis on broad based non-health spending.

As argued above, the coronavirus crisis “emerged at a time of fundamental concern about the global state of democracy” (Flinders 2021). Trust in government in democratic countries had been diminishing since 2008, limiting their perceived legitimacy and the willingness of citizens to voluntarily comply with their advice. Democratic governments are also less able to make use of repressive measures to enforce full compliance with lockdowns and restrictions on personal freedoms than their nondemocratic counterparts. Taking advantage of already existing and effective repressive measures to ensure full compliance at a lower marginal cost we may expect that public compliance is more expensive in democratic regimes especially those where trust in public authorities is low. Combining theoretical expectation with political circumstances at the time of the pandemic we should expect that democratic governments will outspend their nondemocratic counterparts (ceteris paribus) to achieve the same results in terms of general compliance from populations, especially in the context of widespread social distrust discussed above. With respect to empirical results, in their investigation of fiscal consequences of natural disasters, Noy and Nualsri (2011) find evidence that more accountable governments respond more aggressively, and countercyclically, to natural disasters.

$H_{01}$ : Democratic regimes will engage in larger coronavirus financing that nondemocracies.

With respect to institutional apparatus, previous literature has found that the financial decision making process in government can be affected by a wide range of factors including electoral competition, the size and type of actors in the budget process, and the electoral business cycle (Barro 1973; Barro 1979; Johnsen, Shepsle and Weingast 1981; Roubini and Sachs 1989; Alesina and Perotti 1996; Poterba 1996; Hallerberg and von Hagen 1999; Kontopoulos and Perotti 2002; Chang and Tsebelis 2004; Hallerberg, Strauch and von Hagen 2009; Franzese 2010; Hallerberg and Ylaeoutinen 2010; Wehner 2010a, 2010b; Armingeon 2012; Boettke and Powell 2021; Nelson 2021; Wehner and Mills 2021; Furton, Rizzo and Harper 2023; Hallerberg and Wehner 2025).

Electoral Competition and the Political Business Cycle

Electoral competition and the political business cycle literature argues that fiscal policy and public investment decisions will strategically consider political timing (Nordhaus 1975, Rogoff and Sibert 1988, Brender and Drazen 2005). Governments are expected to overspend in election years to buy the support of myopic voters while paying for it in the sobering aftermath after having secured another term in office. Empirical evidence has been mixed with some studies confirming a political business cycle affecting real macroeconomic outcomes. Specifically, past research has shown that the size and effectiveness of strategic fiscal manipulation by incumbent governments will depend on the expected likelihood on an incumbent victory, the size and fragmentation of executive and legislative branches of government and, the maturity of democracies (Zupan 1990, Besley and Case 1995, Franzese 2002, Brender and Drazen 2005, Wehner 2013, Seiferling 2020). A key part of these findings is the assumption that voters will reward fiscal manipulations based on the timing of those decisions and be sufficiently myopic to forget about the early years of tighter budgets. In the context of the coronavirus pandemic the political business cycle literature suggests that financial decisions made by democratic governments in response to the pandemic will take into account the date of the next election where incumbents facing re-election during the pandemic should be expected to outspend incumbents who face re-election dates later in the future (Pulejo and Querubin 2021, Lokshin, Rodriguez-Ferrari and Torre 2024).

$H_{02}$ : Government who are facing elections during the pandemic will engage in larger coronavirus financing than government who do not face elections during the pandemic.

For incumbent facing re-election in future years expectations become trickier. Theory suggests that the further way an election is the less generous will be the current fiscal package, however, these expectations are based on a normal business cycle. Political business cycles in the context of an exogenous shock are less well developed in the theoretical literature, but practice has shown that governments tend to take on a ‘whatever it takes’ approach perhaps with the expectation of being rewarded for generous spending programs, even in the years after the crises ends. In the context of the pandemic, large scale spending programs were largely applauded by the public. Ministers of finance across the world giddily revealed expansionary measures to ensure households could continue to buy food and pay their bills during the lockdown while ensuring that public health infrastructure was not overloaded. The most politically prized (and economically criticized) of these was the use of stimulus payments to households. For example, the mailing of $1,200 checks (followed by $600 and $1,400) in the USA or ¥100,000 in Japan were largely applauded by the public. The immediate payoff from the announcement of generous cash injections led to a “sharp increase” in the popularity of national leaders (Flinders 2021). In short, for governments with sufficient fiscal space and financial depth to access cheap money, “what is not to like about the stimulus payments?” (Romer 2021).

In the context of the political business cycle the downside to cash injections come from the sobering aftermath and consequences of spending large amounts of borrowed money. For example, in the aftermath of the pandemic, inflation rates soared by 2023, widespread fraud was exposed from a “rapid implementation of large new spending and loan programs that came with an unusually high risk” (UK NAO 2023), and debt service payments grew to around 10% of government revenue in developed economies like the UK. Once the spending spree ends and the financial consequences of rapid and large packages are soberly reflected upon, an incumbent government's post pandemic popularity will likely dove. From this, governments who are facing re-election several years after the onset of the pandemic may consider that voters will punish them for damaging public finances and imposing financial burdens on taxpayers/consumers. This leads to the hypothesis that governments whose election dates are further away may take this sober reflection into account when making financial decisions.

$H_{03}$ : The further away an election year is, the less generous will be coronavirus spending packages.

Executive Fragmentation and Polarization

The literature on executive fragmentation in budgeting takes its roots from the Tragedy of the Commons problem (or, ‘law of 1/N’). There are two main findings that come from this strand of literature. First, where the number of meaningful actors involved in the budget making process increases, there is a greater need to make conciliatory concessions to all actors who want to ‘bring home the bacon’ to their home constituencies. A more saturated number of actors involved in the budgeting process should lead to a higher propensity to overspend as compared with a single decisive actor. Secondly, due to the bargaining process that is required with large number of actors, fiscal reactions should be expected to take longer amount of time resulting in a delayed financial response (Wehner 2010b) which can be further exacerbated by ideologically division or polarization (Roubini and Sachs 1989, von Hagen and Harden 1994, Alesina and Perotti 1996). In the context of a pandemic where discretionary financial decisions need to be made in a timely manner it could be expected that countries with a more saturated number of meaningful actors in the budget process may design larger rescue packages but with bigger lags relative to countries with a unified and strong executive who should be able to make and implement swift policy measures (ceteris paribus).

$H_{04}$ : Highly concentrated and unified executives will engage in larger coronavirus financing than fragmented ones early in the pandemic but, over time, more fragmented executives will engage in larger coronavirus financing than unified ones.

Trust and Effectiveness in Public Institutions

From the discussion above, widespread distrust in public authorities was a unique feature of the coronavirus pandemic. While nondemocracies were more able to make use of repression in response to social backlash against restrictions, democratic governments arguably faced higher marginal costs to achieve the same results in terms of compliance with restrictions in individual freedoms ( $H_{01}$ ).

In an early review of the literature Devine et al. (2021) provide a comprehensive summary of the role that political trust played in the implementation, compliance, mortality, risk perception and consequences of the coronavirus pandemic. The results from a survey of twelve papers suggested that compliance was higher where there was higher levels of trust in those delivering the guidelines and policy messages along with their political affiliation (Van Bavel et al. 2020; Devine et al. 2021; Goldstein and Wiedemann 2022). Further evidence from Han et al. (2021) found that governments who were “perceived as well organized, disseminating clear messages and knowledge on COVID-19, and perceived fairness were positively associated with trust in government.” These results (along with those from Lazarus et al. 2020) suggest that more trust in those delivering advice leads to higher levels of voluntary compliance.

From this, It should be expected that in countries where citizens have high levels of trust in public authorities, regardless of regime type, there would be high levels of voluntary compliance with the advice given even if these put restrictive measures on personal freedoms, whereas; in countries where citizens had low levels of trust in public authorities more costly measures including the enforcement of new laws and/or educational campaigns would be required to ensure compliance. From this, we should expect that financing covid in distrustful societies will be costly.

$H_{05}$ : Governments with lower levels of trust will need to engage in larger coronavirus financing than governments who are widely trusted.

It is also possible that the financial effect of trust in public institutions will depend on their effectiveness (Han et al. 2021). Where public institutions are effective based on high quality professional civil service and low levels of political interference, we should expect effective responses to covid especially where civil servants have high degrees of independence from political pressures (for example, when it comes to public procurement for lucrative covid contracts). From this perspective we should expect that more effective governments will be more efficient when making public finance decisions and will be outspent by less effective governments. The effect of trust ( $H_{05})$ may also depend on whether public institutions are effective or not: in countries where levels of social trust are high and public institutions are effective, we should expect that the cost of implementing effective measures to be cheaper than in countries where social trust is low, and institutions are effective. Equally, in countries where institutions are ineffective and social trust is low, we should expect that the implementation of covid measures to be more expensive as political interference leads to distortions and higher marginal costs.

$H_{06}$ : The effect of trust on coronavirus financing will depend on the effectiveness of public institutions.

These six hypotheses are tested using a globally representative panel of countries using a combination of unique coronavirus datasets drawing on both micro level survey data from the Wellcome Trust and the IMF's Database of Fiscal Policy Responses to Covid 19.

The Data

Measuring financial responses to the coronavirus from a globally representative sample of countries is not an easy task. Fortunately, the IMF built a unique Database of Fiscal Policy Responses to Covid 19 which covers financial measures taken by governments on a quarterly basis over the June 2020-July 2021 period for 192 countries. The database breaks down financial measures into categories of government expenditure (health and non-health), traditional financial contracts (loans, equity injections) and contingent liabilities (guarantees, quasi fiscal operations). One unique feature of the empirical strategy in this paper is the measurement of financial responses on a per capita basis. While past empirical work (Elgin and Yasar 2024) has relied on fiscal responses as a percentage of GDP, the use of this denominator overlooks the fact the size of a fiscal response should depend on the number of potential coronavirus hosts after controlling for other factors. Measuring financial responses on a per capita basis not only provides a more comparative metric, but has the additional benefit of easy interpretation for the parameter estimates in the regressions below – a unit change in each independent variable tells us how much more or less government spent per person on each financial option (ceteris paribus).

Figure 1 below shows the variance across countries for top twenty and bottom twenty countries. Considering that these financial responses are measured on a per capita basis the differences are quite stark with all bottom 20 countries (who took on some measures¹) spending under $100 per capita and top 20 advanced economies like the US, Singapore, Australia, Ireland New Zealand, Canada, Switzerland, Japan, the UK and Germany spending over $6,000 per capita (up to over $15,000 in the case of Germany, the US, and Japan).

Figure 1.

Top 20 and Bottom 20 Coronavirus Financial Responses (Spending and Financial Risk).

Figure 2 below shows the unconditional relationship between government financial responses over regime type (level of democracy, based on Economist EIU data). Because a significant amount of the variance in covid financing will be explained by country wealth, the scatterplot is color coded across high income economies (red), emerging markets (green) and low-income economies (dark blue). There does appear to be a cluster of advanced economies with democracy scores above six who engaged in extensive health and non-health spending while also being more willing to take on fiscal risk in the form of financial contracts and contingent liabilities. Although it is difficult to separate the wealth effect from a regime type effect there is some evidence here that high income nondemocracies underspent their democratic counterparts on a per capita basis.

Figure 2.

Per Capita Financial Responses to Covid: Regime Type.

The unconditional evidence fort a political business cycle is less clear. In Figure 3 below it appears that governments who faced elections during the pandemic (0 on the x-axis) did not engage in larger spending programs or increase their willingness to take on fiscal risk. In fact, the unconditional data suggest that governments with election dates further into the future were more willing to engage in generous financing although this relationship could be explained by economic and epidemiological factors.

Figure 3.

Per Capita Financial Responses to Covid: Electoral Cycle.

Using the Herfindahl index as a proxy, Figure 4 shows the unconditional relationship between per capita covid financing and fragmentation (a score of 1 on the x-axis represents a unified executive). While Figure 4 does not consider timing (this will be explored further as the dataset covers five quarters), there is some indication that unified executives outspend more fragmented ones. From the theoretical discussion above we would only expect this to take place in the early stages of the pandemic due to a bargaining lag with more fragmented executives. Over time, we should expect that financing from more fragmented executives will overtake unified ones (this will be examined further in the analysis section).

Figure 4.

Per Capita Financial Responses to Covid: Government Fragmentation.

The unconditional relationship between per capita covid financing and average levels of trust in government is shown below in Figure 5. As with the previous figures it is difficult to disentangle the country wealth effect from a potential trust effect but, it does appear that there is a potentially higher cost for high income economies as we move along the x-axis (from 1 = a lot of trust to 4 = no trust at all).

Figure 5.

Per Capita Financial Responses to Covid: Social Trust.

With respect to right hand side variables there are several macroeconomic and financial factors which need to be considered before making any meaningful cross-country comparisons across political regimes, social trust, and institutions. Macroeconomic factors, including annual and quarterly GDP per capita and unemployment data were sourced from the OECD and IMF World Economic Outlook (WEO). Because policymakers will consider changes in unemployment rates during lockdowns (rather than the absolute number), these were first differenced over the sample period. As in Romer and Romer (2019), fiscal space is proxied using gross debt to GDP ratios sourced from WEO. To avoid endogeneity issues fiscal space is measured using 2019 debt levels (as a percentage of GDP) - a country's fiscal space available to manage covid will depend on the pre-covid size of debt not the debt accumulated to finance pandemic spending. Governments should also be limited by government effectiveness – ineffective governments may be less likely to have access to the resources to make meaningful large-scale interventions or to instill sufficient confidence in financial markets to acquire cheap new financing. Government effectiveness is measured using the World Bank World Governance Indicators (WGI)'s government effectiveness which “captures perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies” (Kaufmann, Kraay and Mastruzzi 2010). With respect to financial determinants, central banks made large purchases of government debt in primary and/or secondary markets to finance government rescue packages which has been ignored in some previous empirical analysis of covid spending. This is measured using central bank claims on central government as a percentage of GDP which was sourced from the IMF's International Finance Statistics (IFS).² Financial depth may influence the degree to which government can use traditional financial channels to inject liquidity into firms and households. This is proxied using private credit by deposit money banks (as a percentage of GDP) and sourced from IFS. Policymakers would have also considered epidemiological and demographic information including the number of coronavirus cases and deaths (high levels of infection will need higher immediate cash injection responses and higher investment in health infrastructure) and population density. Epidemiological information including coronavirus cases and deaths was sourced from the World Health Organization (WHO). Composition of the workforce may also play a role in governments response where countries with higher proportions of workers in services can lead to higher interventions as these workers hunker down in their homes. To control for this, the proportion of total workforce working in services is sourced from World Banks World Development Indicators (WDI).

Regime type data was sourced from Freedom House political rights and civil liberties and the Economist Intelligent Unit (EIU) democracy index. Both indicators are used in the regression framework to ensure robustness of the results. Other political data was sourced from the Database of Political Institutions (DPI, Cruz, Keefer and Scartascini 2021). Unfortunately, DPI data was only available until 2020 which required manual coding to fill in relevant series for 2021. Variables related to changes in, or ideology of, government were left unchanged if there were no executive elections that year and/or if there was an election that resulted in a victory for the incumbent. In cases where an executive election did take place in 2020 or 2021 and there was a change in government (i.e., in the USA), the data was manually recoded to account for this change. Social trust data was sourced from the Wellcome Global Monitor Wave 2 which asked respondents how much they trusted the national government in their country (1 – “a lot”; 2 – “some”; 3 – “not much”; 4 “not at all”). Country level means were computed from 1,000 individual responses per country and trust scores were inverted for ease of interpretation in the regression results below (1 – “not at all”,…, 4 – “a lot”).

In cases where only annual data was available (WEO, Freedom House, WDI, DPI) quarterly data was linearly imputed from annual over the short sample period.

Specification and Results

From the discussion above we are interested in explaining the variance in both the size and composition of government financial responses to the coronavirus pandemic. Beginning with total cash injections (increase in deficit) government responses will depend on macroeconomic conditions, state capacity, fiscal space, borrowing costs, automatic stabilizers, central banks willingness to purchase government debt, demographic factors, epidemiological information available to government, and the political/institutional context discussed above (regime type, electoral business cycle, executive structure, and social trust):

\begin{aligned} Δ δ_{j t} = α_{j} + X β + Z γ + ω (r e g i m e_{j t}) + φ (e b c_{j t}) + τ (e l e c_{j t}) + θ (e x e c_{j t}) \\ + δ (t r u s t_{j}) + ν (e f f e c t i v e_{j t}) + η (e f f e c t i v e_{j t} * t r u s t_{j} \end{aligned}

(1)

Where:

$Δ δ_{j t}$ is the per capita above the line (change in deficit) measures taken in response to the coronavirus pandemic in country j at time t.

$t o t a l_r e s p o n s e_{j t}$ is the per capita sum of all financial measures taken in response to the coronavirus pandemic in country j at time t .

$X$ is a matrix of macroeconomic and financial variables (GDP per capita, unemployment, automatic stabilizers³, financial depth, labor force composition).

$Z$ is a matrix of epidemiological and demographic variables (covid cases, deaths, population density).

$r e g i m e_{j t}$ is the regime type of country j at time t (based on Freedom House and Economist EIU).

$e b c_{j t}$ considers the electoral budget cycle (years from next election).

$e l e c_{j t}$ equal 1 where an election takes place in country j at time t (0 otherwise).

$e x e c_{j t}$ is a dummy variable representing executive ideology (1 = right).

$t r u s t_{j}$ measures average levels of social trust in government/public authorities/vaccines in country j in 2020 (based on Wellcome trust survey data).

$e f f e c t i v e_{j t}$ measures the quality and professionalism in the civil service of country j at time t (based on World Governance Indicators).

And $α$ , $β, γ$ , $ω, φ, τ, θ$ , $δ, ν, η$ are unknown parameters to be estimated.

Equation (1) can be estimated separating health and non-health financial responses to covid with the same set of right-hand side variables:

{\begin{matrix} Δ δ_{j t}^{h e a l t h} = α_{j} + X β + Z γ + ω (r e g i m e_{j t}) + φ (e b c_{j t}) + θ (e x e c_{j t}) + \\ δ (t r u s t_{j}) + ν (e f f e c t i v e_{j t}) + η (e f f e c t i v e_{j t} * t r u s t_{j}) \\ Δ δ_{j t}^{n o n h e a l t h} = α_{j} + X β + Z γ + ω (r e g i m e_{j t}) + φ (e b c_{j t}) + θ (e x e c_{j t}) + \\ δ (t r u s t_{j}) + ν (e f f e c t i v e_{j t}) + η (e f f e c t i v e_{j t} * t r u s t_{j}) \end{matrix}

(2)

Lastly, financial responses that increase government financial risk but do not affect deficits (increases in financial claims including equity injections, asset purchases, loans) and contingent liabilities are modeled with the same set of right-hand side variables:

\begin{aligned} Δ f a_{j t} = α_{j} + X β + Z γ + ω (r e g i m e_{j t}) + φ (e b c_{j t}) + θ (e x e c_{j t}) + δ (t r u s t_{j}) \\ + ν (e f f e c t i v e_{j t}) + η (e f f e c t i v e_{j t} * t r u s t_{j}) \end{aligned}

(3)

\begin{aligned} c l_{j t} = α_{j} + X β + Z γ + ω (r e g i m e_{j t}) + φ (e b c_{j t}) + θ (e x e c_{j t}) + δ (t r u s t_{j}) \\ + ν (e f f e c t i v e_{j t}) + η (e f f e c t i v e_{j t} * t r u s t_{j}) \end{aligned}

Where:

$Δ f a_{j t}$ is the per capita below the line (increase in financing) measures taken in response to the coronavirus pandemic in country j at time t, and,

$c l_{j t}$ is the per capita increase in contingent liabilities in response to the coronavirus pandemic in country j at time t .

Equations (1)–(3) were estimated using a random effects model with a GLS estimator for an unbalanced panel of 80 countries over a five-quarter period (Q3 2020–3 2021). Results for the global sample of countries are shown below for the determinants of cash injections and traditional financing/contingent liabilities are shown below in Tables 1 and 2, respectively.

Table 1.

Results: Cash Injections per Capita (Health and Non-Health Expenditure).

Dependent var	Total Deficit (per capita)		Health Expenditure (per capita)		Non-health Expenditure (per capita)
	(1) Baseline	(2) Interaction	(3) Baseline	(4) Interaction	(5) Baseline	(6) Interaction
GDP per capita	0.09***(0.01)	0.09**(0.01)	0.01***(0.002)	0.01***(0.002)	0.08***(0.01)	0.09***(0.01)
$Δ$ Unemployment	59.74(72.00)	41.10(72.94)	30.80(25.94)	23.26(26.13)	28.62(62.73)	18.84(63.65)
Fiscal space (Gross debt/GDP 2019)	−8.37(5.17)	−10.40**(5.29)	−1.35(1.06)	−1.89*(1.06)	−5.03(4.80)	−6.50(4.92)
Central bank claims on government (QE)	57.41***(12.60)	58.01***(12.74)	8.15***(2.64)	8.01***(2.62)	44.03***(11.60)	43.84***(11.79)
Financial depth	3.66(3.73)	4.43(3.79)	−0.26(0.79)	0.01(0.79)	2.41(3.59)	3.05(3.67)
% empl services	−7.80(12.37)	−5.43(12.58)	−1.19(2.54)	−0.36(2.53)	−6.58(11.65)	−4.62(11.88)
Government effectiveness	−657.64*(343.35)	675.17(832.32)	−6.66(73.08)	422.20**(166.76)	−624.73**(317.83)	503.16(783.59)
Social trust	227.81(322.31)	192.17(325.67)	−44.63(66.93)	−54.55(66.23)	250.57(303.65)	209.77(308.44)
Effectiveness*trust		−569.86*(321.43)		−189.06***(66.48)		−482.12(302.33)
Democracy (EIU)	119.74(123.30)	142.68(124.79)	−7.39(26.11)	1.75(25.99)	129.52(115.11)	146.69(116.76)
Executive ideology (right)	−205.52**(100.13)	−202.75**(99.58)	−49.20(37.35)	−51.94(36.97)	−205.52**(90.27)	−207.60**(89.82)
Herfindahl Index	672.39*(394.21)	412.97(453.04)	119.37(82.39)	21.14(116.62)	554.93(368.89)	422.40(422.62)
Herfindahl Index*time		63.48(68.70)		23.31(25.76)		23.44(61.70)
Election year	112.49(126.82)	117.39(127.69)	57.61(42.31)	56.86(42.38)	44.77(110.32)	40.68(111.32)
Time until next election	24.19(26.98)	22.53(26.86)	11.95(10.12)	12.04(10.03)	21.82(24.10)	21.94(24.02)
Covid Cases (cumulative)	0.06(0.03)*	0.06*(0.03	0.04***(0.01)	0.04***(0.01)	0.01(0.03)	0.01(0.03)
Population density	−0.38(0.52)	−0.41(0.53)	0.07(0.10)	0.07(0.10)	−0.41(0.49)	−0.41(0.50)
Time (linear)		83.42(60.17)	39.59***(12.18)	23.20(22.25)	112.44***(29.96)	92.77*(54.48)
Obs	345	345	327	327	327	327
Countries	80	80	79	79	79	79
Quarters	5	5	5	5	5	5
R²	0.70	0.70	0.44	0.47	0.67	0.67

Standard errors in parenthesis.

*** p < 0.01; ** p < 0.05; * p < 0.1.

Table 2.

Results: Traditional Financing and Contingent Liabilities per Capita.

	Traditional financing		Contingent liabilities
	(7) Baseline	(8) Interaction	(9) Baseline	(10) Interaction
GDP per capita	0.01***(0.006)	0.01**(0.006)	−0.01(0.01)	−0.01(0.01)
$Δ$ Unemployment	−61.87(105.09)	−25.92(108.67)	78.40(65.55)	72.49(65.71)
Fiscal space (Gross debt/GDP 2019)	2.09(3.27)	2.89(3.22)	7.74(8.05)	6.86(8.29)
Central bank claims on government (QE)	−14.17(9.04)	−11.89(8.94)	46.60***(17.37)	51.27***(17.67)
Financial depth	1.15(2.32)	0.34(2.34)	13.47***(5.81)	13.16***(5.87)
% empl services	−4.59(8.36)	−5.08(8.21)	18.21(23.83)	15.16(24.29)
Government effectiveness	57.96(218.69)	−729.07(501.11)	324.11(490.27)	−289.51(1447.03)
Social trust	56.65(200.50)	−18.07(201.90)	48.68(607.17)	−37.23(634.82)
Effectiveness*trust		350.82*(209.24)		255.41(555.10)
Democracy (EIU)	−10.68(78.29)	−47.76(81.21)	109.37(191.17)	102.12(193.09)
Executive ideology (right)	−45.27(120.77)	−51.26(121.11)	16.76(86.89)	26.06(86.73)
Herfindahl Index	−32.16(251.62)	185.02(385.06)	10.58(573.82)	−236.91(603.10)
Herfindahl Index*time		−74.11(88.31)		86.97(58.49)
Election year	−125.61(143.05)	−115.67(143.57)	433.54***(124.28)	474.89***(126.54)
Time until next election	−29.08(33.70)	−27.68(33.99)	26.41(22.67)	24.55(22.60)
Covid Cases (cumulative)	−0.03(0.03)	−0.03(0.03)	0.02(0.02)	0.03(0.02)
Population density	−0.07(0.35)	−0.09(0.34)	1.18(0.84)	1.23(0.86)
Time (linear)	66.53*(40.43)	117.88(72.65)	21.15(29.73)	−40.88(52.71)
Obs	215	215	263	263
Countries	54	53	64	64
Quarters	5	5	5	5
R²	0.30	0.32	0.50	0.50

Standard errors in parenthesis.

*** p < 0.01; ** p < 0.05; * p < 0.1.

Unsurprisingly, two of the most robust determinants of cash injection financial responses are country wealth (GDP per capita) and central bank willingness to purchase government debt (central bank claims on government). Figure 6 gives an idea of the magnitude of differences in health and non-health spending per capita across GDP per capita (6a) and central bank claims on government (6b).

Figure 6.

Cash Injections, GDP per Capita and Central Bank Claims on Government. (a) GDP per Capita. (b) Central Bank Claims on Government.

Considering the results from 1a and 1b in the context of our six hypotheses, we reject $H_{01}$ at conventional levels of significance across all financing options for government. It has been argued that political systems should face different costs when implementing covid measures (especially the ability of governments to enforce lockdowns) and theoretical foundations expect that regimes who have cheaper access to repressive measures should be able to take advantages of them, however, the evidence here suggests that, once we control for other factors, there are no significant differences in cash injections (health or non-health) or in traditional finance and contingent liabilities across regime type.

Moving to $H_{02}$ there is no evidence of significant differences across traditional financial instruments and cash injections when comparing election years with non-election years (consistent with McCannon and Hall 2021), however, there is evidence that election year governments will make significantly higher use of contingent liabilities in response to the pandemic, with election year governments taking on around $475 per capita more risk in the form of contingent liabilities than non-election year governments. There is a convincing causal story for governments in election years to take advantage of the ability to receive political benefits from facilitating the delivery of public goods without increasing the deficit or debt. From this perspective, contingent liabilities offer a tempting short-term opportunity for governments who wish to engage in expansionary measures but do not want to pay the price when communicating the associated increase in governments deficit/debt in an election year. Figure 7 below shows the government's willingness to take on additional risk in the form of contingent liabilities across financial depth and election years. The former would provide a conduit, or partner, for engaging in contingent liabilities (for example, the Bounce Back Loan scheme in the UK facilitated by commercial banks).

Figure 7.

Per Capita Financial Responses to Covid: Election Years and Contingent Liabilities.

There is little support for $H_{03}$ at conventional levels of statistical significance (The further away an election year is, the less generous will be coronavirus spending packages). This could be due to the urgent nature of spending requirements in response to an exogenous shock where politicians reacted quickly taking into account current circumstances (for example if an election was taking place that year $(H_{02}$ ) but not consider future electoral repercussions).

With respect to $H_{04}$ (highly concentrated and unified executives will spend more in early pandemic years but, over time, more fragmented executives will out spend unified ones), there is evidence that unified executives will outspend fragmented ones in early 2020 with a magnitude of $672 per capita, but no evidence that fragmented executives will outspend unified ones over time. This could be due to the short sample period (five quarters) which, in political time, is relatively short and may not be sufficient to capture higher spending from bargaining among fragmented executives. Figure 8 shows the predicted relationship between government fragmentation and covid spending (total deficit, health and non-health).

Figure 8.

Per Capita Financial Responses to Covid: Government Fragmentation.

The findings for $H_{05}$ and $H_{06}$ are more interesting and pronounced once we consider the interactive effect of government effectiveness and trust. Starting with $H_{05}$ (governments who faced lower levels of trust from the general public will need to spend more than governments who are widely trusted), there doesn’t appear to be a statistically significant effect of average levels of trust (based on averaged micro-level Wellcome Trust data) on covid financing. However, once government effectiveness data are incorporated into the model, the interactive relationship becomes statistically significant in terms of the total effect on deficits and health spending. In columns 1 and 5, which do not consider an interaction effect, more effective governments spend significantly less per capita on covid than ineffective ones suggesting that these governments are more efficient in their spending decisions which are less likely to face political pressure. The magnitude of this effect is large with a one-unit change in government effectiveness leading to a reduction of $658 per capita total deficit spending on covid. While low levels of trust are more costly for effective governments ( $H_{05})$ , this is not the case for ineffective governments. Once we consider the interaction between trust and effectiveness, the financial implications diverge - ineffective governments outspend effective ones and this effect diverges as trust increases. For example, as trust increases, ineffective governments increase their spending perhaps taking advantage of social trust when making financial decisions that may be less effective in implementation or allow for higher political rents. On the contrary, effective governments tend to spend less as trust levels increase suggesting that they are able to take advantage of social trust when minimizing the cost of financing a response to the coronavirus pandemic. These findings suggest that the most prudent governments are those who are effective and highly trusted. Figure 9 shows the conditional relationship between trust and covid spending (total deficit (a) and health spending (b)).

Figure 9.

Per Capita Financial Responses to Covid: Government Effectiveness and Social Trust. (a) Total Deficit Spending, (b) Health Spending.

Reflecting on the definition of government effectiveness, ineffective government who deliver low quality public services and have low levels of (overall and procedural) credibility and independence from political pressure are more likely to fall victim to political pressure for unnecessarily costly coronavirus financial packages. In short, these findings suggest that highly trusted effective governments were able to capitalize on trust premiums to minimize covid spending while highly trusted ineffective governments took advantage of trust premiums to maximize the size of their spending packages.

Robustness Check: Automatic Stabilizers

As explicitly noted by the IMF, it is difficult to derive conclusions based on direct comparison with this data given the range of country specific circumstances leading to wide variations in the way covid was financed. Many of these factors were controlled for in Tables 1 and 2, however, automatic stabilizers pose a unique challenge for at least two reasons. Firstly, automatic stabilizers are not observable and need to be estimated. We know that these played a key role during covid as a large range of policies were already in place relating to areas most impacted by covid including health and employment. These policies are vast and vary significantly over time and across countries making them not possible to directly measure. Past estimates of how government revenue or expenditure react to the business cycle have been found to suffer from endogeneity problems, an inability to separate discretionary actions from automatic stabilizers, and results being dependent on the choice of cyclical indicators. This challenge is made even more difficult with a global sample of countries that includes low income and emerging markets. In the case of developing/ emerging market economies, Fatas and Mihov (2001) noted that, “it is not clear whether one can describe these [less developed] economies as having a well-defined business cycle … which exhibit short-run fluctuations around a balanced growth path.” Secondly, countries with large automatic stabilizers will face less pressure to engage in discretionary spending depending on the design of the stabilizers which will vary across countries. Where automatic stabilizers provide strong insurance mechanisms during a pandemic there is less of a need for policymakers to take discretionary measures but, it is not easy to accurately measure these for a global sample of countries. During the coronavirus pandemic, where built it insurance mechanisms should reduce the size of discretionary measures necessary to provide fiscal relief and stimulate supply, we should expect that this will affect discretionary coronavirus financial responses (our dependent variables).

This paper overcomes these challenges by estimating cyclically adjusted balances using long time series data (1980–2019) from the IMF WEO and Government Finance Statistics Yearbook (GFSY) databases and following the methodological framework in Fedelino, Ivanova and Horton (2009). To ensure our estimates are reasonable we verify them against those in the IMF WEO database (for available countries) with highly correlated results with the caveat that these estimates may be less accurate for low-income countries with less well-defined business cycles. For consistency and ease of interpretation automatic stabilizers are measured as a percent of GDP.

The results in Tables 3 and 4 are encouragingly similar to those in 1a and 1b. The sample size is significantly lower due to data limitations for automatic stabilizer estimates but, the main findings (discussed in Section 5) are relatively unchanged. Specifically, regime type plays does not play a role in determining financial responses to the coronavirus pandemic (reject $H_{01}),$ electoral competition continues to play a significant role in governments willingness to increase risk through contingent liabilities ( $H_{02}$ ), there is no evidence of a political business cycle effect ( $H_{03}$ ), there continues to be evidence that unified executives will significantly outspend fragmented ones but no evidence that fragmented governments will, over time, outspend their unified counterparts ( $H_{04})$ . The evidence regarding the effect of government effectiveness and social trust confirms that effective governments will be outspent by ineffective ones which continues to diverge as trust increases ( $H_{05,} H_{06}$ ).

Table 3.

Results: Cash Injections per Capita (Health and Non-Health Expenditure).

Dependent var	Total Deficit (per capita)		Health Expenditure (per capita)		Non-health Expenditure (per capita)
	(1) Baseline	(2) Interaction	(3) Baseline	(4) Interaction	(5) baseline	(6) Interaction
GDP per capita	0.09***(0.02)	0.10**(0.02)	0.01***(0.003)	0.02***(0.003)	0.08***(0.01)	0.10***(0.02)
$Δ$ Unemployment	18.33(109.20)	30.20(111.85)	22.36(39.30)	2.54(39.86)	−8.18(95.37)	−36.01(97.68)
Fiscal space (Gross debt/GDP 2019)	−22.31***(8.05)	−25.25***(8.11)	−3.71**(1.71)	−4.32**(1.71)	−14.28*(7.36)	−16.49**(7.53)
Central bank claims on government (QE)	91.21***(19.74)	91.42***(19.82)	13.28***(4.30)	12.68***(4.29)	66.17***(18.02)	65.29***(18.40)
Financial depth	7.64(6.42)	9.79(6.49)	0.59(1.42)	1.29(1.44)	3.20(6.39)	5.38(6.58)
% empl services	1.81(19.01)	−3.62(19.09)	0.68(3.94)	−0.89(3.95)	2.99(17.82)	−1.95(18.23)
Government effectiveness	−1208.61**(532.60)	1754.13(1646.47)	−75.05(114.12)	738.14**(337.47)	−1109.39**(487.23)	1503.64(1558.66)
Social trust	365.32(620.45)	1140.98(739.04)	−133.42(129.08)	67.32(150.30)	437.68(585.64)	1083.62(694.58)
Effectiveness*trust		−1257.07*(658.96)		−350.37***(137.60)		−1112.17*(624.41)
Democracy (EIU)	91.49(218.04)	198.05(223.70)	−30.38(47.50)	−3.62(48.27)	161.22(204.61)	243.63(211.44)
Executive ideology (right)	−245.87*(139.55)	−236.71*(138.69)	−62.00(52.17)	−63.64(51.47)	−265.85**(126.43)	−269.18**(125.63)
Herfindahl Index	1479.58**(641.14)	1109.58(700.66)	238.27*(141.03)	83.63(182.32)	1223.23**(595.35)	1038.34(652.65)
Herfindahl Index*time		137.49(102.15)		49.52(38.06)		60.98(91.99)
Election year	33.71(168.72)	57.24(168.91)	36.32(57.85)	40.75(57.64)	−32.01(147.70)	−27.41(148.18)
Time until next election	13.04(37.59)	9.18(37.41)	12.67(14.19)	11.97(14.01)	22.31(33.92)	21.77(33.77)
Covid Cases (cumulative)	0.04(0.04)	0.06*(0.04)	0.04***(0.01)	0.04***(0.01)	0.002(0.03)	0.01(0.03)
Population density	−0.37(0.51)	−0.45(0.81)	0.13(0.16)	0.13(0.16)	−0.32(0.75)	−0.36(0.76)
Time (linear)	166.98***(45.63)	54.30(89.42)	51.13***(16.69)	12.80(32.49)	136.90***(41.63)	83.43(80.76)
Automatic Stabilizer (cyclical balances/gdp)	−876.23(951.81)	−733.60(949.62)	−66.86(193.14)	−40.68(192.03)	−1207.92(887.37)	−1093.46(901.00)
Obs	239	239	229	229	229	229
Countries	54	54	53	53	53	53
Quarters	5	5	5	5	5	5
R²	0.68	0.69	0.44	0.47	0.65	0.66

Standard errors in parenthesis.

*** p < 0.01; ** p < 0.05; * p < 0.1.

Table 4.

Results: Traditional Financing and Contingent Liabilities per Capita.

	Traditional financing		Contingent liabilities
	(7) Baseline	(8) Interaction	(9) Baseline	(10) Interaction
GDP per capita	0.02*(0.009)	0.01(0.01)	−0.02(0.02)	−0.02(0.02)
$Δ$ Unemployment	−97.88(156.66)	−9.51(159.23)	127.38(106.69)	105.26(108.01)
Fiscal space (Gross debt/GDP 2019)	5.17(4.81)	6.81(5.19)	9.67(9.99)	7.88(10.30)
Central bank claims on government (QE)	−21.63(13.22)	−16.97(14.41)	39.98*(29.80)	46.40**(22.23)
Financial depth	−1.44(3.54)	−3.04(3.90)	3.37(9.14)	3.25(9.41)
% empl services	−4.89(11.92)	−2.48(12.90)	30.23(36.57)	30.54(37.63)
Government effectiveness (State Capacity)	144.28(316.47)	−1768.36*(1043.74)	82.60(647.26)	−606.51(2164.61)
Social trust	138.19(325.36)	−396.66(443.98)	448.57(863.73)	268.47(1032.47)
Effectiveness*trust		842.10*(432.30)		273.82(836.80)
Democracy (EIU)	41.74(140.19)	−48.21(156.69)	566.79*(312.92)	575.23*(325.21)
Executive ideology (right)	−57.28(158.07)	−61.64(155.94)	4.21(110.00)	25.06(109.86)
Herfindahl Index	45.77(410.43)	540.63(568.08)	409.71(748.11)	98.97(776.59)
Herfindahl Index*time		−145.52(114.18)		119.05(79.47)
Election year	−244.34(189.98)	−253.58(188.25)	486.34***(149.50)	540.72***(152.13)
Time until next election	−34.42(43.45)	−28.50(42.66)	34.96(29.35)	31.65(29.17)
Covid Cases (cumulative)	−0.04(0.04)	−0.05(0.04)	0.02(0.03)	0.03(0.03)
Population density	0.08(0.46)	0.18(0.49)	2.30*(1.14)	2.28*(1.17)
Time (linear)	88.18*(51.18)	199.24*(95.10)	30.36(38.20)	−54.99(70.60)
Automatic Stabilizer (cyclical balances/gdp)	−776.44(575.23)	−953.20(1882.57)	−4340.36***(1394.09)	−4355.77***(1435.70)
Obs	167	167	200	200
Countries	40	40	46	46
Quarters	5	5	5	5
R²	0.31	0.33	0.56	0.56

Standard errors in parenthesis.

*** p < 0.01; ** p < 0.05; * p < 0.1.

Limitations

There are at least three important limitations in this paper. The first is the realistic potential for covid related data to vary in quality and accuracy across countries. Especially in the case of covid cases and death, government's ability to accurately monitor and report accurate data likely varied across countries (see Morris and Reuben 2020). Moreover, the definition of a covid death may have varied across countries, in which case, even perfectly accurate data would not be comparable. Coronavirus financing data may also vary in quality across countries. Ministries of finance and statistical offices who recorded financial measures were working in a challenging environment during the pandemic which may have led to different degrees of accuracy across a global sample of countries. Fortunately, these are control variables in this paper and not of central interest, however, accurate covid data may still have effects on the parameter estimates for economic and political determinants.

A second limitation is the lack of baseline scenarios – what if high income countries took no financial action? While there are a sizeable number of countries who took no financial measures in the IMF database (48), these tend to be developing countries with limited capacity. Combining the first and second limitations it is possible that governments whose financial response to the coronavirus pandemic were small in magnitude were also accompanied by poor data leading to the conclusion that limited government response is the best option to ensure low number of cases and deaths.

Lastly, the microdata for average levels of social trust in government is cross sectional. It would have been useful to understand the dynamics for this indicator in the leadup to, and during, the pandemic. Data on trust in government does exist from the Edelman Trust barometer but for a more limited sample of countries. In short, a comparable global dynamic sample of micro data on trust in government would have been preferential to our cross-sectional alternative.

Conclusion

The coronavirus pandemic hit the world in the midst of a fractured and polarized political environment. Since the GFC in 2008, public trust in government had been on a continuous decline which made it more difficult to credibly advocate scientific solutions for a public health issue. The seemingly one-dimensional focus on scientific considerations fueled by catastrophe reporting and mass hysteria from the media neglected realistic concerns and nuances relating to responsible economic policymaking. To avoid imposing unnecessary burdens on future taxpayers and/or the creation of a Higgsian ratchet effect where the financial or regulatory covid responses lingered long into the future it was important that policymakers took measures which were credible and effective in the short term, didn’t enrich political elites, and minimized any expansion the government's balance sheets (long term effects). In this precarious environment, publicly advocated solutions became entangled with politics and political discourse leading to a question of what role, if any, political and institutional factors played in determining what financial measures government took in response to the coronavirus pandemic and how responsible, expensive and, long lasting they were.

Building on theoretical expectations from the political economy literature this paper derived a series of six testable hypotheses drawing on a global sample of eighty countries covering the Q3 2020–Q3 2021 period. The results suggest that politics did play a role in determining financial responses with governments placing a greater emphasis on short term circumstances due to the immediacy of decision making but, forward-looking strategic measures did not play a significant role. The main results in this paper suggest that i) election year governments will make significantly higher use of contingent liabilities in response to the pandemic with election year governments taking on around $475 per capita more risk in the form of contingent liabilities (compared with non-election year governments), ii) unified executives outspent fragmented ones in early 2020 at a magnitude of $672 per capita (but no evidence that fragmented executives will outspend unified ones over time), iii) more effective governments spent less per capita on coronavirus responses than ineffective ones, the magnitude of this effect being notably large - a one unit change in government effectiveness led to a reduction of $658 per capita total deficit spending, and, iv) once we consider the interaction between trust and effectiveness, the financial implications diverge - ineffective governments outspend effective ones and this effect diverges as trust increases. These results raise questions about the extent to which the financial consequences will persist into the future. This is especially important given the per capita magnitude of covid measures and their lasting impact on government balance sheets (or the calling of contingent liabilities) in a post covid environment where inflationary pressure and rising interest rates have created lasting effects for households and government that are still being played out today. Equipped with a new armada of academic research and evidence perhaps next time will be different.

Footnotes

Acknowledgements

The author would like to thank Ms. Daoyu Sun for her exceptional research assistant work on this paper.

Data Availability Statement

All data and code can be made available for full replicability of the results.

Declaration of Conflicting Interests

The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Mike Seiferling

Use of AI

During the preparation of this work no AI, or AI assisted, tools were used.

Notes

Author Biography

Mike Seiferling is an assistant professor in the School of Public Policy. His area of research is in public finance focusing on the interlinkages between macroeconomics, accounting, finance and political science.

References

Alesina

Perotti

. 1996. “Fiscal Discipline and the Budget Process.” American Economic Review 86 (2): 401–407. http://www.jstor.org/stable/2118160.

Armingeon

2012. “The Politics of Fiscal Responses to the Crisis of 2008-2009.” Governance 25 (4): 543–565. https://doi.org/10.1111/j.1468-0491.2012.01594.x.

Barro

R. J.

1973. “The Control of Politicians: An Economic Model.” Public Choice 14 (1): 19–42. https://doi.org/10.1007/BF01718440.

Barro

R. J.

1979. “On the Determination of Public Debt.” Journal of Political Economy 87 (5): 940–971. https://doi.org/10.1086/260807.

Besley

Burgess

. 2002. “The Political Economy of Government Responsiveness: Theory and Evidence from India.” The Quarterly Journal of Economics 117 (4): 1415–1451. https://doi.org/10.2139/ssrn.319012.

Besley

Case

. 1995. “Does Electoral Accountability Affect Economic Policy Choices? Evidence from Gubernational Term Limits.” The Quarterly Journal of Economics 110 (3): 769–798. https://doi.org/10.2307/2946699.

Boettke

Powell

. 2021. “The Political Economy of the COVID-19 Pandemic.” Southern Economic Journal 87 (4): 1090–1106. https://doi.org/10.1002/soej.12488.

Boodram

B. A.

Balaban

Dayton

Falade-Nwulia

Latkin

Miller

Uzzi

. 2022. “A Longitudinal Study of Vaccine Hesitancy Attitudes and Social Influence as Predictors of COVID-19 Vaccine Uptake in the US.” Hum Vaccine Immunother 18 (5). https://doi.org/10.1080/21645515.2022.2043102.

Brender

Drazen

. 2005. “Political Budget Cycles in New Versus Established Democracies.” Journal of Monetary Economics 52 (7): 1271–1295. https://doi.org/10.1016/j.jmoneco.2005.04.004.

10.

Bursian

Furth

. 2015. “Trust Me I am a Central Banker.” Journal of Money, Credit and Banking 47 (8): 1503–1530. https://doi.org/10.1111/jmcb.12282.

11.

Chang

E. C. C.

Tsebelis

. 2004. “Veto Players and the Structure of Budgets in Advanced Industrialized Countries.” European Journal of Political Research 43 (3): 449–476. https://doi.org/10.1111/j.1475-6765.2004.00161.x.

12.

Cruz

Keefer

Scartascini

. 2021. The Database of Political Institutions (DPI): 2020, IDB Open Data. https://doi.org/10.60966/gn3s-at31.

13.

Devine

Gaskell

Jennings

Stoker

. 2021. “Trust and the Coronavirus Pandemic: What Are the Consequences of and for Trust? An Early Review of the Literature.” Political Studies Review 19 (2): 274–285. https://doi.org/10.1177/1478929920948684.

14.

Edelman. 2012. Edelman 2012 Trust Barometer. January 2012.

15.

Elgin

Yasar

. 2024. “Democracy and Fiscal Policy Response to COVID-19.” Public Choice 198 (1-2): 25–45. https://doi.org/10.1007/s11127-023-01107-3.

16.

Fatas

Mihov

. 2001. “Government Size and Automatic Stabilizers: International and Intranational Evidence.” Journal of International Economics 55 (1): 3–28. https://doi.org/10.1016/S0022-1996(01)00093-9.

17.

Fedelino

Ivanova

Horton

. 2009. Computing Cyclically Adjusted Balances and Automatic Stabilizers. IMF Technical Notes and Manuals.

18.

Flinders

2021. “Democracy and the Politics of Coronavirus: Trust, Blame and Understanding.” Parliamentary Affairs 74 (2): 483–502. https://doi.org/10.1093/pa/gsaa013.

19.

Franzese

R. J.

2002. “Electoral and Partisan Cycles in Economic Policies and Outcomes.” Annual Review of Political Science 5 (1): 369–421. https://doi.org/10.1146/annurev.polisci.5.112801.080924.

20.

Franzese

R. J.

Jr.

2010. “Multiple Policymakers: Veto Actors Bargaining in Common Pools.” Paper for the Comparative Politics Workshop, Duke University, February 5, 2010.

21.

Funk

C. D.

Laferriere

Ardakani

. 2020. “A Snapshot of the Global Race for Vaccines Targeting SARS-CoV-2 and the COVID-19 Pandemic.” Front Pharmacol. https://doi.org/10.3389/fphar.2020.00937.

22.

Furton

G. L.

2023. “The Pox of Politics: Troesken’s Trade-off Re-Examined.” Public Choice 195 (1-2): 169–191. https://doi.org/10.1007/s11127-022-01002-3.

23.

Furton

G. L.

Rizzo

M. J.

Harper

D. A.

. 2023. “The Political Economy of Public Health.” Public Choice 195: 1–3. https://doi.org/10.1007/s11127-022-01031-y.

24.

Goldstein

Wiedemann

. 2022. “Who Do You Trust? The Consequences of Political and Social Trust for Public Responsiveness to COVID-19 Orders.” Perspectives on Politics 20 (2): 412–438. https://doi.org/10.1017/S1537592721000049.

25.

Greer

S. L.

King

Massard da Fonseca

Peralta-Santos

. 2021. Coronavirus Politics: The Comparative Politics and Policy of COVID-19. Ann Arbor, MI, USA: University of Michigan Press. https://doi.org/10.3998/mpub.11927713.

26.

Gros

Nowak-Lewhmann

Roth

. 2012. “Has the Financial Crisis eroded Citizens’ Trust in the European Central Bank? Panel Data Evidence for the Euro Area, 1999-2011.” Center for European Governance and Economic Development Research, no./ 124.

27.

Gros

Roth

. 2010. “The Financial Crisis and Citizen Trust in the European Central Bank.” CEPS Working Document No. 334, July 2010.

28.

Hallerberg

Strauch

von Hagen

. 2009. Fiscal Governance in Europe. Cambridge: Cambridge University Press.

29.

Hallerberg

von Hagen

. 1999. “Electoral Institutions, Cabinet Negotiations, and Budget Deficits Within the European Union.” In Fiscal Institutions and Fiscal Performance, edited by Poterba

James M.

von Hagen

Jürgen

, 209–232. Chicago: Chicago University Press.

30.

Hallerberg

Wehner

. 2025. “Regimes, Leaders, and Lockdowns: Who Responded More Quickly to the COVID-19 Pandemic?” Journal of Politics (In Press). http://eprints.lse.ac.uk/id/eprint/127190.

31.

Hallerberg

Ylaeoutinen

. 2010. “Political Power, Fiscal Institutions and Budgetary Outcomes in Central and Eastern Europe.” Journal of Public Policy 30 (Special Issue 01): 45–62. https://doi.org/10.1017/S0143814X09990213.

32.

Han

Zheng

Cristea

Agostini

Belanger

Gutzkow

Kreienkamp

Leander

N. P.

. 2021. “Trust in Government Regarding COVID-19 and its Associations with Preventive Health Behaviour and Prosocial Behaviour During the Pandemic: A Cross-Sectional and Longitudinal Study.” Psychological Medicine 53 (1): 149–159. https://doi.org/10.1017/S0033291721001306.

33.

Higgs

1987. Crisis and Leviathan: Critical: Episodes in the Growth of American Government. New York, USA: Oxford University Press. https://archive.org/details/crisisleviathanc0000higg/page/n3/mode/2up.

34.

Johnsen

Shepsle

K. A.

Weingast

. 1981. “The Political Economy of Benefits and Costs: A Neoclassical Approach to Distributive Politics.” Journal of Political Economy 89 (4): 642–664. https://doi.org/10.1086/260997.

35.

Kaufmann

Kraay

Mastruzzi

. 2010. “The Worldwide Governance Indicators: Methodology and Analytical Issues.” World Bank Policy Research Working Paper No. 5430.

36.

Kontopoulos

Perotti

. 2002. “Fragmented Fiscal Policy.” Journal of Public Economics 86 (2): 191–222. https://doi.org/10.1016/S0047-2727(01)00146-3.

37.

Kreps

Prasad

Brownstein

J. S.

Hswen

Garibaldi

B. T.

Zhang

Kriner

D. L.

. 2020. “Factors Associated with US Adult's Likelihood of Accepting COVID-19 Vaccination.” JAMA Netw Open 3 (10): e2025594. https://doi.org/10.1001/jamanetworkopen.2020.25594.

38.

Lazarus

J. V.

Ratzan

S. C.

Palayew

Gostin

L. O.

Larson

H. J.

Rabon

Kimball

El-Mohandes

. 2021. “A Global Survey of Potential Acceptance of a COVID-19 Vaccine.” Nature Medicine 27 (2): 225–228. https://doi.org/10.1038/s41591-020-1124-9.

39.

Lokshin

Rodriguez-Ferrari

Torre

. 2024. “Electoral Cycles and Public Spending During the Pandemic.” Review of Development Economics 28 (3): 1077–1107. https://doi.org/10.1111/rode.13094.

40.

Marks

J. H.

2020. “Lessons from Corporate Influence in the Opioid Epidemic: Towards a Norm of Separation.” Bioethical Inquiry 17: 173–189. https://doi.org/10.1007/s11673-020-09982-x

41.

McCannon

B. C.

2021. “Do Governors Lead or Follow? Timing of Stay-at-Home Orders.” Eastern Economic Journal 47 (4): 506–518. https://doi.org/10.1057/s41302-021-00204-9.

42.

McCannon

B. C.

Hall

J. C.

. 2021. “Stay-at-Home Orders Were Issued Earlier in Economically Unfree States.” Southern Economic Journal 87 (4): 1138–1151. https://doi.org/10.1002/soej.12495.

43.

McGuire

M. C.

Olson

. 1996. “The Economics of Autocracy and Majority Rule: The Invisible Hand and the Use of Force.” Journal of Economic Literature 34 (1): 72–96. https://www.jstor.org/stable/2729410.

44.

McLaughlin

D. M.

Mewhirter

Sanders

. 2021. “The Belief That Politics Drive Scientific Research & Its Impact on COVID-19 Risk Assessment.” PLoS One 16 (4): e0249937. https://doi.org/10.1371/journal.pone.0249937.

45.

Miozzi

Powell

. 2023. “The Pre-Pandemic Political Economy Determinants of Lockdown Severity.” Public Choice 197 (1-2): 167–183. https://doi.org/10.2139/ssrn.4444772.

46.

Morris

Reuben

. 2020. “Coronavirus: Why Are International Comparisons Difficult?” BBC News, 2020.

47.

Nelson

2021. “The Timing and Aggressiveness of Early Government Response to COVID-19: Political Systems, Social Culture and More.” World Development 146: 105550. https://doi.org/10.1016/j.worlddev.2021.105550.

48.

Nordhaus

W. D.

1975. “The Political Business Cycle.” The Review of Economic Studies 42 (2): 169–190. https://doi.org/10.2307/2296528.

49.

Noy

Nualsri

. 2011. “Fiscal Storms: Public Spending and Revenues in the Aftermath of Natural Disasters.” Environment and Development Economics 16 (1): 113–128. https://doi.org/10.1017/S1355770X1000046X.

50.

Olson

1965. The Logic of Collective Action. Cambridge, MA, USA: Harvard University Press. http://commres.net/wiki/_media/olson.pdf.

51.

Olson

1993. “Dictatorship, Democracy, and Development.” American Political Science Review 87 (3): 567–576. https://doi.org/10.2307/2938736.

52.

Poterba

J. M

. 1996. “Budget Institutions and Fiscal Policy in the U.S. States.” NBER Working Paper 5449.

53.

Pulejo

Querubin

. 2021. “Electoral Concerns Reduce Restrictive Measure During the COVID-19 Pandemic.” Journal of Public Economics 198: 104387. https://doi.org/10.1016/j.jpubeco.2021.104387.

54.

Ramey

V. A.

2019. “Ten Years After the Financial Crisis: What Have We Learned from the Renaissance in Fiscal Research?” Journal of Economic Perspectives 33 (2): 89–114. https://doi.org/10.1257/jep.33.2.89.

55.

Rogoff

Sibert

. 1988. “Elections and Macroeconomic Policy Cycles.” The Review of Economic Studies 55 (1): 1–16. https://doi.org/10.2307/2297526.

56.

Romer

C. D.

2021. “The Fiscal Policy Response to the Pandemic.” Brookings Papers on Economic Activity 2021 (1): 89–110. https://www.jstor.org/stable/27093821 Contribution from The Johns Hopkins University Press .

57.

Romer

C. D.

Romer

D. H.

. 2019. “Fiscal Space and the Aftermath of Financial Crisis: How It Matters and Why.” NBER Working Paper Series 25768. http://www.nber.org/papers/w25768.

58.

Rosi

van Vugt

F. T.

Lecce

Ceccato

Vallarino

Rapisarda

. 2021. “Risk Perception in a Real-World Situation (COVID-19): How It Changes from 18 to 87 Years Old.” Frontiers in Psychology 12: 528. https://doi.org/10.3389/fpsyg.2021.646558.

59.

Roubini

Sachs

J. D.

. 1989. “Political and Economic Determinants of Budget Deficits in the Industrial Democracies.” European Economic Review 33 (5): 903–933. https://doi.org/10.1016/0014-2921(89)90002-0.

60.

Seiferling

2020. “Fiscal Deficits and Executive Planning Horizons.” Political Science Research and Methods 8 (2): 329–343. https://doi.org/10.1017/psrm.2018.69.

61.

UK National Audit Office. 2023. Tackling Fraud and Corruption against Government. National Audit Office, March 2023.

62.

Van Bavel

J. J.

Baicker

Boggio

P. S.

Capraro

Cichocka

Cikara

Crockett

M. J.

, et al. 2020. “Using Social and Behavioural Science to Support COVID-19 Pandemic Response.” Nature Human Behaviour 4 (5): 460–471. https://doi.org/10.1038/s41562-020-0884-z.

63.

von Hagen

Harden

L. J.

. 1995. “National Budget Process and Fiscal Performance.” European Economic Review 1995 (3): 771–779. https://doi.org/10.1016/0014-2921(94)00084-D.

64.

Wälti

2012. “Trust No More? The Impact of the Crisis on Citizens’ Trust in Central Banks.” Journal of International Money and Finance 31 (3): 593–605. https://doi.org/10.1016/j.jimonfin.2011.11.012.

65.

Wehner

2010a. “Cabinet Structure and Fiscal Policy Outcomes.” European Journal of Political Research 49 (5): 631–653. https://doi.org/10.1111/j.1475-6765.2009.01914.x.

66.

Wehner

2010b. “Institutional Constraints on Profligate Politicians: The Conditional Effect of Partisan Fragmentation on Budget Deficits.” Comparative Political Studies 43 (3): 208–229. https://doi.org/10.1177/0010414009347828.

67.

Wehner

2013. “Electoral Budget Cycles in Legislatures.” Legislative Studies Quarterly 38 (4): 545–570. https://doi.org/10.1111/lsq.12027.

68.

Wehner

Mills

. 2021. “Cabinet size and governance in Sub-Saharan Africa.” Governance. ISSN 0952-1895.

69.

Wintrobe

1990. “The Tinpot and the Totalitarian: An Economic Theory of Dictatorship.” American Political Science Review 84 (3): 849–872. https://doi.org/10.2307/1962769.

70.

Zupan

M. A.

1990. “The Last Period Problem in Politics: Do Congressional Representatives Not Subject to a Re-Election Constraint Alter Their Voting Behaviour?” Public Choice 65 (2): 167–180. https://doi.org/10.1007/BF00123797.

Distrust Deficits,Government Effectiveness and Coronavirus Financing

Abstract

Keywords

Introduction

Background: Political Baggage and the Great Financial Crisis

Theory

Regime Type

Electoral Competition and the Political Business Cycle

Executive Fragmentation and Polarization

Trust and Effectiveness in Public Institutions

The Data

Specification and Results

Robustness Check: Automatic Stabilizers

Limitations

Conclusion

Footnotes

Acknowledgements

Data Availability Statement

Declaration of Conflicting Interests

Funding

ORCID iD

Use of AI

Notes

Author Biography

References