Honesty by Convenience: Corruption Tolerance in Ecuador

Daniel Sánchez Pazmiño

Attitudes towards corruption may be a strong determinant of its incidence. Using survey data from the AmericasBarometer, binary-outcome empirical models are estimated to discover the key determinants of increased corruption tolerance in Ecuador between 2014 and 2016. It is found that two key variables may have influenced this increase. First, people who approved of the President’s job performance were initially less likely to justify bribes, yet by 2016 they started justifying corruption more. Second, people who identified closer to the political right justified corruption more in 2016 as well. These variables explain the increase as the percentage of people who approved the President’s performance decreased and the percentage of people identified with the political right increased. It is also found that people who were either employed or outside the labor force justified corruption more in 2016 than those who were unemployed.


1. Introduction

“Even if you are from [my political party], I will fulfill my duties. If you steal, steal well! Justify well! But do not let your affairs be seen, comrades.”[1] Uttered publicly by Rosa Cerda, Ecuadorian congresswoman for the Napo province (Castro 2021), these comments met widespread criticism around the country, although the remarks were initially met by cheers from the audience she addressed. However, Cerda’s declarations did not transcend an eight-day suspension (Ordóñez 2021), and the whole event was soon forgotten by most citizens.

This episode is only one of many corruption-related scandals that occurred in Ecuador, a middle-income country in South America. The country has seen increased COVID-19 vaccine inequality (Taj, Mitra, and Politi 2021), weakened public health services (Celi 2020), policymakers charging fees for political positions (Espinosa 2021), lost Social Security funds (Pesantes 2020), a former president convicted (Valencia 2020) as well as two vice presidents impeached and removed on charges of corruption (León 2020), among others. However, it is almost as if these scandals no longer cause outrage. At most, they cause a sigh of disappointment or social media outrage which dwindles shortly after.

Figure 1: Corruption Tolerance (%) Choropleth Map in 2019. The map shows corruption tolerance percentages across Latin America in 2019, where Ecuador places third among the most corruption-tolerant countries. Data from the ®AmericasBarometer 2018/19.

This apparent ambivalence has seen Ecuador place well above the corruption median in the world according to both Transparency International’s and World Bank’s corruption indexes. About 90% of voting-age Ecuadorians believe that at least half of politicians are corrupt, and more than a quarter of them admit having been involved with bribes in 2019, according to the AmericasBarometer (AB) survey by the Latin American Public Opinion Project (LAPOP). However, a mere 8.08% consider corruption to be the most serious problem faced by the country. In fact, 25.38% of Ecuadorians believe that paying a bribe is justified. Corruption tolerance has risen 11.79 percentage points from 2014 to 2019. Furthermore, Figure 1 shows that Ecuador is also one of the countries with the highest corruption tolerance in the region.

This article aims to investigate the determinants of the largest corruption tolerance increase in Ecuador—from 2014 to 2016 (as shown in Figure 2). This period coincided with two key events. First, the popularity of the governing regime sharply dropped for the first time in a decade (Quillupangui 2016). Second, the country faced an economic recession (Weisbrot et al. 2017). The article will seek to investigate the increase’s determinants by estimating binary-outcome models through logistic regression, which relates the probability of tolerating corruption to several individual-level public opinion and economic indicators. Survey data from the AmericasBarometer is used for the empirical modeling. It is determined that changes in presidential job approval as well as in political wing preferences during 2014 and 2016 could have influenced the increase. It is also found that those not unemployed (employed or outside the labor force) justified corruption more in 2016 relative to those who were unemployed.

Figure 2: Percent of Ecuadorians who justify corruption, by year. The evolution of corruption tolerance in Ecuador. Error bars show design-adjusted 95% confidence intervals.

Changes in attitudes toward corruption are important for studying corruption incidence, as a higher degree of corruption tolerance will eventually lead to more corruption (Ariely and Garcia-Rada 2019; Campbell and Göritz 2014). Learning what drives corruption tolerance can then foster better policymaking and citizen attitudes, which steer individuals away from dishonest acts. The argument in this article proceeds as follows. Section 2 provides the economic and political background for the argument. Section 3 reviews the relevant literature. Section 4 explains the empirical methodology. Section 5 presents and discusses the results from the estimation of empirical models. Section 6 concludes.


2. Economic and Political Background

Ecuador is a middle-income country located in upper South America next to Colombia and Peru. Its GDP for 2022 is projected to be $115.47 billion US dollars, with an expected growth rate of 2.68%.[2] Its population size notwithstanding, Ecuador is a naturally and ethnically diverse country, yet seems anchored to issues that have tormented it since its beginning as a nation. Hanratty (1991, xxi) identified four key issues that have determined the social and economic trajectories of the country: (i) a skewed social structure, (ii) persistent regional rivalries, (iii) a considerable dependence on oil, and (iv) a lack of strong political institutions. As of late 2022, these issues still dominate Ecuador’s political and social environment. Since the increase in corruption tolerance occurred during a key period in which two of these issues were most apparent, it is important to briefly review these mechanisms.

Ecuador’s modern economic history originated in the late 1960s with the discovery of oil fields in the Ecuadorian Amazon in 1967, along with its subsequent nationalization in the following years (Empresa Pública PetroEcuador 2013). The economy grew at rates never seen before, which led to important social and economic transformations in Ecuadorian society (Hanratty 1991; Hurtado 2007). The nationalization of the oil industry greatly increased public revenue, allowing for expansionary fiscal policy and overall growth. Unfortunately, this policy became a double-edged sword as it also increased Ecuador’s dependency on global market price fluctuations since Ecuador’s fiscal policy became tied to the ability to sell oil at a high price. Regardless of public investment, the country has been an underdog in economic terms as its GDP per capita has stagnated, while comparable South American nations have seen considerable growth, especially after 1990. An economic crisis in the late 90s led the country to its official dollarization, which further reduced the government’s role in managing the economy since monetary policy was no longer a possibility.

Political instability interacted with the dependence on commodity prices to hinder growth. Ecuador’s modern political history started in 1979 when the population was able to break a decade-long series of dictatorial regimes by electing a new president and a new constitution. However, the return to democracy did not mean stability: between 1979 and 2006, the country had 12 presidents, and, on average, Ecuador sees major protests against the government every six years (Loaiza 2022). This constant political instability disallows the establishment of long-term economic policies that can address the unhealthy dependence on commodities.

In 2006, a left-leaning government was elected, which concentrated power in the executive branch and engaged in significant reform through public spending. This government enjoyed high approval ratings for most of its tenure until 2016 (as seen in Figure 3). The key was the leader, rather than the party or its ideals. Branding himself as “the biblical underdog” (Hedgecoe 2009, paragraph 4), charismatic academic Rafael Correa distanced himself from the country’s political elite and constantly denounced corruption and injustice in the system. The new government promised a radical change in 2007 and did deliver, in a way, as it gave Ecuador a politically stable, though totalitarian, environment, as well as other changes in political and economic mechanisms (Weisbrot et al. 2017).

Figure 3: Ecuadorian public opinion indicators, 2004-2019. Time series for political public opinion questions asked in the AmericasBarometer. Error bars show 95% design-adjusted confidence intervals.

Figure 3 shows that the President reached all-time high popularity in 2014 and then a severe drop in 2016. This trend is seen through the percentage of people who approve of the President’s job performance and the percentage who report confidence in him. Another notable change in the political landscape of this period was the way that the voting-age population identified politically. There was a notable increase in the number of people who identified as the ‘right’ of the political spectrum, while those who identified with the ‘left’ did not see significant changes.

The administration’s high popularity allowed Correa to vanquish every political opponent. During most of his tenure, there was no need for any legislative pact to pass policy, as he enjoyed well over a two-thirds majority in all political apparatuses. This government was often criticized due to its totalitarian practices, yet the average Ecuadorian voter appeared deaf to this. Institutionality, democratic values, the separation of powers, etc., seemed abstract and far from appealing to a traumatized nation. The new government convinced the people that it had been the political right that had destroyed the country, an idea that has haunted the current presidency of conservative businessman Guillermo Lasso, elected in April 2021.

As Ecuador entered a severe recession due to plunges in commodity prices and a fatal earthquake in 2016, the Correa administration was forced to take widely unpopular austerity measures. Furthermore, a significant amount of corruption accusations appeared against top government officials, which planted the seed of a deep investigation into a complex corruption scheme involving top government officials and large corporations (Villavicencio et al. 2019), which ended in a capture order for Correa in 2020. Several narratives started to be constructed by government officials to explain the flaws and accusations denounced at that point. These included reducing corruption accusations to “political persecution” or unfounded claims spread before elections (Meléndez and Moncagatta 2017).

Regarding the economic recession, Orozco (2015) holds that although the commodity price collapse in 2008 was greater, there was little reduction in economic activity in 2008 because international financing and savings left over from oil funds were used to keep government expenditure high. In 2016, savings eroded, and government debt grew bigger, stagnating the economy. A politically weakened Correa left Ecuador for Belgium in 2017 after giving up power to his political successor, Lenín Moreno, who later turned his back on Correa.

Understanding these mechanisms, it is clear that Ecuador proves to be an excellent setting for studying the determinants of corruption tolerance. The economic and historical background has created a laboratory to study how people react to highly corrupted environments in the presence of exogenous shocks. Any analysis of the corruption tolerance increase during the 2014-2016 period must necessarily account for the events that happened at the time. In the next section, I analyze the related literature to construct a framework that will allow me to use these events as potential determinants for corruption tolerance at the individual level.


3. Corruption: A Basic Framework

The literature on corruption mostly focuses on corruption incidence and how it may determine other economic and social outcomes. While less attention has been given to the corruption tolerance phenomenon, a key finding of this subset of the literature is that the more tolerance and exposure to corrupt acts, the more likely it is that these will spread across individuals. Ariely and Garcia-Rada (2019) discuss experimental findings showing that individuals who pay a bribe or are requested to pay one are more likely to behave dishonestly in subsequent ethical dilemmas. Gino et al. (2009) show that subjects exposed more to dishonest behaviors are more likely to engage in them.

An empirical study of corrupt organizations by Campbell and Göritz (2014) shows that initial exposure to dishonest acts can create an organizational culture fostering corruption among its members. The corrupt culture may change the behavior of otherwise honest individuals through social pressure, notably when principles such as ‘the ends justify the means’ are perceived as the organization’s core values. Specifically for the variable at hand, Carlin (2013, 6) proposes that:

[B]ribery has a self-perpetuating mechanism: if the rule of law is so weak that state actors are brazen enough to solicit bribes and self-interested citizens feel justified in paying them, the supply and demand of bribery will converge to form strong social behavioral norms.

It is adequate to place corruption in a basic framework that will inform the empirical modeling, keeping two key channels in mind: the social and economic payoffs that corrupt acts imply. I build this framework based on the implications of a microeconomic model of corruption (Shleifer and Vishny 1993), the effect of social payoffs on economic outcomes (Akerlof 1980), and a behavioral theory of corruption normalization (Ashforth and Anand 2003).

Shleifer and Vishny (1993) model bribes in a way where a public official trades public goods in exchange for bribes. Private agents then pay them to receive the goods and the consumer surplus that any transaction brings. This payoff might be understood as an individual economic incentive to engage in corrupt acts: paying the bribe allows the use of a desirable public good or allows for quicker access to it. Thus, economic convenience could be an important determinant of how people behave around corruption: people may tolerate dishonesty if it means a positive economic payoff.

On the other hand, there might also be moral considerations in deciding to tolerate or engage in corruption. While the economic payoff of paying or receiving a bribe may be positive, the moral connotation of the act may bring shame or rejection from society. Avoiding a bad image can very well become an important determinant of the decision to engage in corruption. Nevertheless, in environments where this is tolerated the negative social payoff of bribing might be smaller, which decreases the social payoff of being honest. Akerlof (1980) holds that social payoffs might change economic outcomes in a significant way, deviating from the neoclassical equilibria. How the social payoffs of corrupt acts are determined is key, as it could be assumed that most of the time the economic payoff of bribes is positive for the corrupt individual.

Ashforth and Anand (2003) develop a model to explain how corruption is normalized or tolerated in an organization, which helps to understand how these social payoffs are determined. The implications of this model imply that social payoffs of being corrupt should be decomposed into effects related to the institutionalization, rationalization, and socialization of corruption.

Leadership in the organization is very relevant for institutionalization behaviors in Ecuador, considering its recent historical background. Ashforth and Anand (2003) propose that leaders need not engage in corrupt acts to foster their normalization, as they can simply facilitate or ignore the initial corrupt acts to have subordinates start normalizing corruption. Subordinates do not second guess their superiors’ decisions due to the habit of obedience, which is more prevalent in highly hierarchical organizations.

Two other mechanisms are involved in the normalization of corruption. The rationalization mechanism of corruption is especially important, as it can be modeled at the individual level. This mechanism involves corrupt individuals rationalizing corruption to “avoid the adverse effects of an undesirable social identity” (Ashforth and Anand 2003, 13). Relevant to the present context is the denial of responsibility rationalization, in which corrupt individuals become convinced that they have no choice but to engage in corrupt acts due to external circumstances.

Denial of responsibility also involves individuals seeing their own corruption as a form of retribution against unfair actions previously exerted on them. Another example of denial of responsibility is when corrupt acts are justified because actors perceive those who denounce corruption as illegitimate authorities with motives other than the organization’s well-being.

The socialization mechanism considers the peer effects of corruption, wherein dishonest practices are ‘taught’ to organization newcomers. Newcomers will initially be induced to change their attitudes toward corrupt beliefs, then peer pressured to escalate these practices. Since newcomers strive to be accepted, they adopt these dishonest behaviors as their own, while they also rationalize them to avoid the social costs of dishonesty. Later, the newcomers become the ones that exert peer pressure on future members.

Having established a framework that will allow for better modeling of corruption tolerance, it is useful to look at what the literature has found with the variable at hand. Singer et al. (2016) found that for every Latin American country in 2014, at least 60% of the respondents perceived their governments to be corrupt, but a much smaller proportion considered corruption the most important problem in their countries. They also found that those who justify corruption have been exposed to a bribe in the past.[3] Other significant determinants of corruption tolerance in 2014 were age and the urban-rural dichotomy. Younger participants tend to justify corruption to a higher degree, a robust finding through time and across countries of the region. Those living in rural settings also tend to justify corruption more.

Lupu (2017) shows that corruption tolerance has been growing consistently in the region and that the average Latin American country has about a fifth of its population believing that corruption is justified. Between 2014 and 2016, corruption tolerance grew from 17.4% to 20.5% throughout the region. Older citizens, as well as those exposed to corruption previously, are more prone to justifying it. The level of perceived corruption also appears to be a significant determinant. Lupu (2017, 67) concludes, therefore, that corruption may have become “a self-fulfilling prophecy: as more and more citizens perceive that corruption is more widespread, they also become more likely to condone it.”

Regarding Ecuadorian corruption tolerance, Moscoso (2018) demonstrates that corruption is also perceived to be widespread despite not being regarded as an important problem. Montalvo (2019) finds that the general trend in which younger people justify corruption more also applies to Ecuador. For the same round, Moscoso and Moncagatta (2020) find that age and interest in politics are significant predictors of corruption tolerance, as well as exposure to corruption, as was found by Lupu (2017) for the whole region. The empirical evidence can support corruption becoming a known inconvenience for daily life in the country rather than an unacceptable threat to the system, perceiving it as endemic to the political and social environments.


4. Methodology

The AmericasBarometer (AB) survey from the Latin American Public Opinion project is used in this paper to investigate the corruption tolerance increase in Ecuador. This survey was administered in Ecuador and other Latin American countries from 2004 to 2019, at about two-year intervals. It asks about public opinion matters, including democracy and corruption, among others. The open-access AB databases available on the LAPOP website are used for the empirical models. Table 1 presents descriptive statistics for all variables used.

Table 1: Descriptive statistics for all variables. Descriptive statistics table with estimates (Est.) and robust standard errors (SE), where age, years of education, and the external and internal political efficacies are arithmetic means. All other variables are percentages. Standard errors are adjusted for survey-design effects.

The empirical models estimated in this study will use the 2014 and 2016 rounds of the AB in Ecuador, with n2014 = 1489 and n2016 = 1545. The survey is based on a multi-stage national probability design, with design-adjusted errors of ±2.5% and ±1.9%, respectively, each year (LAPOP 2014; LAPOP 2017). Both surveys are self-weighted. However, 95% confidence intervals for the descriptive statistics, adjusted for survey-design effects, are presented when relevant.

The empirical analysis is concerned with the EXC18 question: “Do you think, given the way things are, sometimes paying a bribe is justified?” (Moscoso 2018, 96, [originally asked in Spanish]). The question has been asked in all survey rounds in Ecuador and is the last one after a set of questions regarding corruption exposure and perception. This variable (ctol) is equal to 1 when the respondent answers ‘Yes,’ 0 when the answer is ‘No,’ and dropped from the model otherwise. All models have ctol as the explained variable, and responses to other questions are used as regressors.

In order to identify the changes in behavior that led to the increase, the survey rounds are pooled, and the following general model is estimated:

Where R is a vector of controls and x* is a key regressor whose change across time may have significantly influenced the rise of corruption tolerance between 2014 and 2016. This key regressor interacts with a year dummy, y16, which equals unity for 2016 observations. The complete regressors’ vector includes all variables in R, the key regressor x*, and the interaction term. The parameters vector θ includes vector β as well as β0δ0, and δ1G is the link function. In this article, I follow the literature and use a logistic function as G.

Consider the partial effect of the key regressor x* on P(ctol = 1 | X):

The parameter δ1 would then measure the ceteris paribus effect of a change in the key regressor x* from 2014 to 2016 in ctol. Therefore, the coefficient of interest in this study is δ1. If there has been a change in 2016 in x* which significantly influences corruption tolerance, δ1 should be statistically significant. Further, a δ1 coefficient not statistically different from zero would mean that individuals with and without this key characteristic are equally likely to justify corruption across time. Average partial effects are shown for all models. I use survey-weighting to adjust for complex survey design effects, as suggested by Castorena (2021). Since the sample is self-weighted, survey weighting does not affect magnitudes, only standard errors.


5. Results

As seen in Section 2, two economic variables significantly changed during the corruption tolerance increase period: the percentage of people who reported a worse economic situation as well as unemployment. Variables that proxy attitudes in the political landscape also significantly changed: the percentage of people who confide in the President, the percentage who approve the President’s performance, and the percentage of people who identified with the political right. These variables were used for simple empirical models, which follow the equation below.

Where the key regressor x* can be: a dummy variable set to unity for respondents who answered that their economic situation is worse (Model 1), a dummy variable set to unity for those who reported being unemployed (Model 2), a discrete variable with numbers 1-7, where higher values imply a higher degree of confidence in the President (Model 3), a discrete variable with numbers 1-5, with higher numbers indicating a higher rating of the President’s job performance (Model 4), or a discrete variable with numbers from 1-10 where 1 is the extreme left and 10 is the extreme right (Model 5). Table 2 presents the coefficients of the logistic model, and Table 3 presents their associated average partial effects. It shows that an unemployed person is 5.9% more likely to justify corruption. Additionally, a respondent who answered one degree higher in their confidence in the President was 2.4% less likely to justify it. Finally, a person who rated the President’s job performance one unit higher was 4.4% less likely to justify corruption. All other partial effects are not significant.

Table 2: Logit coefficients for baseline models. Logit coefficients of baseline models with design-adjusted std. errors. *p < 0.1, **p< 0.05, ***p < 0.01.

Consider the logit coefficients in Table 2. The coefficient for the year dummy confirms the significance of the corruption tolerance increase in 2016, which is lost when considering interaction terms with confidence in the President and has a negative sign with the other political variables. The inclusion of unemployment and economic situation does not eliminate the significance of the year dummy. Model 1 suggests that a person who reports having a worse economic situation does not tolerate corruption differently than those who report the same or equal economic situation. According to Model 2, respondents who were unemployed were more likely to justify corruption than those who were not.[4] The interaction term in this model has a negative sign, which shows that the effect of unemployment in 2016 was lower than in 2014, meaning unemployed people justified corruption less after political instability set in.

Table 3: Average partial effects for logit models in Table 2. Average partial effects for models in Table 2, with design-adjusted std. errors. *p < 0.1, **p< 0.05, ***p < 0.01.

Models 3 and 4 display the same relationship: people who either trust or approve of the President to a higher degree also tolerate corruption less. A more zealous supporter of the regime believed bribes were not justified. However, this appears to change in 2016. The interaction terms for both variables are significant and positive: in 2016, supporters started to justify corruption more. This relationship could explain the jump in corruption tolerance as regime support eroded in 2016, which meant that the number of non-supporters was higher, and these respondents justified corruption more than supporters. Also, the remaining supporters started to justify bribes to a higher degree. In Model 3, the significance of the year dummy is lost, while in Model 4 the sign is reversed.

The coefficients in Model 5 show that a person who identifies closer to the political right does not justify corruption more or less relative to those identifying closer to the political left. However, the interaction term shows that people answering higher values of this variable justified corruption more in 2016. Once again, the significance of the year dummy is lost when considering this variable. With a higher number of respondents identifying with the political right wing, and who appear to justify corruption more, it is understandable how overall corruption tolerance increased.

Figure 4: Graphical representations of corruption tolerance across key explanatory variables. The panels show the percent that justifies corruption across the groups used as explanatory models in Table 2. Error bars represent the 95% confidence intervals adjusted for design effects.

These findings are supported by Figure 4. According to panel (a), in 2014, only 12.03% of those not unemployed justified corruption, while in 2016, the figure increased to 27.03%, a very close percentage to unemployed people who justified it in 2016. The time difference between these point estimates is not statistically significant, which means that in 2016 the effect of unemployment on corruption tolerance approached zero. Thus, Figure 4, along with Model 2 of Table 2, shows that it was not the unemployed who started to justify corruption less, but the people who were not unemployed that started to justify it more.

Panels (b) and (c) of Figure 4 show that the percentage of people who either confided in or approved the President and justified corruption increased significantly between 2014 and 2016. This means that the negative effect of supporting the executive in 2016 was smaller than in 2014, as confirmed by the interaction term in Models 3 and 4 of Table 2.

In panel (d) of Figure 4, four different political groups are considered: the left, right, center, and those who did not answer the question. All four groups saw increases in the percentage of group members who justify corruption. All increases in corruption tolerance are significant, except for those who identify with the left wing.

Table 4: Logit coefficients for modified models. Logit coefficients of the modified models with design-adjusted std. errors. *p < 0.1, **p< 0.05, ***p < 0.01.
Table 5: Average partial effects for models in Table 4. Average partial effects for models in Table 4, with design-adjusted std. errors. *p < 0.1, **p< 0.05, ***p < 0.01.

Now the general model, as described by the equation in Section 4, is estimated with the key regressors and a set of controls at the individual level. I keep the variables that yielded statistically significant interaction terms with the year dummy in Table 2, except for confidence in the President, as job approval ratings contemplate the same effects. Coefficients are displayed in Table 4, and average partial effects are depicted in Table 5.

These models include multiple control variables suggested by Moscoso and Moncagatta (2020) and Lupu (2017). Of these, only age is significant and has a negative effect on corruption tolerance. A person older by one year is four percentage points less likely to justify corruption. Political efficacy indicators are included too. The external political efficacy question, which asks if respondents believe that politicians serve the interests of the people, has no statistical significance. Internal political efficacy asks about how well the respondent understands politics, and this control is significant. A person who understands more about the country’s politics is more likely to justify corruption, and the estimated increase in corruption tolerance probability is about 1.5 percentage points.

While Moscoso and Moncagatta (2020) find that none of the political efficacy variables are significant for corruption tolerance in 2019, they find that interest in politics is significant and has a positive effect. That finding is reversed here; interest in politics is significant yet portrays a negative relationship between the two—more interest in the country’s politics is negatively related to corruption tolerance. A person who reports being interested in politics is about 3.5 percentage points less likely to justify corruption. While the two questions may appear to be similar, they imply different attitudes to politics.

The political efficacy question asks if citizens are politically aware, and the second one asks if they are interested in entering politics. Separating these two questions may imply that attitudes of apathy or pragmatism to the political society are separated from an ‘idealist’ attitude towards it by those who would like to enter politics. A control for years of education is also added and is significant, communicating that more educated respondents are less likely to justify corruption. Other things equal, an additional year of education is related to a six percentage points reduction in corruption tolerance. This finding is intuitive considering that more education may mean more knowledge about the costs of corruption. The social payoffs for being honest may also be higher as higher education may entail a better economic position, which makes engaging in corrupt acts less economically attractive.

Exposure to corrupt acts (paying or being asked to pay a bribe) is also strongly correlated with tolerance. A person who has been exposed to some form of bribing is about 15% more likely to justify corruption. The causality direction is unclear as it might be possible that a predisposed tolerance to corruption due to external factors makes citizens more likely to be in environments where corruption flourishes. Corruption perceptions, on the other hand, play no role in determining corruption tolerance for this period.

A dummy variable equal to unity for respondents who have recently attended a protest is added and is very significant. Other things equal, a person who has attended a protest is about 7% more likely to justify corruption. This relationship might be related to the denial of the victim corruption tolerance explanation proposed by Ashforth and Anand (2003). People who attend protests are likely to reject the current state of things, which may induce a feeling of contempt against society. They may believe dishonest acts are justifiable in these circumstances because they feel corrupt acts can be ‘retribution’ by alleging that small corruption acts are nothing compared to grand corruption scandals. Since they have ‘declared’ their rejection of the system in general, they have surrendered to its flaws and have no social incentives to remain honest.

Most importantly, Table 4 shows that results in Table 2 are robust to several controls suggested by the literature. It is still true that unemployed respondents justified corruption more in 2014 and less in 2016. People who approved the President’s job were less likely to justify corruption in both years, but their rejection was weaker in 2016. Finally, while political identification was not significant in 2014, it was in 2016, and people who identified as closer to the political right were more likely to justify corruption.

It is possible that those initially unemployed justified corruption more because of their ‘steady state’ of corruption tolerance; unemployed people are economically disadvantaged, which gives them incentives to engage in corrupt actions that can yield positive economic payoffs. Additionally, as they cannot enter the job market, they might feel alienated from society, which might decrease social or moral incentives to remain honest. The change in corruption tolerance in 2016 is more difficult to understand. It is possible that, since the recession, many have lost jobs and have had relatively short unemployment spells. The recently unemployed may not feel too alienated from society and thus have not adopted an attitude of pragmatism toward the current circumstances. Savings or family income may support the recently unemployed, which makes them less desperate and more prone to take the ‘moral high ground.’ These factors all contribute to them still feeling like a part of society, which reduces their rationalization of corruption. However, over longer unemployment spells, desperation may trigger more pragmatic points of view, which can lead to higher corruption tolerance in the future.

To better understand the implications of the political variables’ coefficients and their change over time, consider a key effect on corruption normalization: leadership. Therefore, supporters of the regime faced higher social sanctions when justifying corrupt behavior, as this may have implied that the economic and political model they supported was flawed. However, by 2016, the popularity of the government saw a sharp decrease and rationalization narratives appeared. A statement by the President represents a particularly relevant example: a regime-affiliated newspaper portrayed how Correa qualified the Panama Papers as a selective fight against corruption, which is nothing but another kind of corruption, as well as a “strategy by power groups to destabilize democratically elect governments” (Telégrafo 2016, paragraphs 5-7). If the legitimacy of those who denounce and control corruption is questioned by an important authority of the organization, corrupt acts can be more easily normalized (Ashforth and Anand 2003). Thus, if there was a greater incidence of corrupt acts and numerous attempts by the authorities to justify them, it is understandable how supporters of the regime started to justify corruption more.

Results also show how people who identified with the political right became more corruption tolerant in 2016. It is unclear if there is a causal relationship between the political right and corruption tolerance. This is because it has been determined that in Ecuador, the answer to the political identification question has little to do with the traditional principles of the political wings. Rather, it is possible that the political self-identification of Ecuadorians follows a multidimensional perspective (Moncagatta and Poveda 2020), not accurately measured with an indicator like the one used here.

A potential explanation for the direction of this effect is that those who identified with the right do so partially because they consider themselves against the ruling government. This theory is reasonable considering the increase in the percentage of ‘rightists’ from 2014 to 2016, which aligned with the regime’s downfall. Additionally, it is possible that anti-regime attitudes formed under a common set of ideas rather than under a political party or figure since, during President Correa’s tenure, the opposition forces did not materialize strongly behind a party or leader (Meléndez and Moncagatta 2017). It is sensible to believe that no political wing has any particular preference for justifying or rejecting corruption, as notable academics (Holcombe and Boudreaux 2015) and politicians (Morris 2021) associated with both wings have denounced corruption. Anti-regime respondents, rather than those who actually identified with the political right, might rationalize corruption as a form of retribution, as proposed by Ashforth and Anand (2003) and discussed by Adoum (2000) in the Ecuadorian case.

Some limitations are worth discussing. One of the most important issues is the possible differences across individuals in their understanding of ‘bribes.’ Even though the EXC18 question mentions paying a bribe, the idea that comes to mind for respondents may be outside the mentioned hypothetical situations. What respondents think about when reading “paying a bribe” could vary. This openness implies that observations are not homogeneous. Another issue is the social desirability bias; the corruption tolerance variable may be considerably mismeasured due to this phenomenon, and instances of social desirability bias may be heterogeneous across unobserved characteristics that are correlated to our key regressors.


6. Conclusions

The degree to which citizens of a country justify corruption is a topic worth careful study, given that the more corruption is normalized, the more likely it is that actors in that environment engage in it. This trend occurs because corruption necessarily implies both social and economic payoffs, so when the social payoff of being honest is eliminated through a justification of dishonest acts, the economic payoff now drives an individual’s decision to participate in this behavior.

In Ecuador, the data from the AmericasBarometer survey has shown that corruption tolerance has risen since 2014, with the most significant increase being between 2014 and 2016. Binary-outcome logit models were implemented to find the determinants of this increase. It was found that changes in presidential job approval and political wing preferences during the 2014 and 2016 period could have influenced the corruption tolerance increase. It was also found that those not unemployed justified corruption more in 2016 relative to those who were unemployed. While this trend does not explain the corruption tolerance increase, it is an interesting finding which may foreshadow a considerable increase in corruption tolerance after the COVID-19 pandemic if corruption tolerance is a lagged function of economic conditions.

Considering this empirical evidence, the jump in corruption tolerance between 2014 and 2016 is explainable. The economic recession brought about by the collapse of commodity prices, the dependence on government expenditure, and the earthquake of April 2016, combined with the numerous accusations of corruption against government officials, deteriorated regime support. The recession led to a decrease in the percentage of people who approved of the President and an increase in the percentage of people who identify with the political right. Also, a decrease in the number of people who did not justify corruption and an increase in the number of people who did was identified. All of this would account for the increase in corruption tolerance.

The most robust findings of the literature are confirmed. Exposure to corruption is a strong predictor of corruption tolerance, so people exposed to bribes are more likely to justify corruption. Also, age is a negative predictor of corruption tolerance, a troubling finding revealing a flawed education system and the lack of attention given to the political inclusion of younger citizens. Education is identified as a negative predictor, but only for 2016. This regressor may have a significant effect on how people behave toward dishonest behavior, as pointed out by Adoum (2000), who considers academic dishonesty as a precedent for political corruption.

These findings suggest obscure details about the way that Ecuadorians behave toward corruption. The considerable amount of consequences for acting corruptly in recent years has not made people tired of dishonesty. In fact, it seems that it has only made them more willing to engage in it. The opposition groups to President Correa’s regime, which often cite corruption scandals as arguments against left-leaning politicians, have seemingly become more open to the idea that corruption is inherent to politics and can be justified if it suits their needs. A similar criticism can be aimed at people who participate in protests and are later found to be other sources of corruption tolerance.

Nevertheless, this phenomenon is not isolated to opposition groups and is also found among regime supporters. When corruption became the norm among leaders, supporters became pragmatic regarding corrupt acts. Both of these lines of reasoning entail that corruption will keep happening regardless of who is in power, as both parties in politics have found a way to allow deceit to exist. Calls for honesty have been bent to such an extent that they have become devoid of true meaning, only used if such honesty works to the convenience of those speaking about it.

The costs of corrupt behavior are well-documented in the literature. They challenge the validity of democratic systems (Moscoso 2018), destroy wealth, distort markets, and hinder economic growth and income distribution (Shleifer and Vishny 1993; Singer et al. 2016). Corruption can even add to human misery by shortening life expectancy (Siverson and Johnson 2014), a result expected to appear soon in Ecuador, considering the extensive corruption incidence during the COVID-19 pandemic. The problem of corruption is a politically and emotionally charged discussion point. While policy-making and legal action might be ways to change attitudes toward corruption, it will be difficult to fully eliminate corruption this way. The principle of honesty by convenience must be vanquished through individual action and reflection so that dishonesty is reprehended enough to influence social incentives and provide an escape from the atrocious evils that corruption espouses.



I thank the Latin American Public Opinion Project (LAPOP) and its major supporters (the United States Agency for International Development, the Inter-American Development Bank, and Vanderbilt University) for making the data available. I also thank Universidad San Francisco de Quito for providing the funding necessary for the copyrighted AmericasBarometer data. I am eternally grateful to Professor Santiago José Gangotena, my supervisor, LAPOP scholars J. Daniel Montalvo and Paolo Moncagatta, and Professor Julio Acuña for their feedback on the initial versions of this article, as well as Alejandra Marchán, who helped me throughout the entire editing process. Dedicated to the loving memory of Jorge Pazmiño.



Presidente Correa pide “campaña mundial” para revelar “toda la información” de ‘Panama Papers’ (2016). El Telégrafo. https://www.eltelegrafo.com.ec/noticias/economia/8/presidente-correa-pide-campana-mundial-para-revelar-toda-la-informacion-de-panama-papers.

La asambleísta que dijo ‘si roben, roben bien’ se enreda más (2021). Plan V. https://www.planv.com.ec/confidenciales/confidencial-politica/la-asambleista-que-dijo-si-roben-roben-bien-se-enreda-mas.

Adoum, J. (2000). Ecuador: señas particulares (6th ed.). Eskeletra Editorial.

Akerlof, G. (1980). A Theory of Social Custom, of Which Unemployment May be One Consequence. The Quarterly Journal of Economics, 94(4), 749.

Ariely, D. and Garcia-Rada, X. (2019). Contagious Dishonesty: Dishonesty begets dishonesty, rapidly spreading unethical behavior through a society. Scientific American, 321(3), 63–66.

Ashforth, B., and Anand, V. (2003). The Normalization of Corruption in Organizations. Research in Organizational Behavior, 25, 1-52.

Campbell, J., and Göritz, A. (2014). Culture Corrupts! A Qualitative Study of Organizational Culture in Corrupt Organizations. Journal of Business Ethics, 120(3), 291-311.

Carlin, R. (2013). Violations of the Rule of Law Fuel Tolerance of Bribery in the Americas: Americas Barometer Insights No. 88. LAPOP. https://www.vanderbilt.edu/lapop/insights/IO888en.pdf

Castorena, O. (2021). Survey Weights in AmericasBarometer data. LAPOP. https://www.vanderbilt.edu/lapop/insights/IMN007en.pdf.

Castro, M. (2021). ¿Quién es Rosa Cerda, la asambleísta de Pachakutik? GK. https://gk.city/2021/07/19/rosa-cerda-asambleista-pachakutik-roben-bien/.

Celi, E. (2020). El reparto de hospitales: el nuevo escándalo de la actual Asamblea. Primicias. https://www.primicias.ec/noticias/politica/reparto-hospitales-escandalo-nuevo-actual-asamblea/.

Cohen, M., Lupu, N., and Zechmeister, E. (Eds.). (2017). The Political Culture of Democracy in the Americas, 2016/17: A comparative study of Democracy and Governance. https://www.vanderbilt.edu/lapop/ab2016/AB2016-17_Comparative_Report_English_V2_FINAL_090117_W.pdf.

(2018). Cultura política de la democracia en Ecuador y en las Américas, 2016/17: Un estudio comparado sobre democracia y gobernabilidad. Latin American Public Opinion Project. https://www.vanderbilt.edu/lapop/ecuador/AB2016-17_Ecuador_Country_Report_V3_12.20.18_W_01.15.19.pdf.

El Telégrafo (2016). Presidente Correa pide “campaña mundial” para revelar “toda la información” de “Panama Papers”. El Telégrafo. https://www.eltelegrafo.com.ec/noticias/economia/8/presidente-correa-pide-campana-mundial-para-revelar-toda-la-informacion-de-panama-papers.

Empresa Pública PetroEcuador (2013). El petróleo en el Ecuador: la nueva era petrolera. Quito. Empresa Pública PetroEcuador. https://www.eppetroecuador.ec/wp-content/uploads/downloads/2015/03/El-Petr%C3%B3leo-en-el-Ecuador-La-Nueva-Era.pdf.

Espinosa, C. (2021). Fiscalía formulará cargos contra la exvicepresidenta de la República María Alejandra Vicuña. El Comercio. https://www.elcomercio.com/actualidad/seguridad/fiscalia-cargos-exvicepresidenta-maria-alejandra-vicuna.html

Gino, F., Ayal, S., and Ariely, D. (2009). Contagion and Differentiation in Unethical Behavior: The Effect of One Bad Apple on the Barrel. Psychological Science, 20(3), 393-398.

Hanratty, D. (Ed.). (1991). Area Handbook Series. Ecuador: a country study (3rd ed.). https://www.loc.gov/item/91009494/.

Hanratty, D. (1991). Introduction. In D. M. Hanratty (Ed.), Area Handbook Series. Ecuador: a country study (3rd ed.), xxi-xxv.

Hedgecoe, G. (2009). Rafael Correa: an Ecuadorian journey. openDemocracy. https://www.opendemocracy.net/en/rafael-correa-an-ecuadorian-journey/.

Holcombe, R., and Boudreaux, C. (2015). Regulation and corruption. Public Choice, 164(1-2), 75-85.

Hurtado, O. (2007). Las costumbres de los ecuatorianos. Editorial Planeta del Ecuador, C.A.

LAPOP. (2014). AmericasBarometer 2014: Sample Design. LAPOP. http://www.vanderbilt.edu/lapop/ab2014/AB-2014-Tech-Info-112114-W.pdf.

LAPOP. (2017). AmericasBarometer, 2016/17: Technical Information. LAPOP. https://www.vanderbilt.edu/lapop/ab2016AmericasBarometer_2016-17_Sample_Design.pdf.

León, J. (2020). Ecuador’s Former President Convicted on Corruption Charges: The outcome of Rafael Correa’s closely watched trial is likely to affect Ecuador’s political landscape for years to come. The New York Times. https://www.nytimes.com/2020/04/07/world/americas/ecuador-correa-corruption-verdict.html.

Loaiza, Y. (2022). Un golpe cada 6 años: la trágica historia de derrocamientos y convulsión social del Ecuador. Infobae. https://www.infobae.com/america/historia-america/2022/06/25/un-golpe-cada-6-anos-la-tragica-historia-de-derrocamientos-y-convulsion-social-del-ecuador/.

Lupu, N. (2017). Corruption in the Americas. In M. Cohen, N. Lupu, and E. Zechmeister (Eds.), The Political Culture of Democracy in the Americas, 2016/17: A comparative study of Democracy and Governance, 49-67.

María, J. (2020). Ecuador’s Former President Convicted on Corruption Charges. The New York Times. https://www.nytimes.com/2020/04/07/world/americas/ecuador-correa-corruption-verdict.html.

Meléndez, C., and Moncagatta, P. (2017). Ecuador: Una década de correísmo. Revista De Ciencia Política (Santiago), 37(2), 413-448.

(2020). The Political Culture of Democracy in Ecuador and in the Americas, 2018/19: Taking the Pulse of Democracy. Latin American Public Opinion Project. https://www.vanderbilt.edu/lapop/ecuador/AB2018-19-Ecuador-Country-Report-Eng-V2-W-200903.pdf.

Moncagatta, P., and Poveda, A. (2020). The Politicization of Citizens and Ideological Polarization in Ecuador. In P. Moncagatta, A. Moscoso, S. Pachano, J. Montalvo, and E. Zechmeister (Eds.), The Political Culture of Democracy in Ecuador and in the Americas, 2018/19: Taking the Pulse of Democracy (pp. 73-88): Latin American Public Opinion Project.

Montalvo, J. (2019). Spotlight on Corruption Tolerance in Ecuador. In E. Zechmeister & N. Lupu (Eds.), Pulse of Democracy (p. 67). Nashville, TN: LAPOP.

Morris, S. (2021). Corruption and Anti-Corruption from the Left and Right in the Americas. Rice University. https://www.bakerinstitute.org/media/files/files/b298016d/bi-brief-072021-usmx-anticorruption.pdf https://doi.org/10.25613/61ZB-RW45.

Moscoso, A. (2018). Corrupción y democracia en el Ecuador: Efectos de la corrupción en la opinión pública sobre la democracia, el apoyo al sistema y la tolerancia política. In J. Donoso, P. Moncagatta, A. Moscoso, S. Pachano, J. Montalvo, and E. Zechmeister (Eds.), Cultura política de la democracia en Ecuador y en las Américas, 2016/17: Un estudio comparado sobre democracia y gobernabilidad. Latin American Public Opinion Project, 85-113.

Moscoso, A., and Moncagatta, P. (2020). Attitudes Towards Corruption in Ecuador. In P. Moncagatta, A. Moscoso, S. Pachano, J. Montalvo, and E. Zechmeister (Eds.), The Political Culture of Democracy in Ecuador and in the Americas, 2018/19: Taking the Pulse of Democracy. Latin American Public Opinion Project, 114-123.

Ordóñez, V. (2021). Rosa Cerda por su apología ‘roben bien’ es suspendida como asambleísta por ocho días y sin sueldo. El Universo. https://www.eluniverso.com/noticias/politica/rosa-cerda-por-su-apologia-roben-bien-es-suspendida-como-asambleista-por-ocho-dias-y-sin-sueldo-nota/.

Orozco, M. (2015). Las crisis del 2009 y 2015 son distintas. El Comercio. https://www.elcomercio.com/actualidad/negocios/crisis-2009-2015-son-distintas.html.

Ortiz, R. (Ed.). (2013). Rafael Correa Delgado: Biografías de Líderes Políticos. Barcelona Centre for International Affairs. https://www.cidob.org/biografias_lideres_politicos/america_del_sur/ecuador/rafael_correa_delgado.

Pesantes, K. (2020). Escándalo de corrupción en hospital del IESS suma siete nuevos acusados. Primicias. https://www.primicias.ec/noticias/lo-ultimo/corrupcion-hospital-iess-nuevos-acusados/.

Quillupangui, S. (2016). Un 58% de la población desaprueba la gestión del presidente Rafael Correa, según Cedatos. El Comercio. https://www.elcomercio.com/actualidad/politica/desaprueba-gestion-rafaelcorrea-ecuador-cedatos.html.

Shleifer, A., and Vishny, R. (1993). Corruption. The Quarterly Journal of Economics, 108(3), 599-617.

Singer, M., Carlin, R., and Love, G. (2016). Corruption in the Americas. In E. Zechmeister (Ed.), The Political Culture of Democracy in the Americas, 2014: Democratic Governance across 10 Years of the AmericasBarometer. LAPOP, 143-163.

Siverson, R. M., and Johnson, R. A. (2014). Politics and Parasites: The Contribution of Corruption to Human Misery. International Studies Quarterly, 58(1), 199-206.

Taj, M., and Politi, D. (2021). ‘V.I.P. Immunization’ for the Powerful Rattles South America. The New York Times. https://www.nytimes.com/2021/02/25/world/americas/covid-south-america-vaccine-corruption.html?.

Valencia, A. (2020). Ecuador’s former president found guilty of corruption. Reuters Media. https://www.reuters.com/article/us-politics-ecuador-idUSKBN21P373.

Villavicencio, F., Zurita, and Christan. (2019). Arroz Verde: La Industria del Soborno (1st ed.). Editorial La Fuente.

Weisbrot, M., Johnston, J., and Merling, L. (2017). Decade of Reform: Ecuador’s Macroeconomic Policies, Institutional Changes, and Results. Washington DC. Center for Economic and Policy Research. https://cepr.net/images/stories/reports/ecuador-2017-02.pdf.

(2016). The Political Culture of Democracy in the Americas, 2014: Democratic Governance across 10 Years of the AmericasBarometer. LAPOP. http://www.vanderbilt.edu/lapop/ab2014/AB2014_Comparative_Report_English_V3_Updated_040517_W.pdf.

(2019). Pulse of Democracy. Nashville, TN. LAPOP. https://www.vanderbilt.edu/lapop/ab2018/2018-19_AmericasBarometer_Regional_Report_10.13.19.pdf.

  1. Translated from Cerda, 2021 in “La asambleísta que dijo ‘si roben, roben bien’ se enreda más” (2021), paragraph 2.
  2. Data is from the IMF’s World Economic Outlook data set for October 2022.
  3. The original wording by the authors in the AB reports is corruption victimization. Here, this variable is referred to as corruption exposure, to account for the possibility that the respondent can be either a victim of corruption by being forced to pay a bribe or the initial corrupt agent who offers to pay one.
  4. In this case, not being unemployed means either being employed, salary and hours worked notwithstanding, and not being in the labor force (students, rentiers, among others). Results are robust to include an employment variable.


Share This Book