When Justifiable Anger Leads to Bad Policy: The Unintended Consequences of Colombia’s Anticorruption Referendum

Last August, Colombia held a national referendum on seven anticorruption measures. Despite the fact that six of these measures had previously been proposed in, but failed to pass out of, the lower house of the legislature, popular support for the measures was overwhelming: each measure received 99% “Yes” votes. The referendum did not pass, however, because even though more people voted “yes” on the referendum than voted for the current President, under Colombian law the referendum would only pass if a quorum of 12.1 million citizens voted, and the 11.6 million voters who turned out fell short of that number. Nonetheless, proponents of the referendum declared it a success because it has put public pressure on Colombia’s political leaders to implement these measures. And indeed, President Duque has convened an anticorruption roundtable and vowed to implement all seven measures by December 2018.

Is this a good idea? It’s certainly the case that Colombia needs to do more to combat corruption, which is estimated to cost Colombian taxpayers at least $17 billion a year. But it’s not clear that all of the proposed solutions, though doubtless well-intended, are good public policy. I won’t attempt a comprehensive review of all seven measures here. I’ll put to one side discussion of those measures that focus on improving transparency (for example, by publicizing government budgets, legislators’ voting records, and public officials’ tax returns and asset declarations) or on making penalties more severe (for example, requiring those convicted of corruption to serve their full sentences, and nullifying government contracts with parties convicted of corruption). Rather, I want to address two measures that target Colombian legislators: one of these measures would impose a three-term limit, while the other would substantially cut legislators’ pay.

These two measures appear to reflect understandable public anger at how legislators have abused their positions for private gain. But this retributive impulse may produce bad policy. Indeed, both term limits and salary cuts are likely to prove counterproductive in the fight against corruption in Colombia.

  • Term limits. Many anticorruption organizations have called for legislative term limits, on the logic that limiting the time legislators and other public officials can remain in office erodes their ability to build relationships with special interests and solicit bribes. However, term limits may also lead to more corruption, for three reasons. First, once officials are in their last term, they no longer have to worry about their re-election chances, and without that democratic check, legislators may be more inclined to engage in corrupt activities because they no longer have to earn the public’s vote. Second, legislators who know they will only be in office for a relatively short period of time may adopt the mindset of grabbing what they can while they can, without as much concern for long-term consequences. When politicians know there’s a possibility they could build a long career in government, they may be take a longer-term view. Third, introducing relatively short term limits can increase legislative turnover and instability, and this can make it easier for special interest groups to “capture” legislators. Many legislators are not experts in all policy areas when they come into office, and therefore often rely on information provided by outside stakeholders to make decisions. Thus, legislators with limited experience may be more reliant on special interests than career politicians, as the latter can build up their own expertise over time, leaving them potentially less susceptible to outside influence and more independent in their decision-making.
  • Cutting legislative pay. The pay cut proposed by the referendum would reduce the legislator pay cap from 40 times Colombia’s average salary to 25 times the average salary—a 40% salary cut for lawmakers, who currently make the equivalent of US$124,000 per year. It is unclear if this measure is meant to combat corruption, or is just a form of retribution against elected officials, given the perception that most or all of them are tainted by corruption. However, there is ample evidence to suggest that paying public officials, including legislators, a low salary can increase the chances of corruption, as low-paid public officials and civil servants have stronger incentives to supplement their official salaries by soliciting bribes or misappropriating public funds (see here, here, and here, but see a potential counterexample here). While Columbian lawmakers earn very high incomes by Colombian standards and would still be wealthy under this new scheme, most people who experience a sudden 40% drop in income would feel pressure to find some way to make up the difference and maintain their previous standard of living. Thus, this measure may in fact increase a legislator’s motivation to engage in corrupt activity.

Overall, while the Colombian anticorruption referendum represented a significant movement in the fight against corruption in Colombia, it seems that perhaps retributive impulses may have led to the inclusion of certain measures that will prove counterproductive, at least with respect to fighting corruption.

19 thoughts on “When Justifiable Anger Leads to Bad Policy: The Unintended Consequences of Colombia’s Anticorruption Referendum

  1. Really interesting post. I’d be curious where the seven measures initially included in the referendum came from. Do they reflect societal consensus about what changes are necessary? Or, did the civil society group that proposed the referendum put together the list of reforms which were then voted for in-mass? It seems like Duque’s mandate will (or maybe should) be influenced by the answer to that question. Either way, I’m genuinely amazed by how high public officials’ salaries are in Colombia. I could imagine that if one wanted to decrease the salary of public officials at least to some extent — which seems necessary — there might be some strategic value in suggesting a higher decrease at the outset only to have it rolled back a bit.

    • Thank you for your comment, Hilary! The effort to get these anti-corruption measures on the ballot was led by two Senators from the Green Alliance Party, Claudia Lopez Hernandez and Angelica Lozano Correa, who created the Consulta Popular Anticorrupción movement. They turned to the referendum process after failing to pass any of their proposed measures in the legislature. They pulled together the list of measures en masse and got signatures from 5% of the population to propose the referendum. The final referendum was approved in the Senate in order to get on the ballot.

      In terms of civil society, most of the major political parties in Colombia did not support the referendum, with some saying that the referendum represented more empty measures, or that they preferred other measures that were proposed in Congress. Some experts said the referendum would do little to curb corruption in the police and judiciary. However, Andres Hernandez, Senior Programme Coordinator for Transparency International Colombia, noted, “”If millions of citizens vote in favour of it, it will have a strong symbolic effect on the political class.” For many, it seemed like a movement by the people against the political elite. But you bring up a good point that it is important to consider the source of these proposed measures in order to understand why they were proposed in such a way.

      Perhaps you are right that Senators Hernandez and Correa expected the referendum not to pass and so they wanted to set their initial ask on the high end in order to get some progress on this issue. However, they came very close to passing the referendum, meaning that their strategy could have backfired.

      For more info on the thoughts of civil society groups/political parties:

  2. Thanks, Christina. Your points make a lot of sense. I wonder, on the second point, how legislative pay affects the pool of candidates interested in public office. And it seems like the analysis cuts in different ways. Intuitively, the higher the salary the more likely the job is to attract those coming from high-paying jobs who might have otherwise balked at taking too large a pay cut for public service; at the same time, the higher the salary perhaps the more likely it is that lower-income office-seekers will see the position precisely because of the salary and as a means to amassing wealth (and a politician with that mindset heading into office would be at a heightened risk of corrupt behavior). But with lower salaries, the job may only attract those who still see the job as paying relatively well compared to their prior lower paying job or those who previously held higher-paying jobs but are confident that 1) serving the public is still worth it (great), or 2) they can find ancillary benefits to recoup for the last income (terrible).

    My analysis is admittedly very reductive, not only because it ignores the independently wealthy as potential candidates. Your post seems targeted to the reactions of the current legislators to their slashed paychecks, which is certainly an important component, but I imagine that many of the most impactful effects will be seen down the road.

    • Thanks for your thoughts, Kees! I actually had not considered the impact that this referendum might have on pool of candidates that may run for office in the future. The term limits should allow for a wider variety of candidates as people are term-limited out of office and new faces come in. I agree that the most charitable interpretation of what might happen is that good-hearted public servants will heed the call and run for office to serve their country and not to make money. Another possible view: seeing the rapid decline in salaries may project the notion that one’s salary in public office is volatile. This may discourage those who need a steadier source of income and are more risk-averse from running, while encouraging those who are more risk-taking to run. Not all risk-takers are criminals, but committing a crime (particularly a political crime like corruption) requires taking some risk. Thus, having more risk-takers in the pool of candidates may lead to more corrupt activity.

  3. Thanks for the great post, Cristina. Like Hilary, I’ve also been wondering about the legislative salary component of this set of proposals. I wonder if there’s a particular sweet spot of compensation, or maybe it’s better to think of it as a threshold, below which corruption becomes significantly more likely. This is sort of a bizarre comparison, but I’m thinking of those studies that show that, in the U.S., money does correlate with people’s overall happiness up to a certain point, and then every dollar an individual earns beyond that doesn’t seem to have anything to do with happiness. Is it possible there might be a similar effect associated with corruption? Perhaps not, but I don’t know. It’s unclear to me whether or not this threshold would be a matter of dollar amounts or, as you’ve framed it, a certain multiplier above the average salary of a jurisdiction’s citizenry, but it’s at least not totally crazy to me that, once a public official is paid enough more than most of her constituents or colleagues, the chance that she might engage in, say, petty corruption would drop significantly. (Even if the chances that she might engage in grand corruption were unaffected). I have no idea if anyone has looked into this question, and, admittedly, it’s only tangentially related to your piece here on Colombia so I don’t expect that you’ve research it. But I’m putting it out there because this kind of question seems potentially relevant to an evaluation of the legislative pay measure in Colombia’s referendum. If, it turns out, that 25x average salary is still above a hypothetical threshold multiplier, then this measure might not be as problematic as we’d otherwise expect.

    • Thanks for your comment, Jason! Many economists have theorized that if you raise the salaries of civil servants, their incentive to engage in petty corruption should go down. For the “good legislators” who only felt that they needed to engage in corruption to feed their families or maintain their standard of living, this alternative source of funding should be enough to dissuade them from continuing to be corrupt. However, I imagine that there would be many factors that could go into a public employee’s cost-benefit analysis when deciding whether to engage in corruption. For starters, how likely are you to get caught? If the likelihood of facing prosecution has not changed, a legislator may take the pay raise and continue engaging in corruption. There may be a social norm of corruption that one would have to address with a corresponding educational campaign. Also, the shift in salary should be gradual to avoid the scramble that can come from people feeling a 40% pay cut. All this to say that I think there are many possible factors that could go into finding the threshold/ideal salary for legislators to minimize the number engaging in petty corruption. Thanks again for your thoughts!

  4. Thanks Cristina. This is a really interesting post, both in terms of the fact that such an anticorruption referendum was held in Colombia and in terms of the arguments that you raised regarding two of the measures that were eventually implemented.

    I have to say that I find it extremely intriguing that despite the enormous percentage of people who voted “yes” on the referendum, the total number of voters was not sufficient to legally pass the referendum. Do you happen to know what theories (if any) have been put forth to explain this phenomenon?

    With regard to your argument against legislators’ term limits: While I do not know about Colombia specifically, I think that generally speaking, the fact that an official can serve for more than one term (or in other words, be re-elected) might in and of itself lead to more corruption. Since that official’s re-election is naturally dependent on people’s support, there is a serious risk that she would employ various corrupt means to increase the amount of people who would vote for her. In this sense, one may argue that term limits are not favorable only because the legislators can remain in office for a shorter period of time, but also because they have fewer opportunities to be re-elected.

    • Thank you for your comment, Guy! There was a quorum requirement for the referendum. My understanding is that the referendum needed 12.1 million participants in order to pass. So, even if the remaining 0.4 million people had voted “No” the referendum would have passed. The proponents of the referendum just needed to meet the minimum number of overall participants. I imagine this is to prevent referenda from passing and affecting the entire nation when only a small fraction of the population actually votes.

      You do bring up a good point that opportunities for re-election can introduce the potential for corruption. One counterpoint might be that allowing for this level of democratic accountability may minimize corruption, as legislators have to earn the votes of those they represent. Provided that there are strong limits on campaign contributions, legislators may stand to benefit more from addressing the needs of the people than from serving the interests of corrupt actors.

      • Thanks for taking the time to respond to my comment, Cristina! Regarding the insufficient number of participants in the referendum, I cannot help but wonder whether many people decided to stay home that day and not to vote merely because they were sure that the vast majority of voters would vote anyway in favor of the proposed anticorruption measures. (That in turn raises the question – which I am not sure can really be answered – whether such a referendum was actually required to begin with to pass these measures).

  5. Great post! I was interested to learn more about the ways in which new term limit and legislator salary reforms may backfire. But I also wonder to what extent these dangers have already been accounted for by the other reforms in the anticorruption referendum “bundle.” For example, you mention that some of the anticorruption reforms deal with increasing transparency. It seems plausible to me that increased transparency could mitigate some of the opportunistic wealth-seeking of legislators who know they aren’t up for reelection.

    • Good point, Keller! Among the reforms included in the bundle, there are two measures that aim to specifically restrain potentially corrupt legislators:
      1. Requiring members of Congress to be open about which bills they proposed, how they voted, and who lobbied them.
      2. Requiring elected officials to disclose their assets, income, and tax returns.
      I think these measures sound very promising. My point in this post is that the term limits and pay cut introduced in the referenda could work against these potentially very effective reforms.

  6. This is such an interesting post! In regards to the term limits, I understand your concerns about relatively short term limits, but I wonder what you consider to be “relatively short”. It seems to me that a three-term limit seems to be a reasonable amount of time, and might allow for legislators to develop some expertise without having to fully rely on special interest groups. Is there a term limit that you would suggest instead?

    • Thank you for your comment, Signa! I agree that the length of the term limit is important for analyzing whether it would be counterproductive in the fight against corruption. For me, it would be important to have a term that is long enough for a legislator to build some expertise and confidence in governing in order to be less vulnerable to capture. On a separate but somewhat related note, in Colombia’s neighbor Peru, there is a referendum going on today that would ban the immediate re-election of legislators to the same post. While the length of the term is important, it is equally important to allow legislators the opportunity to run for re-election in order to ensure that they remain democratically accountable to the people that they serve.

  7. Thank you, Cristina, for your smart take on the two proposed anticorruption measures. I am curious about the term limits you discuss here. What were the considerations in Colombia in proposing the legislative term limits? In particular, I wonder whether the term limits—in addition to the relationship piece you discuss—were considered as a method of getting rid of the bad apples in the legislature that continue to stay in office despite their corrupt behavior. Three terms (although I have no knowledge of how long they are) seem appropriate to ensure that effective legislators are able to enact meaningful initiatives while in office, while ensuring that the corrupt legislators do not stay too long. But, as you mention, having term limits may backfire in certain circumstances. I’m curious how these proposed measures would play out if enacted.

    • Thanks for your comment, Vicky! You are right to point out that one upside to instituting term limits is to force out long-time corrupt actors. However, there is one other potential problem with the term limits provision proposed in the referendum. The limit of three terms applies to the same position, meaning that one person cannot serve more than three terms as Senator. However, they can still run for the House of Representatives and serve three terms in that office. Therefore, what may seem like a reasonable limit could actually double because of the bicameral nature of the legislature. In fact, in Peru, there is a referendum happening today that would not allow representatives to serve more than one term in the legislature. In response to this proposed measure, the current legislature added another question to the referendum that would add a second chamber to the legislature, some say precisely to give themselves another position to run for. All this to say that the “bad apples” in the Colombian legislature may be able to skirt these new limits and remain in power for longer.

  8. Thanks for such an interesting post, Christina! I especially appreciated your point about term limits since it contradicts the common view (and campaign tagline of challengers) that incumbents tend to become more captured over time.

    What you said about retribution in the second bullet jumped out to me. When I began thinking about term limits and legislative pay cuts as part of a strategy of retribution against politicians, the logic of these measures seemed clearer to me—as a more-or-less coherent effort to punish lawmakers for their excesses. Setting aside for a moment the side effects you lay out (and acknowledging that they may well dwarf the intended effects of the measures), there is some force to an argument that (1) if lawmakers have engaged in corruption, we should impose costs on them, perhaps in the forms of lower salaries and terms limits, to deter future corruption. One might go further and say that (2) even if most lawmakers are corruption-free, these measures signal public frustration with lawmakers’ failure to pass anticorruption measures themselves. If lawmakers see the measures as punishment—rather than as anticorruption policies per se—the changes might motivate the legislature to take corruption seriously. Coupled with a credible promise of higher salaries and longer terms if lawmakers someday clean up their acts or adopt reforms, these measures make some sense.

    Despite its appeal, I see at least two problems with the argument above. First, the punitive signal to legislators seems fuzzy at best, since the measures are framed as anticorruption reforms and not as retribution for corrupt lawmakers or as incentives to take corruption seriously. Second, and more importantly, these measures would only change future legislative behavior if each individual lawmaker thinks her fortunes will rise or fall because of what she does. But because lawmakers are a large and diverse group, I think there are real problems of monitoring, collective action, and short-termism in trying to incentivize an entire legislature with measures like these. A single legislator, for instance, would probably be right to conclude that her individual actions will have little bearing on aggregate legislative salaries and term limits. (Not to mention the possibility that a salary increase unaccompanied by longer term limits might come too late, after she had already served three terms!)

    • Thank you for your comment, Jonathan! It is interesting to consider the deterrence effect of these “retributive policies” as their own way of dis-incentivizing legislators from engaging in corrupt acts and from incentivizing them to pass anti-corruption measures. Another proposed measure from the referendum that may support this version of events is the measure requiring those convicted of corruption crimes to serve their jail terms without the possibility of house arrest. This could serve as a strong signal to corrupt legislators that their status as legislators and/or “their friends in high places” cannot get them out of trouble if they were ever caught. I agree that there may be a problem with trying to “punish” all legislators for not caring enough about corruption. Aside from the concerns you brought up, another way this may backfire is by leading legislators to pass fluff legislation that sounds like it would fight corruption, but not actually give it any teeth. This could make it harder for the public to discern which laws are genuine anti-corruption efforts and which are just meant to please them.

  9. When reading through your post, I could not help, but wonder whether a referendum in such a field makes sense at all… I mean, most of these questions (as also highlighted above by others here in the comments) are highly controversial, complex and have very far reaching consequences. I feel like some of the questions that were asked are actually too complex to be allowed for citizens to decide in a popular vote… I mean, I would like to see detailed numbers and forecasts, analysis of practices in similar countries, etc., before I would even form an opinion. Making this a popular vote just feels a bit… populist? At the same time, I understand that if the state of affairs is really grim and it becomes hard to bring any political change however well-based and grounded – then a referendum can actually act as a good way to kick start reforms as it clearly shows the opinion of the citizens – and politicians still care about them (if for nothing less, then for their votes).

    • Ruta, I agree with many of the sentiments you are describing here. I am actually in the process of writing a second post analyzing whether popular referenda are a useful tool in the fight against corruption. Broadly speaking, I think there is a lot of risk with using popular referenda in this area because of the complexity of the policies at hand and the potential for retributive policies to take hold in an effort to galvanize more public support for the referendum. However, I do think the method could hold some promise in terms of forcing the issue of corruption to center stage and making politicians act against their own self-interest.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.