Syndicate content

Opinion polls, priorities, and leadership

Nicholas van Praag's picture
     Interpreting their parents' concerns.  Photo © Curt Carnemark / World Bank

Polls matter to political leaders because they reflect public sentiment about what is important and what is not.  They are no crystal ball, however, and can be interpreted and acted upon in very different ways.  President Alvaro Uribe of Colombia, who will step down later this month, used polling data in his first term to justify a decisive break with past policies.  It is an interesting example for the WDR team as we look at how leaders in conflict-affected states prioritize policy actions and the extent to which they do so by listening to the people.

By 2002, after years of zero growth and declining living standards, Colombia's institutions were under increasing pressure from the narco-clans, the militia and the FARC.

With presidential elections in the offing, polls showed the country was almost equally divided between those who wanted more focus on social programming and poverty alleviation and those who thought the government needed to get tough and crack down on violence.

The response of a succession of governments to similar polling data over the previous decade reflected their quite different takes on popular sentiment. As a result, the policy pendulum swung from security to social programming and back. Some presidents started soft and ended hard when negotiations with the violent factions failed.

Then, in the presidential elections in 2002, Alvaro Uribe, the governor of the nation’s third largest state, came from obscurity to win the presidency. He did so by offering an alternative vision for tackling the country’s problems. It was not a choice between social programming and security, he insisted. People wanted both—and rightly so.

He argued that security and the legitimacy of the state were a necessary precursor to economic and social investments and insisted that the two policy strands should be rolled out with a short lag between them.

Essentially, Uribe re-interpreted what previously had been seen as lack of consensus among the electorate on which of the two courses to follow. He believed that people wanted both security today and economic hope for tomorrow. His response was to exercise determined leadership on the sequencing and timing of what he saw as the necessary policy actions.

The Colombia example shows that polling, however good, is not a substitute for vision. Nor does it necessarily make hard decisions any easier to take. What proved crucial in Colombia was the way President Uribe interpreted opinion data and used it to justify a new approach to old challenges.

This confirms what is coming through from the WDR research: common to successful leaders is their ability to redefine citizen and elite expectations and to transform public policies and institutions in ways that enable the state to address immediate and long-term sources of conflict.


Submitted by Anonymous on
While I find the thrust of the blog interesting, the writer uses facts in a lose way that breaks the logic of the point he is making. Fact 1: President Uribe DID NOT use polling data to justify the break from past policies. His governorship in Antioquia and his campaign for the presidency were completely centered around the "security" issue, and he did not justify himself with the country after his electoral victory. Fact 2: Polling is an imperfect instrument, especially when the samples being polled do not refect the population properly. This problem is quite acute around the world, but more so in less developed countries. A good example of how mistaken they can be comes presciely from Colombia. Prior to the first round presidential election of last Sunday (May 30), polls were placing the two front runners neck to neck. As it turned out, one of the candidates doubled the other one in the number of votes and share. Reliable? Good enough to base major policy decisions? Fact 3: Gratuitous may be a way of describing the statement "The response of a succession of governments to similar polling data over the previous decade reflected their quite different takes on popular sentiment." Nick seems to be ignoring that President Samper was elected with very strong support, at least financial, from one of the narco-clans, overshadowing polls and tainting his approach on security issues. He also seems to ignore the fact that political parties have strong views on some issues, which is what explains why Conservative President Pastrana's approach in the second half of the 90s was along the same lines Conservative President Betancour had followed in the early 80s. Fact 4: President Uribe did not come from obscurity as Nick states. He had been a star Senator before being a star Governor. His achievements both in the Senate and in Antioquia--the "California" of Colombia, were quite notorious and widely recognized at all levels. Fact 5: President Uribe did not break the mold on social spending issues during the campaign that took him to the presidency of Colombia. His key message then was that Colombia needed domestic security for growth. The strength of his governemnt in the social area came with the technical team that he chose to lead the key ministries. Fact 6: If the message in President Uribe's political campaign was domestic security for growth and he ended up winning by a wide margin, can anyone validly claim that he "re-interpreted what previously had been seen as lack of consensus among the electorate"? Fact 7 (trivial): President Uribe will be stepping down on August 7, not next month as Nick mentions. Sure, the Colombian example shows that leadership matters. The story on the polls seems more like "magical realism." One would hope that the WDR will be grounded in more solid research.

Thanks to the anonymous reader who sent these comments about my blog on President Uribe and polling. Some points are on the mark; others less so. But all are interesting and I appreciate the effort. President Uribe was indeed a successful Governor and Senator before running for President but he was relatively unknown nationally. Polls show that fact and I believe this makes his campaign that much more heroic. The debate about polling is often polarizing. Polls are not crystal balls, as I said in my blog. But they are a badly-needed instrument to understand where public opinion –- and base voters -- stand in relation to the policies of a leader. They cannot replace a leader's vision. Rather, they help leaders understand how best to communicate their vision. Before the break up of the peace talks between the Pastrana administration and the Farc, polling data showed some schizophrenia: people hated the guerrillas but were torn between a military confrontation and peace talks. There is a long tradition of peace talks in Colombia with wide popular support. In 2001 and 2002 people changed their views dramatically. While it is true that Uribe had security as his banner issue early on, it is also true that polling data showed how public opinion had changed and moved towards his position. During his second administration, probably based on polling information that showed growing concern for economic and social issues, he tried to link progress on security with the improvement of social and economic conditions. Again, he was not using the polls as a replacement for his vision but, like most leaders, he used them to help read the map. And yes, President Uribe is stepping down in August. I thought Colombia had a shorter gap between election day and inauguration. I got that wrong. Apologies.

Submitted by Anonymous on
The response from anonymous is on the mark. Problem with these blogs is people commenting and making superficial statements on countries and issues they know little about, and then drawing conclusions that do not hold much water. It is not about polls, this is a trivial point. Colombians have been debating long and hard about how to achieve peace, and have done tremendous research--the polls are the least interesting aspec of Colombia's fascinating experience.

Add new comment