This is a guest post by Gregory J. Love and Ryan E. Carlin
What makes political executives powerful? Students of comparative political institutions have painstakingly pored through parchment and practice to elaborate detailed indicators of executive powers. Such measures have advanced our understanding of executive power tremendously. Yet, arguably, one of executive’s most powerful tool can’t be found in the text of constitutions, legal statues, or norms. And unlike the powers deriving from these sources, this crucial tool is often ephemeral and downright fickle. It is the power of popularity.
A quick glance at news coverage of democratic national leaders will show just how fascinated the media and political elites alike are with popularity ratings. This is true whether the head of government is a president or a prime minister, whether scheduled elections are next week or years away, or whether the leader is enjoying a “honeymoon” or is a “lame duck.” But why are approval ratings of such widespread interest when, in theory, citizens’ decisions vis-à-vis representation and accountability are taken at the ballot box?
For prime ministers, popular support is central to job security. When leaders are viewed as “weak” – often a synonym for being disliked by the electorate – they risk votes of no confidence and internal party removal. Of course, popularity matters not only at its nadir but at its apex, as well. High ratings strengthen the prime minister’s hand when it comes to keeping coalition partners and dissident party factions in check. The threat of a no-confidence vote rings hollow when the incumbent prime minister’s party stands to gain at new elections. Popular prime ministers, thus, can more easily see their agenda implemented.
For presidents, public approval is no less powerful. As Stimson put it, “If the real power of the presidency is not directly proportional to the most recent Gallup popularity rating, it is not far from it” (1976, 2). Yet institutional features of presidentialism – fixed terms, separation-of-powers, and in some cases non-concurrent legislative or sub-national elections – add nuance to presidents’ ability to harness public approval. Popularity is probably more valuable early in their terms, when they seek to make hay during a “honeymoon” period. Popularity’s political value is lowest when term limits turn presidents into lame ducks. Dynamics that do not match a cyclical pattern of honeymoon, decay, and a slight uptick as elections near may be precursors of instability. For example, extremely unpopular presidents can face the same prospects as a prime minister: removal from office (such as Dilma Roussef in Brazil).
While the attention that political elites and the public at large shower on executive approval ratings is warranted, we know relatively little about executive approval in comparative perspective. Aside from the (inconsistent) evidence that the economy matters and some vague sense that rally ‘round the flag events do, too, it is entirely unclear how context shapes the dynamics, drivers, and policy implications of executive approval.
The Executive Approval Database 1.0
Our ability to carry out comparative research on executive approval has been seriously limited by lack of high quality comparative longitudinal macro-public opinion data. To address this shortcoming, in 2016 the Executive Approval Project (EAP) launched the Executive Approval Dataset (EAD 1.0).
Producing comparable data on presidential and prime minister approval across countries, time, and regions presents myriad challenges. First is data collection. The EAP’s goal is to gather the universe of time series (survey marginals) tapping respondents’ assessments of executive performance. Its web-based visualization tool shows the breadth of data collection. Most of 30+ countries in the EAD have between 12 and 36 approval series from public and private polling sources.
The second challenge is aggregation. Differences in question wording, response categories, and periodicity as well as missingness complicate valid analyses with comparable data. Premised on the idea that these series tap the same underlying construct, the EAD uses Stimson’s (1991; 2015) dyads algorithm to create a single, continuous, and smoothed measure (to account for sampling variance and differing frames). A detailed description of the algorithm and software (WCalc) can be found here. While Stimson’s approach is widely used, and easily implemented via the EAP’s visualization tool, it is not the only option. Researchers are welcome to download the EAD’s input series from the website and use their own aggregation approaches.
The EAD 1.0 contains over 11,000 survey marginals and 330 time series pertaining to over 140 presidents in 18 Latin American countries. Using these data, we produce the following figure, which represents the most broadly comparable measurement of presidential approval to date.
Research Applications of the EAD
Over the past six years a growing number of researchers have used the EAD. We can group the existing work into two broad categories: (1) papers explaining the dynamics of executive popularity and (2) papers focusing on the effects of executive approval on politics and policy.
In the first group of projects, researchers look at factors such as economic performance, corruption and scandals, terrorism, populist rhetoric and policies, and institutional constructs to understand when leaders fail (or succeed) in the eyes of the public. As for the consequences of executive popularity, the works are fewer but growing rapidly. New papers explore the role of approval on policy agendas, executive decrees, and other aspects of the executive-legislative process. A bibliography of papers using the EAD can be found here.
The EAP aims to update and expand the dataset’s coverage in Europe, Asia, North America, Latin America, and Africa for the EAD 2.0 release in summer 2018. To that end, the EAP is establishing Country Teams with partner institutions around the world. And we are always looking for new partners!
If you wish to contribute, or to collect, executive approval data for a particular country, or set of countries, we’d love to enlist your help. Please contact the authors for more information.
The Executive Approval Project Core Team is:
Ryan Carlin, Georgia State University
Jonathan Hartlyn, University of North Carolina at Chapel Hill
Timothy Hellwig, Indiana University
Gregory Love, University of Mississippi
Cecilia Martínez-Gallardo, University of North Carolina at Chapel Hill
Matthew Singer, University of Connecticut