Political Institutions, Policymaking Processes and Service Delivery

Political Institutions, Policymaking Processes and Service Delivery

Ernesto Stein and Mariano Tommasi

Motivation• Natural inclination of economists is to emphasize policy recipes and reforms

as a way of improving the well-being of people in developing countries.• In Latin America during the 1990s, this has led to the adoption of a wide

variety of reforms, grouped under the “Washington Consensus” • The outcome of these efforts has been disappointing.• Our work is based on the belief that the potential of policy recipes to deliver

depends on the quality of the policymaking process through which those recipes are discussed, approved, implemented and enforced.

• Improvements in the policymaking processes – and a better understanding of these processes when policies are designed – are key in order to improve the quality of public policies, and achieve development objectives.

Reforms and Public Policies• We want to understand why some countries are able to adopt some reforms, while other countries are

not. But focus is broader than that.• We want to explain certain features of public policies, not just on the magical moment in which they

shift (the moment of reform) but on a more permanent basis.• We are interested, for example, in the quality of implementation and enforcement of public policies

once they have been approved;• in the extent to which they produce broad benefits for the population at large or concentrated benefits

to narrow groups;• in the degree to which policies tend to be stable (or change gradually), or shift abruptly every time

there is a change of administration;• in the extent to which there is innovation when policies fail;• in the extent to which different agencies in charge of the delivery of a certain service coordinate their

actions effectively;• All these dimensions of public policies should have an important impact on the quality of service

delivery.

The approach• We view public policies as the outcome of inter-temporal political transactions

among political actors.• Political institutions do not affect policy outcomes directly, but rather through

their impact on the process by which policies are designed, approved and implemented (the PMP).

• Most of the literature on effects of political institutions on policy outcomes tends to focus on a single institutional dimension, and explores its impact on some policy outcome.– Example: impact of proportional representation on government size.

• The workings of the PMP (and the impact on policy outcomes) do not depend on single-factor explanations but rather on a multiplicity of factors, and their interaction.

• For this reason, we adopt a more systemic or “GE” approach.

Country focus• Our approach is very demanding in terms of the knowledge of the institutional

details of the countries under study.• For this reason, in the first phase of this research agenda we focused the research

on very detailed country studies.• Phase I: Argentina, Brazil, Chile, Colombia, Ecuador, Mexico, Paraguay, Peru,

Uruguay and Venezuela. Now we are adding Costa Rica, Guatemala and Jamaica (still preliminary)

• The focus of each country study was to understand as thoroughly as possible the PMP in each country and its institutional determinants, as well as the mapping between the PMP and certain qualities of public policies.

• These country studies can then be used as inputs to try to understand differences across countries in the quality of service delivery in specific sectors (we are working on several sector studies for our annual report right now).

Outer features of public policies• For the country papers, the focus of the methodology was to explain

not the content of policies, but rather certain common features or qualities of public policies, characterized as“outer features”– Stability vs. volatility– Adaptability vs. rigidity– Coordination / coherence– Quality of implementation and enforcement– Public vs. private regardedness

• These features are often more naturally linked to the institutional environment than the content of policies themselves. – Content may shift back and forth within a weak institutional environment,

while features such as stability or adaptability may remain the same.

• Focus on outer features also allows comparability across policy issues.

Outer features of public policies• For sector studies, however, it may be more convenient to

explain some aspects of policies linked to their content, in addition to some outer features.

• For example, we want to know whether the national and subnational governments effectively coordinate the delivery of education services…

• …but we also want to know whether countries are able to introduce teacher incentives.

• We want to know if the regulation of public utilities is consistent over time (as opposed to shifting from capture to expropriation and back, depending on the government in power)…

• …but we also want to know whether countries are able to privatize public utilities.

The Policymaking Process (PMP)• Within our framework, the PMP takes center stage. A lot of

effort is spent in each country paper characterizing the PMP: – Which are the key actors that participate in it?

– What powers and roles do these actors have?

– What incentives and capabilities do these actors bring to the table?

– How long are their time horizons?

– What are the characteristics of the arenas in which they interact?

– How frequent are these interactions?

– What is the nature of the transactions they engage in?

Characterizing countries’ PMP• How did authors go about characterizing their PMPs?• By combining first hand knowledge of key policy actors and

their interactions with secondary sources and, in some cases, with new empirical work. Examples:– In Colombia and Paraguay, the authors illustrate their claims about the

changing role of the legislature in the PMP by carrying out empirical analysis of legislative activity in the different periods they consider.

– In Brazil, authors present empirical evidence about the connection between votes for the government in Congress and the appropriation of legislator’s budget amendments on the part of the executive, providing support for their story that the key exchange is one of pork for support.

– In Ecuador, team provides detailed account of the mechanics and the currencies used by the government to form coalitions and to (try to) keep them together (often unsuccessfully).

PMP II: The cast of characters• Series of background papers for IPES focused on the role of

different actors and institutions in Latin America’s PMPs• Political parties and party systems• The legislature (and the legislators)• The president • The cabinet• The bureaucracy• The judiciary• Regional authorities (their role in the national PMP)• Business interests• The media• In each case, we look at their role in the PMP across countries,

taking into account relevant interactions with other actors

(features of)PUBLIC

POLICIES

Y

POLICY-MAKING

GAME

XxZ→Y

FUNCTIONINGOF

POLITICALINSTITUTIONS

(rules ofpolicymaking

game)

X

General Equilibrium Interactions

BASICINSTITUTIONS

&HISTORY

features ofspecific

policy issues

Z

The framework

(features of)PUBLIC

POLICIES

Y

POLICY-MAKING

GAME

XxZ→Y

FUNCTIONINGOF

POLITICALINSTITUTIONS

(rules ofpolicymaking

game)

X

General Equilibrium Interactions

BASICINSTITUTIONS

&HISTORY

features ofspecific

policy issues

Z

The framework

PMP Policy Outcomes• We believe certain characteristics of the PMP play a key role in

determining policy outcomes.• One key dimension: ability of political actors to achieve cooperative

outcomes – that is, their ability to strike and enforce inter-temporal political agreements.

• In political environments that facilitate such deals, public policies will tend to be of higher quality.

• They will be more stable (less sensitive to realization of political shocks), more adaptable to changing economic conditions, and it will be easier to adopt welfare-improving reforms (since commitment to compensate losers would be credible).

• In environments that hinder cooperation, policies may be either too unstable (subject to political swings) or too inflexible, and many desirable reforms will not be adopted.

• Key question: what determines cooperation?

What determines cooperation?

• Number of actors with substantial impact on the policymaking game (related to number of veto players).

• Discount rate of actors (related to their tenure in office).• Extent of convergence or divergence of preferences.• Availability of credible enforcement technologies (such as an

independent judiciary, or a strong bureaucracy to which certain public policies can be delegated).

• Link between PMP and characteristics of public policies modeled by Spiller and Tommasi (2003)

ST Model: Games of political cooperation• Consider set of political actors who have a common interest in having policy

respond to economic shocks, but there is conflict due to different preferences.• Random political shocks shift the relative power of the players.• Optimal policies are invariant to the realization of political shock (i.e., do not

depend on the party in power) but contingent on economic shock.• If actors are patient enough, optimal policies can emerge in equilibrium. But if

their discount rate is high, full cooperation will not be achieved and policies will depend on the party in power (policy volatility).

• The existence of credible enforcement technologies (such as an independent court) which bound the players to their agreements may contribute to generate cooperative outcomes.

• Alternatively, players may agree to tie their hands and adopt rigid rules, which may restrict political opportunism at the cost of excessive rigidity.

• Absent the ability to enforce intertemporal deals, promises to compensate losers will not be credible, and some welfare-improving reforms will not be adopted.

(features of)PUBLIC

POLICIES

Y

POLICY-MAKING

GAME

XxZ→Y

FUNCTIONINGOF

POLITICALINSTITUTIONS

(rules ofpolicymaking

game)

X

General Equilibrium Interactions

BASICINSTITUTIONS

&HISTORY

features ofspecific

policy issues

Z

The framework

PI PMP• In turn, key aspects of the PMP are determined (among other

things) by the nature of the political institutions.

• For example, the number of actors, as well as their preferences, may be shaped by the nature of the electoral rules, which affect the configuration of the legislature.

• Constitutional rules (such as agenda setting power of the president) may affect the interaction between the relevant actors.

• Rules regarding the workings of political parties, as well as electoral rules, may determine whether the primary arena where political transactions take place is the legislature, or whether the key actors are the party leaders (or the governors).

(features of)PUBLIC

POLICIES

Y

POLICY-MAKING

GAME

XxZ→Y

FUNCTIONINGOF

POLITICALINSTITUTIONS

(rules ofpolicymaking

game)

X

General Equilibrium Interactions

BASICINSTITUTIONS

&HISTORY

features ofspecific

policy issues

Z

The framework

Policy outcomes may differ across issues• Different policy issues differ in their transaction characteristics.• Some may be more demanding than others in terms of the

enforcement of inter-temporal deals.• Some may have special outside enforcement mechanisms (such as

international trade agreements for the case of trade policy). • They may involve key specific actors (such as the teachers union for

education, or the Central Bank for monetary policy)• They may be played in different arenas.• They may differ on the degree of divergence of preferences.• Policy outcomes will be the result of the interaction between the

global PMP and the specific transaction characteristics / key actors/ relevant arenas of the sectors in question.

• Thus, policy outcomes will differ across issues, even within a given institutional environment.

The PMP in action: specific sector studies

• Tax reform / tax policy

• Budget Institutions

• Privatization / Regulation

• Social Protection

• Health

• Education

• Decentralization

Example 1: Education• Characterized by the presence of a very powerful actor: the teacher’s

union, highly informed and organized (in contrast to beneficiaries which are neither informed nor organized).

• Some key informational problems: policy impact visible only in long run, actions by schools and teachers hard to monitor.

• These features are common across countries. • Yet the way the game is played, and to some extent the outcomes, are

different across countries, as these characteristics are mediated by different PMPs.

• The discussion of Chile and Mexico this afternoon will illustrate this point.

• Navarro (2005) argues that some key differences are driven by the fact that, under different PMPs, the game is played in different arenas, some more transparent than others, involving different number of players, etc.

Example 2: privatization and regulation• Ability to enforce inter-temporal deals is crucial in the case of public utilities,

characterized by large sunk investments and thus by the possibility of expropriation by a opportunistic government.

• Opportunism exacerbated by massive number of consumers, who are also voters and can reward the government for expropriation (even when time consistency considerations will restrict future investments).

• On the other hand, capture of regulators by the regulated can tilt balance in the opposite direction.

• Bergara and Pereyra (2005) show how different country PMPs, combined with these features that are common across countries, contribute to produce different outcomes in Argentina, Chile, Peru and Uruguay.

• For example, in Argentina, the lack of cooperation in global PMP leads to wide swings from capture to expropriation, depending on the preferences of the administration in office.

• More on this tomorrow morning, as Mario will present the study.