The Role of Governors in the National PMP: A Comparative Perspective of Latin American Cases Francisco Monaldi IESA and UCAB Inter-American Development.

The Role of Governors in the National PMP:A Comparative Perspective of Latin American Cases

Francisco MonaldiIESA and UCAB

Inter-American Development Bank

February, 2005

2

Road Map

• The Role of Governors in the PMP across Latin America

• The Sources of the Governors’ Bargaining Power

• Political Institutions and the Role of Governors

• Party Structure and the Role of Governors

• The Role of Governors in Established Federal Countries (Argentina and Brazil)

• Political Decentralization and the PMP (Colombia, Mexico and Venezuela)

3

Governors in the Region

• Significant differences across the region

• In Argentina and Brazil governors have often played a highly significant role in the national PMP

• In Colombia, Mexico and Venezuela they did not play a significant role until the last decade

• In Bolivia, Chile, Costa Rica, Ecuador, Paraguay, Peru and Uruguay they still play a minor role

• With political decentralization, trend towards increasing role across the region

4

The Players and their Incentives

• Presidents and national party leaders care about national policy

• Governors, depending on the institutional setting, might care for regional issues at the expense of national policies

• Coordination / Commons Problem

5

How do Governors Play a Role

• Influence on provincial delegation to the legislature

• Control over some policy areas that might block national policy and attempts to reform

• Capacity of creating fiscal and macroeconomic problems provides them with some bargaining power

• Influence in presidential reelection and national party finance

6

Political Institutions and Role of Governors in the PMP

1. Method of selection of governors

2. Authoritarian vs. Democratic

3. Federal vs. Unitary Constitutional Structure

4. Territorial Bicameralism

5. Malapportionment

6. Timing of elections (concurrency)

7. Fiscal Federalism Rules

8. Presidential Prerogatives

9. Internal Structure of Congress

7

Method of Selection

• Appointment / Removal by President– The principals are the president and national party leaders

– Very limited role in national PMP (e.g. Chile, Costa Rica, Venezuela before 1989)

• Direct Elections– If reelection possible electoral connection creates strong

regional incentives

– Even if there is no reelection they have some regional incentives to advance political career

– Endogenously tends to produce fiscal decentralization

– Most significant role in countries with long standing tradition of regional elections

8

Table I. Some Institutional Variables in Latin AmericaCountry Selection Since Const. Bicam. Terri. Ch. Concurrency (5) Nomination Malap. LC (1) Malap. UC (1) Reelection

Argentina Elected 1912 Federal Yes Yes Partial Subnational 0,14 0,49 YesBolivia Appointed n.a. Unitary Yes Yes n.a. National 0,17 0,38 n.a.Brazil Elected 1891 Federal Yes Yes All Subnational 0,09 0,4 YesChile Appointed n.a. Unitary Yes Yes n.a. National 0,15 0,31 n.a.Colombia Elected 1991 Unitary Yes No All Mixed 0,13 0 YesCosta Rica Appointed n.a. Unitary No No n.a. National 0,02 n.a. n.a.Ecuador App. / Elect. 1998 Unitary No No All National 0,2 n.a. ?Mexico Elected 1979 Federal Yes Yes Partial Mixed 0,06 0,23 NoParaguay Elected 1991 Unitary Yes No All National 0,04 0 ?Peru No Governors (3) n.a. Unitary No No n.a. National 0 n.a. n.a.Uruguay Elected (4) 2000 Unitary Yes No All National 0,03 0 ?Venezuela Elected 1989 Federal No No Partial Mixed 0,07 0,33 (2) Yes

9

Regime / Constitutional Structure

• Authoritarian regimes tend to re-centralize power, including suppressing direct elections in regions (historically military regimes in Argentina and Brazil, more recently Fujimori, Chavez?)

• Important role of sub-national elections in democratic transitions in Brazil and Mexico

• Federal constitutional structure is relevant only when certain elements are in place. In Mexico and Venezuela federalism was not relevant until there were competitive regional elections. The only formally Unitary country in which governors have a significant role is Colombia

10

Table I. Some Institutional Variables in Latin AmericaCountry Selection Since Const. Bicam. Terri. Ch. Concurrency (5) Nomination Malap. LC (1) Malap. UC (1) Reelection

Argentina Elected 1912 Federal Yes Yes Partial Subnational 0,14 0,49 YesBolivia Appointed n.a. Unitary Yes Yes n.a. National 0,17 0,38 n.a.Brazil Elected 1891 Federal Yes Yes All Subnational 0,09 0,4 YesChile Appointed n.a. Unitary Yes Yes n.a. National 0,15 0,31 n.a.Colombia Elected 1991 Unitary Yes No All Mixed 0,13 0 YesCosta Rica Appointed n.a. Unitary No No n.a. National 0,02 n.a. n.a.Ecuador App. / Elect. 1998 Unitary No No All National 0,2 n.a. ?Mexico Elected 1979 Federal Yes Yes Partial Mixed 0,06 0,23 NoParaguay Elected 1991 Unitary Yes No All National 0,04 0 ?Peru No Governors (3) n.a. Unitary No No n.a. National 0 n.a. n.a.Uruguay Elected (4) 2000 Unitary Yes No All National 0,03 0 ?Venezuela Elected 1989 Federal No No Partial Mixed 0,07 0,33 (2) Yes

Political Institutions and the Role of Governors in Latin America

11

Territorial Bicameralism / Malapportionment

• In some countries (federal or not) there exists a territorial chamber, typically with high malapportionment

• Different electorate from presidential election creates separation of powers and purpose (additional veto point)

• Policy scope might be symmetrical to population based chamber

• There exists high malapportionment in many lower chambers in Latin America

• Over-representation has effects on fiscal allocations (Argentina, Brazil, and Mexico)

• If combined with influential governors and regionally motivated legislators can have a significant effect, introducing rural bias

12

Table I. Some Institutional Variables in Latin AmericaCountry Selection Since Const. Bicam. Terri. Ch. Concurrency (5) Nomination Malap. LC (1) Malap. UC (1) Reelection

Argentina Elected 1912 Federal Yes Yes Partial Subnational 0,14 0,49 YesBolivia Appointed n.a. Unitary Yes Yes n.a. National 0,17 0,38 n.a.Brazil Elected 1891 Federal Yes Yes All Subnational 0,09 0,4 YesChile Appointed n.a. Unitary Yes Yes n.a. National 0,15 0,31 n.a.Colombia Elected 1991 Unitary Yes No All Mixed 0,13 0 YesCosta Rica Appointed n.a. Unitary No No n.a. National 0,02 n.a. n.a.Ecuador App. / Elect. 1998 Unitary No No All National 0,2 n.a. ?Mexico Elected 1979 Federal Yes Yes Partial Mixed 0,06 0,23 NoParaguay Elected 1991 Unitary Yes No All National 0,04 0 ?Peru No Governors (3) n.a. Unitary No No n.a. National 0 n.a. n.a.Uruguay Elected (4) 2000 Unitary Yes No All National 0,03 0 ?Venezuela Elected 1989 Federal No No Partial Mixed 0,07 0,33 (2) Yes

Political Institutions and the Role of Governors in Latin America

13

Malapportionment in Senate

0

0,1

0,2

0,3

0,4

0,5

0,6

14

Malappportionment in Lower Chamber

0

0,05

0,1

0,15

0,2

0,25

Source: Snyder and Samuels, 2004

15

Election Timing (Concurrency)

• The timing between national legislative elections, presidential elections, and regional elections determines if the presidential or gubernatorial coattails are relevant for the election of legislators

• In addition presidential coattails or support might be relevant for governors’ elections

• In case congressional elections are fully concurrent with regional elections and separate from presidential gubernatorial coattails have been important and there has been a tendency towards higher fragmentation. In Brazil 1982, 1986 and 1990 and Venezuela 1998 this effect appeared to be relevant

• If presidential coattails are quite significant fragmentation tends to be lower

16

Table I. Some Institutional Variables in Latin AmericaCountry Selection Since Const. Bicam. Terri. Ch. Concurrency (5) Nomination Malap. LC (1) Malap. UC (1) Reelection

Argentina Elected 1912 Federal Yes Yes Partial Subnational 0,14 0,49 YesBolivia Appointed n.a. Unitary Yes Yes n.a. National 0,17 0,38 n.a.Brazil Elected 1891 Federal Yes Yes All Subnational 0,09 0,4 YesChile Appointed n.a. Unitary Yes Yes n.a. National 0,15 0,31 n.a.Colombia Elected 1991 Unitary Yes No All Mixed 0,13 0 YesCosta Rica Appointed n.a. Unitary No No n.a. National 0,02 n.a. n.a.Ecuador App. / Elect. 1998 Unitary No No All National 0,2 n.a. ?Mexico Elected 1979 Federal Yes Yes Partial Mixed 0,06 0,23 NoParaguay Elected 1991 Unitary Yes No All National 0,04 0 ?Peru No Governors (3) n.a. Unitary No No n.a. National 0 n.a. n.a.Uruguay Elected (4) 2000 Unitary Yes No All National 0,03 0 ?Venezuela Elected 1989 Federal No No Partial Mixed 0,07 0,33 (2) Yes

Political Institutions and the Role of Governors in Latin America

17

Fiscal Federalism Rules1. Taxing authority and revenues collected by sub-national officials

2. Expenditure Decentralization

3. Vertical Imbalance

4. Borrowing autonomy

5. Central discretion over transfers

6. Soft budget constraints

18

Expenditure Decentralization and Vertical Imbalance

0%

10%

20%

30%

40%

50%

60%

70%

80%

90%

Exp. Desc. (1) Vertical Imb. (2)

Source: Stein, 1999

19

Fiscal Decentralization

0

0,1

0,2

0,3

0,4

0,5

0,6

Subnational/Total Exp. Ow n Sub Rev/ Total Rev

Source: Rodden, 2004

20

Borrow ing Autonomy

00,5

11,5

22,5

33,5

44,5

5

Argen

tina

Bolivi

a

Brazil

Colom

bia

Mex

ico

Parag

uayPer

uUSA

Borrow ing Autonomy

Source: Rodden, 2004

21

Presidential Prerogatives

• Presidential prerogatives may provide a powerful negotiation tool with governors and legislators

• Partial veto (or line-item veto) can be used to obtain support from legislators (Brazil)

• Power of appointment and removal of governors

• Power of intervention of regional governments

• Discretion over the allocation of resources to the regions

22

Party Structure and the Role of Governors

1. Nomination Powers• By sub-national authorities (Argentina and Brazil)• Partly endogenous to electoral system• Contrast with US primaries

2. Political Careers• At sub-national or national level?

3. Fragmentation and Party Discipline• High fragmentation and low discipline allow for greater

influence of governors

4. Existence of Regional Parties in the National Congress

• Particularly if they are pivotal (Argentina, Venezuela)

23

Governors and the PMP: Argentina• Historically Elected• Formally Federal• Territorial Chamber / High malapportionment• High expenditure decentralization• High vertical imbalance• Governors have significant nomination power• Governors transact congressional support for resources in a

high transaction cost environment• Lack of long-term commitments• Low quality of legislators partially attributable to governors’

influence and incentives• Governors sometimes constitute a veto point (Alfonsin)• Governors have had a credible threat of creating macro

problems

24

Governors and the PMP: Brazil• Historically elected• Concurrency between legislative and regional elections (1982,

1986, 1990)• Formally Federal• High malapportionment (in Senate)• High expenditure decentralization / moderate vertical imbalance• Some influence in nominations• High fragmentation of party system • Significant debate in literature over influence of governors• Governors’ influence has declined in last two presidential terms• Decline in capacity to create macro problems• Alston et al. (2004) argue that the partial presidential veto over

amendments and the internal structure of Congress provide presidents with the upper hand

25

Political Decentralization and the PMP

• The recent trend towards political decentralization has had a significant impact on the PMP of some countries

• Governors and mayors have generally become more relevant players, but there are still large regional differences

• Political decentralization influences the party structure and may have undesirable unintended consequences including a decline in governability

• The opening of multiple arenas of competition has transformed the incentives of politicians', reducing entry and party defection costs

• National parties have sometimes weakened and regional parties have become relevant players

• The decline in the discretionality of transfers from the center may weaken the power of national party leaders

26

Political Decentralization and the PMP: Colombia

• Mayors elected first in 1988. Governors in 1991.• Electoral system with multiple lists has promoted party

factionalism and regional party bosses• National constituency for the Senate has unintendedly

strengthened regional factions• Governors are not (still) key players in nomination

process• Increasing fragmentation in Congress partially

attributable to political decentralization• Constitutional fiscal decentralization• Congress structure helps to provide party discipline

27

Political Decentralization and the PMP: Mexico

• Formally federal, but highly centralized. Until the 1990’s elections were not competitive

• Electoral competition initiated at the sub-national level encouraged democratization

• When PRI dominated, governors were agents of party leaders. Very high party discipline. No-reelection rule gives power to party leaders

• Governors have gained leverage over the central government

• Opposition-ruled regions have gained leverage in fiscal negotiations

• Still high party discipline, but more influence of regional governments in PMP

28

Political Decentralization and the PMP: Venezuela

• Before 1989 party system centralized, concentrated, and disciplined. Nomination by national party leaders in single closed and blocked lists.

• Introduction of regional and local elections and mixed member electoral system, in a context of preference realignment, induced collapse of party system

• High fragmentation and volatility and a decline in discipline

• Role of governors in fiscal laws• Relevant role of regional parties• Re-centralization under Chavez