Democratic changes through Dominant Parties : the cases of Federal India , Malaysia , and Mexico

This article analyzes the dynamics of democratization in India, Malaysia and Mexico. Over the past decade, these countries, which have been governed by dominant parties, adopted or reinvigorated democratic practices. Unlike the predictions that postulated that democratization often occurs in a context of party competition, the article discusses existing mechanisms in dominant party systems that force the ruling party to encourage political transformations as a survival strategy. These three cases reveal that the growing political opposition at the local level was a determining factor behind the adoption of democratizing measures by dominant parties. Therefore, the article builds on the relationship between democratization and local opposition in dominant party systems.


Introduction
India, Malaysia and Mexico are federal countries that have been governed for most part of their contemporary histories by dominant parties.In India, the Indian National Congress Party (INC) has ruled uninterruptedly for 31 years.In Mexico, the Institutional Revolutionary Party (Partido Revolucionario Institucional -PRI) has ruled successively for 71 years.On the other hand, the United Malays' National Organisation (Pertubuhan Kebangsaan Melayu Bersatu -UMNO) has been in power in a row for 56 years in Malaysia.This article purports to explain how dominant parties in these countries can resort to electoral competition as a strategy of political survival 1 .This, in turn, as this article maintains, led to institutional changes that might favor democratization.While it has been acknowledged that dominant party systems behave differently than the other party systems, little has been said about why some dominant party system can be more inclined to democratization.
Understanding democratization as either a change from autocracy to a democratic regime as well as a process of democratic deepening, I argue that dominant party systems can be a source of democratic changes.This article looks on how the dominant party systems use their federal system and electoral institutions to embrace democratic dynamics.This argument is relevant as there is a growing concern over the relevance of parties as agents of change (GREENE, 2007).
Adding to the prediction that democratization occurs under multiparty systems with electoral competition (MAINWARING and TORCAL, 2006), this article holds that dominant party systems can democratize by embracing electoral competition as long as they can use territorial politics to their advantage.In this argument, the multi-level structure plays a critical role in the strategy of the dominant parties.Therefore, it is assumed here that multi-level organization of a country is intertwined with party competition.
Change in the party system has often been associated with patterns of electoral competition (LAVER, 1989;FRAZMANN, 2011).This focus on competition is biased towards explaining changes in political parties through adaptation in the party system.In fact, for a long time single party strategies, internal organization of parties and party positioning on issues have been neglected (HARMAL and JANDA, 1994).Consequently, dominant party systems have often been associated with immobility and stability (ARIAN and BARNS, 1974).Against this predicament, this article seeks to show that dominant parties can initiate and sustain changes in the party system and beyond.This argument will be developed in light of the factors such as the effective number of parties and the control of fiscal resources.These factors provide important hints on the conditions in which dominant parties might opt for an electoral path.Nowadays, several scholars subscribe to the argument that autocratic regimes embrace democratic practices to maintain a firm grip to power (e.g., MILLER, 2013;LEVITSKY and WAY, 2010).
While explaining change in dominant party strategies, I will be taking into consideration parties as central agents of change in an institutional context of multi-level governance, e.g.federalism, which shapes the choices of dominant parties.From a theoretical viewpoint, this is a relevant issue in the study of parties and their systems as the explanation combines the institutional and agency approaches.It is important to remind in this regard that the main paradigms employed to study party system fail to explain more systematically changes overtime and party as agencies of change (SPIESS, 2009).It is important to say that in this analysis exogenous factors such as the end of the Cold War and the demise of the developmental state model have not been included for different reasons.Although the end of the Cold War has created an opportunity for democratization, it has impacted differently the party systems worldwide.In effect, during the post-Cold War period not all countries followed a similar democratizing path (LEVITSKY and WAY, 2010).In the particular context of India, Mexico and Malaysia the Cold War appears to have had different and a minimal impact on their party systems (see VACCARO, 2007).In the particular case of India and Mexico the end of the Cold War reduced the stakes of domestic ideological conflict; however, there are signs that their dominant parties were already out of alignment with the Soviet Union despite an initial affinity.
In terms of changes in the state developmental model, India, Malaysia and Mexico embraced this developmental path, but over time these countries under different process of public sector reform have adopted market economy policies.Yet the process, pattern and degree of public sector reform vary across the countries making it difficult to consider change in the developmental model as an independent variable.
Given the different paces of democratization and the resiliency of dominant party regimes in the aftermath of the Cold War and of the weakening of the developmental state model, it is possible to endorse the idea that other factors have played a more prominent role in the democratization of dominant party systems.
The article will take on the task of explaining the nature of change in the party system in the three country cases by demonstrating that all the dominant parties responded to opposition challenges embracing the strategy of electoral competition coupled with the strategy of changing federal territorial cooperation.Subsequently, showing that these strategies varied in terms of scope and degree across the three countries, I will proceed by looking into the number of effective parties, strength of opposition and the control of federal fiscal resources in order to explain why the political outcome of dominant party survival has been sustained coalition formation in India, no-party alternation in Malaysia, and party alternation in power in Mexico2 .

Dominant Party System, Territorial Politics, and Democratization
The question of definition and characterization of a dominant party system has dominated a large body of literature in political studies.The lack of a consensual definition behind the characterization dominant party systems often leads to conceptual confusion.Since it is beyond the scope of this article to attempt to clarify conceptually party systems, I make a simple distinction between dominant party system and multi-party system adopting the idea that in the former concept a single party wins at least three consecutive elections the office of the national executive under a majority in parliament (SARTORI, 2005).In the later concept, the once dominant party fails either to win successive parliamentary majorities or to hold executive office.However, one-party dominance can in principle occur in both multi-party and dominant party system, as far as the dominant party outdistances itself electorally from the contenders.More specifically, using Pempel's (1990) criteria, a dominant party prevails electorally over the others by retaining a privileged bargaining position over government formation and decision-making over public policy.
Applying this conceptualization to India, Malaysia and Mexico, it is possible to place these three countries under the definition of dominant party system (See Figure 1).Each of these countries has established a dominant party system in different periods by holding minimally free elections, which can be seen in different degrees as -meaningful and manifestly unfair‖ elections (GREENE, 2007, p. 12).This applies to India from 1947 to 1978, to Malaysia from 1957 to the present, and in Mexico from 1929 to 2000.
In the Indian case, the INC started to establish its dominance in the first years following India's independence in 1947.The INC, which initially was a movement of national liberation, consolidated its hegemonic position with the implementation of its vision of a strong central government.The main opposition party the Bharatiya Janata Party (BJP), which started to make electoral advancements in the southern Indian states, eventually became nation-wide party forcing the INC to engage in strategic cooperation.
In Malaysia, a coalition of approximately fourteen parties, the Barisan Nasional (BS), being systematically dominated by the UMNO, controlled the political landscape for several years.The two main opposition parties, the Pan-Malaysian Islamic Party (Parti Islam SaMalaya -PAS) and the Democratic Action Party (Parti Tindakan Democratik -DAP), which have an electoral base that is respectively Islamists and ethnic Chinese, started to threatened the UMNO dominance within and without the BS coalition.
In Mexico, the protracted influence of the PRI on the political dynamics is telling by the long record of the party in holding elected positions: for approximately seventy years it won all national presidential elections and for approximately sixty years it succeeded in winning all state governments.The main opposition party, the National Action Party (Partido de la Acción Nacional -PAN), only started to threaten the PRI's hegemony in the 1990s by winning the gubernatorial elections.A third political party, the Party of Democratic Revolution (Partido de la Revolución Democratica -PRD), was the runner up in two presidential elections, (in 2006 and 2012) becoming a source of opposition in the Mexican political landscape.
In these countries, the transformation of the party system occurred under the growing pressure of the opposition.This forced the dominant parties to be engaged more in competitive elections.In India, the inability of the INC to secure the premiership in 1978 shifted the country from a dominant party system towards a multi-party system, in spite of the dominance of the INC.In a similar way, in Mexico the PRI's loss in 2000 shifted the country slowly to a multi-party system.After 12 years, the return of the PRI to the presidency without being able to guarantee a parliamentary majority confirmed the end of one-party dominance in Mexico.Malaysia is a special case amongst the three countries as the dominant UMNO has been able to maintain its dominance in Malaysian politics both within its political coalition and without.There are many sources of change in party systems (MAIR, 1989;2006).More recently several factors have been identified responsible for transforming party systems in advanced democracies.Among these factors, it is important to highlight mainly the decline of party membership (WHITELAY, 2010) and party's ideological orientation (THOMAS, 1975).In the dominant party system the focus on changes in the political landscape has been largely driven by the growth of opposition to one-party dominance.
In the three case studies being analyzed here, the dominant parties have lost progressively their dominant position (See Table 1).In India's INC and Mexico's PRI these parties failed in some elections to win the premiership and the presidency, respectively.In Malaysia, the UMNO has managed to win all general elections by leading a coalition, although its power inside the coalition has been diminishing over the years.
The turning point for India came in 1991 when the main opposition party to the INC, the BJP, won more than 20 percent of the popular vote in the general election, which was accompanied by a decline in votes for the INC.In Malaysia, the UMNO dominance started to erode in the early 2000s leading to the resignation of prime-minister Mahathir in 2003, after ruling the country for 22 years.In 2004 the UMNO obtained approximately 35 percent of the popular votes in the general election, a sharp fall from 48 percent of the votes received in the 1999 election.In Mexico, the presidential elections are representative of the slow downfall of the PRI.In the 2000 election, the PAN won the presidential race and it also obtained more seats than the PRI in the lower house of the parliament, putting 60 years of PRI's hegemony to an end.The difference between parliamentary and presidential systems in terms of electoral process rests on how the head of the executive is chosen.Although institutionally different in both systems, voters in a parliamentary system, just like in a presidential one, vote strategically to elect the head of the executive branch of government (SHUGART and CAREY, 1992).In terms of voting behavior this means that the difference between the two systems might not be as acute as often claimed.What might play a difference in each system is the room for maneuver through party coalitions, which is expected to be wider in the parliamentary system.Nonetheless, in a dominant party system this possibility is more limited.That said, in a dominant party system the matter of concern should be the influence that parliamentarism and presidentialism play on the way oppositions advance electorally.
Differently from a presidential system, as Mainwaring and Scully (1995) suggest, in a parliamentary system opposition parties have institutional incentives for the formation of coalitions, and, therefore, higher their possibility to be part of a governing coalition or simply be part of a stronger opposition.As such, it is expected that it is easier for the opposition in India and Malaysia to win elections than in Mexico.This reasoning could explain the tardiness and difficulty of the Mexican opposition to make inroads into the political landscape.However, it does not fully explain the long-term ability of the opposition parties to sustain themselves in power and/or exercise strong opposition, as the dominant party can eventually resort to the same coalition tactic to offset the opposition strength.That said, in this article the systems of government will not be the focus of our analysis.
As this article explores the implications of territorial politics in the change and adaptation of three dominant party systems, I now turn to the federal traits of the three country-cases.This exploration is based on the observation that although the dominant parties chose electoral competition as a strategy of political survival (e.g., ARIAN and BARNES, 1974), they varied in terms of their control over territorial politics.
In India, electoral competition was further developed as part of the strategy of the central government in the face of growing pressures from below for change (ARORA, 1992).Evidence of the importance of territorial politics in the opposition of the central government is uncontroversial.India's history with electoral competition shows that political parties started to build coalitions based on policies of state governments.Also electoral competition has been fierce in the states (CHHIBBER and KOLLMAN, 2004).In effect, the decline of the INC, which can be associated with the development of a multiparty system at the subnational level, begins with the electoral advancement of these parties in the 1980s in the state legislatures.State legislatures have become a place for the emergence of traditional regional parties (SÁEZ, 2002).Since the 1989 general election there has been a growing representation of regional parties in the national parliament's House of the People, the Lok Sabha.As a result, since 1989 four national governments have been formed with the help of a regional party.
In Malaysia, the dominance of the UMNO has been dented specially since the 1999 election, when its BS coalition won approximately 56 percent of the vote, a 10 percent fall from the previous general election.The strengthening of the opposition parties in the federative states helped the PAS to retain power in Kelantan and Terengganu.The PAS's victory in Terngganu, which is the homeland of the ethnic Malays and stronghold of the UMNO, was a major defeat for the UMNO.The other opposition party, the DAP, which mobilizes an important number of ethnic Chinese votes, did not perform as well as in the previous elections but as a whole the opposition coalition made inroads in the states and national legislatures.
In Mexico, the opposition party started to become competitive in the electoral arena in the municipalities in the 1980s when the PAN won several municipal elections in the northern Mexican states.However, the PAN only started to threaten the hegemony of the PRI by winning gubernatorial elections.In 1989 for the first time in approximately 50 years an opposition party was governing a federative state out of the existing 32 states.By 1994 approximately 9 percent of the Mexican states were governed by the opposition.In 2000 when the PAN became the winner in the presidential election, over 13 states were governed by either PAN or PRD-led coalitions.
The electoral inroads of the opposition parties in these dominant party systems through sub-national competition show that in a federal context opposition parties can exercise a transformative influence in the political dynamics in the country.The central component of party systems in multi-level context is the linkage between central and sub-national party competition.This linkage creates different possible levels of congruence between parties (THORLAKSON, 2006).Evidence indicates that in federations more decentralization increases party incongruence (THORLAKSON, 2002).This argument has important implications to the way dominant parties respond to opposition and to the manner the strategy of dominant party's survival is carried out.
To reiterate, the hypotheses advanced in this article postulate that in a multi-level context dominant parties respond to opposition pressures using both electoral and territorial mechanisms for political survival.This argument finds resonance in the cases of India, Malaysia and Mexico.In all these countries the dominant party systems have become less influential regionally.
Although in the three countries the influence of the dominant party has been progressively eroded at the sub-national levels, more competitive subnational elections had a different impact in terms of the ascendancy of the dominant party at the central level.
The degree of democratization varied already between the three cases when the dominant parties became more strategic in terms of holding competitive elections.Indiaan already established democracyexperienced important transformation especially in the 1980s and 1990s by expanding elections to the local level.Mexico experienced important changes in terms of electoral competition, moving from restrictive electoral process to a fully competitive one.In Malaysia, which has often been regarded as a pseudo-democracy (CASE, 2004), gerrymandering still exists but elections take place under less restricted civil liberties.
In the three cases, party competition was the response of the central government to growing pressure for changes.In other words, the dominant party has responded with decentralization and electoral competition.Decentralization in India entailed more cooperation with sub-national governments, while in Malaysia and in Mexico it meant non-cooperation.This article holds that this dynamics is associated with some electoral strategies.On the one hand, decentralized cooperation goes together with cooptation, and on the other hand, non-cooperation denoted limited and hostile party competition.The first scenario is found in India, while the other in Malaysia and Mexico.
The territorial strategy of decentralization is followed by an electoral strategy through which the dominant party can either co-opt or compete with the opposition parties electorally.Cooptation will require the dominant party to engage the opposition party in coalition formation.Differently, the option of competition means that the dominant party will be standing for elections.However, it is important to say that if the dominant party decides to run for election they believe that they will win.This is consistent with the observation that autocratic systems hold more elections to legitimize themselves as democratic (SCHEDLER, 2002;LEVITSKY and WAY, 2010).
The interplay between a territorial and electoral strategy leads to variation in the way dominant parties will maintain themselves in power (See Table 2).This is similar to say that the different options chosen by dominant parties in responding to opposition pressure create room for the emergence of different degrees of democratizing dynamics.

The Political Landscape and the Strategy of Dominant Party Survical
This section will examine the overarching strategy of the dominant party in response to the advancement of the opposition.In so doing, the analysis emphasizes the connection between growing relevance of the opposition in the political landscape and changes in the territorial political terrain in the three countries.In addition to this general feature common to India, Malaysia and Mexico, this section will highlight the different responses between these countries.To recapitulate, whereas in India decentralization was the accepted territorial strategy to the waning power of the INC, in Malaysia's EMNO and in Mexico's PRI centralization became the main territorial strategy.

India
It is worth mentioning that the two underlining features of federalism in India has been the reluctance of the central government to transfer power and authority to local governments, and the concentration of fiscal powers in the hands of the central government.These features nonetheless have been accompanied by accommodation of linguistic and religious diversity under the attempt to achieve national integration (FRANKEL, 2005).In the long-run these central features were slowly replaced by inter-governmental cooperation.Changes in party dynamics are in the heart of this adaptation.
The strong accommodative centralization pursued by the central government in the aftermath of the independence meant that the INC would be the epicenter of decision making which manifested itself in a centralized developmental planning and the creation of a federal dynamics that used the party system to achieve cooperation with sub-national governments.After independence, the notion that the Indian developmental goals would be reached with national planning grew stronger among Indian political leaders.This basically entailed that public services and economic development would be concentrated at the center.This in turn enhanced the creation of a centripetal federal dynamic, wherein the interest of sub-national government was closely aligned with the interests of the central government.In this context the INC, which controlled the appointment of the state governors, monopolized the centerstate relations.The signs that the party has absorbed federal dynamics are plenty.The INC's influence on the Indian federal institutions was strong enough that the relationship between the center and the states were played within the institutions of the INC (RAO and SINGH, 2005).
In spite of the strong grip of the INC in the dynamics of the country, slow changes have been taking place over the past thirty years.These changes occurred with transformation of the political landscape.The two underlying elements of these institutional changes have been the advancement of regional parties followed by the decline of a dominant party system.Several analysts underscore that one of the most important changes since the Indian independence has been the transformation of the party system (e.g., DASGUPTA, 2001;KOTHARI, 2006;MAHAJAN, 2007;PRAKASH, 2010).As a consequence, politics at the state level, once an appendix of the center, gained relevance for the national Indian politics.Yadav and Palshikar (2006) call attention to this by suggesting that the regional parties have transformed the political landscape of the Indian states.
In this regard, it is possible to identify according to Kohli (2004), two different phases that reflect distinct political dynamics in India under the dominance of the INC.The first phase is the cooperative years of the Nehru government and the uncooperative years of the Indira Gandhi government.In these two different phases, the INC showed resilience in maintaining a firm grip.After 1977, the premiership of India has been occupied by different coalitions of party that injected a new pattern of doing politics in the country.This new pattern of political dynamics in India is often called the -politics of cohabitation‖ (MEHRA, 2010).These dynamics entail that coalition formation is accompanied by close political cooperation between the main political forces.
In the light of this type of dynamics, the hegemony of the INC should be interpreted with some caution.In the 1977 general elections, following the declaration of state emergency by Indira Gandhi, with the formation of what would be the main opposition party in India, the Bharatiya Janata Party (BJP), the INC started to rely on coalitions to form a government.In India, the advancement of the opposition in the state level required important transformations in the federal dynamics.For example, in the 1980s with the new political landscape in which the INC became less dominant, the central government agreed to establish in 1983 a parliamentary commission (the Sarkaria Commission), which would recommend restrain on the central government's dismissal of governors, the creation of a permanent inter-state council to debate federal issues and the empowerment of local self-governments.In another critical moment in the recent political history in India, the victory of the Janata Dal in 1989 general elections following allegations of corruption involving the leaders of the INC led to important changes in Indian federalism.After 1989 the governing coalition initiated reforms in the system of local government passing two constitutional amendments granting more autonomy to these governments and allowing the states to hold local elections.
In the past decades, as shown above, India has experienced important political changes.The causes of these changes are various; however the renewed territorial demands coming from the sub-national governments have been an important cause of transformation (ARORA, 1992).These sub-national demands required a response from the INC, in the form of gradual fiscal and administrative decentralization which has been accompanied by pre-electoral coalition formation between the main parties at all levels of government.

Malaysia
The augmentation of power and authority of central government vis-à-vis sub-national governments has become a distinguishing feature of Malaysian federalism since 1957, when Malaysia gained independence and promulgated its federal constitution.The UMNO epitomizes the history of a party that capitalizes on the ethnic cleavages to gain popular support from the ethnic Malays while using federal institutions to subdue local elites.
Behind this growing strength of the central government rests its concern to preserve the power stronghold of ethnic Malays.In the context of a dominant party system, the strategy of the central government to maintain hierarchy in the inter-governmental relations has been successful.The perpetuation of the dominant party in the center served well this strategy.Indeed, as Case (2007) proposes, power-sharing arrangements under Malaysian federalism have served to strengthen the political grip of the center.The central government has used the state bureaucracy to favor its Barisan coalition and used uncompetitive elections to legitimize it.More specifically, the political might of the central government was built at the expense of democratic practices such as the creation of competitive but manipulated elections.
This strategy of the central government became more pronounced during the premiership of Mahathir Mohamad from 1981 to 2003.Under Mahathir's government the personalization of power became part of the Malaysian political dynamics (SLATER, 2003;BROWNLEE, 2007).In this personalization, federal dynamics were dictated by the center.In the case of Malaysia it is most evident in the fiscal front.The concentration of fiscal powers at the center coupled with discretionary distribution of fiscal resources has held back the fiscal autonomy of state government in Malaysia while failing to reduce the fiscal gap in Malaysia.
These centralist trends started to change towards the end of the 1990s.The political stability and dominance of the dominant BS coalition was threatened in 1997 with a movement demanding more competitive elections that grew under the reformasi.This movement, which was led by the former Malaysian deputy primeminister, Anwar Ibrahim, was formed under the Barisan Alternatif multi-ethnic coalition with the purpose to challenge the ruling party in the national elections.Under this coalition, the main opposition parties were often able through electoral advancements to create public debate.
Several scholars claim that the reformasi should be interpreted with caution because the BS has maintained control of important state resources and its dominant political force (e.g., WELSH, 2004;CASE, 2004).Even though the more competitive elections have not changed fundamentally the political landscape, political elites in Malaysia started to reach out to local support.Similarly, Derichs (2004) argues that following the 1999 election the opposition coalition, the Barisan Alternatif that had run against the BS failed to produce profound changes due to fear of the ruling Malay elites to lose their hegemonic power.
It is clear that changes in the political landscape in Malaysia have occurred with the gradual growth of the opposition in the Malaysian national elections starting in 1995.However, it still remains to be seen how this political change is translated into a more robust inter-governmental relations and eventually into the increase in power and authority of sub-national elites in Malaysia.Under moderately competitive elections, UMNO has been leading a governing coalition of parties, which has been sustainable to a great extent through a centralized control of resources distributed on the basis of allegiance.

Mexico
From the onset of the promulgation of the 1917 Mexican constitution, the PRI has for the most part dominated electoral politics in Mexico.The overwhelming influence on the political dynamics has been exercised mainly through two mechanisms, namely the fusion between state and party and through the exercise of meta-constitutional powers by the Mexican president.
Given the initial scenario of centralized territorial dynamics, it was unanticipated that the transformation of the Mexican federalism would come from below.The Mexican experience with competitive elections starting in the 1990s supports the argument that the prime source of inertia behind changes in the Mexican federalism came from below.Scholars concur that federal dynamics in Mexico have changed with the introduction of electoral competition (e.g., KLESNER, 1998;HERNANDEZ-RODRIGUEZ, 2003) in a political system dominated by the PRI.While it is clear that the institutional changes from below have created unprecedented circumstances making it difficult to restore the might of the central government but its effect on the state governments' ability to exercise power under this new situation is yet to be discerned (CORNELIUS, 1999).
The reforms in electoral rules were critical for the advancement of opposition parties at the sub-national level of governments.In 1989 PAN won the gubernatorial elections in Baja California.Although the PAN won the mayoral elections in 1983 in two cities -Chihuahua and Ciudad Juarezit lost the subsequent municipal elections.The PAN's municipal government helped the party to win a reputation for managerial efficiency (KLESNER, 1998).Moreover, the electoral advancement of the PAN had an impact within the PRI.According to Hernandez-Rodriguez (2003), competition had an influence on the governors of the PRI who started to challenge the president's power.This, in turn forced the central cadre of the PRI to strengthen the role of the governors within the party and forced the PRI to select candidates for local elections based on their credentials and connection with their locality (HERNANDEZ-RODRIGUEZ, 2003).Also, these changes favored the local elites inside the PRI, which started to strengthen independent power basis.As a result, governors in Mexico even within the PRI became defiant against the centralized Mexican president.
It is clear that from the democratization of local politics in the early 1980s to the national political reforms, several years have passed before real transformation occurred in the dormant Mexican federalism.In spite of the slow pace of changes, the consequences of the electoral competition was far reaching: it contributed to the institutionalization of political institutions e.g.legislative branch (BEER, 2001), it helped to reinforce the structure of the parties at the local level (KLESNER, 2003), it created incentives for political oversight (SOLT, 2004) and it reinforced the power of state governors in some enclaves.
It is widely recognized that the inter-governmental hierarchy that held state and local governments in Mexico subordinate to the center ceased to exert influence in the 1990s.The first move towards reforms came in 1983 with a constitutional amendment to Article 115 granting responsibilities to the Mexican municipalities.In addition to granting responsibilities, the amendment reformed the institutional aspects of municipal structure of government with the establishment of the town council, and the harmonization of the procedures of electing municipal authorities (FOX and ARANDA, 1996).These reforms were limited, as they have not contributed to greater autonomy of municipalities in Mexico.The increase in responsibilities was not matched by an increase in resources of the municipalities, particularly in the rural areas (FOX and ARANDA, 1996).However, on the institutional front, this amendment increased municipal autonomy, as it made more difficult for state governments to remove mayors.It is worth pointing that these reforms have intended to favor the municipalities and not necessarily the states.
Fiscal decentralization became an important demand of the PAN in the 1990s as the party was making important inroads in the gubernatorial elections.In 1995 the central government approved transitory measures on the new fiscal code.This was the result of the PAN's growing pressure to precipitate the transfer of the central government to the sub-national governments.Two important goals were reached: increase in the fiscal transfers (participaciones) and the creation of new revenue sources to states and municipalities.Another important achievement for sub-national governments has been the reduction of central government's discretion on the transfer of resources.Yet, there were reversals on other areas: the transfers to the states were gradually reduced over the years (FALLETI, 2010).
Based on the recent history of electoral changes in Mexico, which has been characterized by higher electoral contestation in a new party system with some signs of multiparty system, it is possible to infer that this was the result of a gradual erosion of a PRI that attempted to use territorial centralization of power to maintain a cooperative federalism based on old patronage practices.In other words, limited party competition together with a centralized cooperation inevitably created an opportunity for the PAN to end in 2000 seven decades of PRI hegemonic rule.

Conditions of change under dominant parties
The purpose of this section is to discuss the conditions under which the different types of outcomes emerged as a consequence of the survival strategy pursued by the dominant parties in India, Malaysia and Mexico.In the Indian case pre-electoral coalition formation with alternating party leadership in the premier office is the main result of the INC's strategies.Differently, in Malaysia the continued electoral dominance of the UMNO has been maintained through a pre-electoral coalition that secured its rule.In Mexico the end result of the territorial and electoral strategies embraced by the PRI has been power alternation.
Underlying these strategies are institutional conditionseffective number of parties and discretional distributions of fiscal resourcesthat can explain the variation in the observed outcomes (See Table 3).In this article it is emphasized that the higher the effective number of parties, the higher the possibility of opposition faced by the dominant party insofar as the opposition parties are able to coordinate an opposition strategy between them3 .In India the high number of parties and high level of position strength together with the lower discretion over distribution of resources inevitably led to more electoral competition requiring the development of cooperative tactics of engagement by the INC with the emergence of a multi party system.With a more moderate number of effective parties and opposition strength, as well as the moderate discretion over distribution of resources, the PRI understood that its historical competitive electoral advantage could be easily maintained.However, with the emergence of the PRD, elections became more competitive culminating in power alternation.In Malaysia the low number of effective number of parties and opposition strength coupled with high fiscal discretion in the distribution of fiscal resources, allowed the UMNO to gerrymander elections to its favor.In order to shed more light into these differences across countries, each of these conditions will be treated with detail below.

Effective number of parties
Considering that the number of parties might have different effects in the party system (e.g., high fragmentation), it is important to understand the distribution of power between them and it is worth looking at the relevance of parties in the dominant party system of India, Malaysia, and Mexico (See Figure 2).Among the three cases, India stands as the country with the highest number of effective parties totaling 4.7 (considering the 1989-2009 period).Mexico (from 1997 to 2009) and Malaysia (from 1957 to 2005) have respectively 2.55 and 1.54 parties4 .In other words, these figures indicate that the party fragmentation is higher in India, being followed by Mexico and Malaysia.
In the line of this article it can be said that higher party fragmentation in India implies that the INC faced more party competition in elections than in Mexico and Malaysia.In response to growing opposition the INC in India had to attract votes from a fragmented electorate and in order to survive under growing factionalism, coalition formation was the best available survival strategy.Some consider this moment as the beginning of the fractionalization of the INC that eventually led to a creation of a multiparty system (e.g., SÁEZ, 2002).
Under a less fragmented party system, the dominant parties in Malaysia stand for elections with a belligerent strategy and resort to manipulation practices (CASE, 2004).Yet, Mexico under a more fragmented party system than Malaysia, the main opposition partythe PANmanaged to capitalize on the malpractices of the initially dominant PRI and eventually attract more votes.However, the advancement of the PAN, which was often seen as a docile opposition coinciding with the emergence of the PRD, increased the contestation of the national elections.Eventually the PAN was able to form coalition with the PRD.In Malaysia the low fractionalization in the opposition allowed the UMNO to keep an upper hand in the electoral competition.
The effective number of parties in a political system to a great extent has to do with the electoral strength of the opposition.In connection with this observation, Greene's (2007) explanation behind the failure of dominant party systems to hold on to power under electoral competition goes through the opposition strength.Greene holds that three main factors that contribute to Source: Reilly (2007) and Golosov (2011).sustained dominance of one party, namely, opposition parties only appeal to a small segment in society, multiple opposition parties do not coordinate among themselves and the dominant party system has a resource advantage over the opposition parties (e.g.voter bias).Focusing on the first two factors, it is possible to suggest that in India the opposition has been strong, in Malaysia weak and in Mexico moderately strong.
In India the electoral inroads made by regional parties at the sub-national level over the past decades added complexity to the Indian party system.The widening of the political opposition has forced the INC to resort to pre-electoral coalition formation at the different levels of government.Under this arrangement, the INC has relied on the regional parties to articulate locally its interests.This dependency is precisely the reason where the strength of the regional opposition parties lies.The handicap of regional opposition parties is that they cannot translate their influence to the national level, where only the BJP is able to exercise opposition.However, the BJP just like the INC, is reliant on coalition formation with the regional parties.In competing with a more pragmatic INC for coalitionable parties and for votes, the internally divided BJP is likely to find fewer parties with which to build coalition due to its narrow pro-Hindu religious ideological orientation.In this political context, the opposition strength rests on the possibility of then building coalition among them.As such, fragmentation in a multi-level context has created a thin balance on the one hand between the national influence of the INC and the regional coalitional parties and on the other, between the INC and the BJP at the national level.
In Mexico, until the late 1990s the ousting of the PRI seemed to be an impossible mission to accomplish.One of the reasons behind this belief was the difficulty of the opposition parties to appeal to the wider public and for them to join forces against the PRI.As a result, the PAN focused in winning elections subnationally.In effect, the states where the PAN won the first gubernatorial elections were precisely the states that the party managed to establish an electoral stronghold that sustains to this date (KLESNER, 2012).While this confirms the difficulty of opposition parties to appeal to a larger public, another opposition party, PRD paved the way to the PAN's first historical presidential victory in 2000.The creation of the PRD in 1989 as a new left party attracted a wider number of electorates.As the PRI was unable to respond to this -new left appeal‖, the PAN took advantage of this polarization and as a more consolidated party was able to overcome the incumbent vote bias.
In Malaysia in two different general elections, in 1995 and in 1999, the opposition gained strength by augmenting the percentage of votes received.Yet the main opposition parties are unable to sustain this support.This is a recurring pattern in the Malaysian general election.This oscillation in electoral support happens even when opposition parties manage to overcome divisions and compete under a pre-election coalition.The failure of gaining continued voter support lay on the difficulty of forming a multi-ethnic alliance (CROUCH, 1996).There are signs that for the opposition to compete electorally with an increasingly catchall UMNO they have to overcome ethnic divisions and propose a convincing reform agenda (WEISS, 2006).

Discretional distribution federal fiscal resources
Decentralized revenue and expenditure not always is a sign of local government autonomy.However, decentralization of fiscal resources coupled with institutionalized mechanisms of resource distribution minimizes the ability to distribute resources in states governed by the dominant party.In other words, the existence of institutional rules of fiscal distribution between levels of government constrains the ability of the central government to distribute fiscal resources discretionally, while limiting the possibility that these resources will be used politically.
In the three cases, the central government exercised certain degrees of discretion of the distribution of federal fiscal resources.The difference between the cases rests on the ability of the central government to use effectively the distribution of resources to ensure political gains.In the Indian case the lack of transparent rules of fiscal transfers has been giving leeway for the central government to use resources according to political motivation but the ability of the central government to use it extensively is limited.In Mexico, the central government exercised discretion over the distribution of resources to the state governments but so did the governors in the case of municipal transfers.In practice, the discretional powers of the Mexican governors limited the ability of the central government to have any direct political gains in the distribution of municipal resources.Certainly, Malaysia is the country that epitomizes the pivotal role of discretion in the distribution of fiscal resources to obtain political gains.
In India, fiscal federalism has evolved towards a concentration of the financial resources at the center and towards an increasing reliance of the subnational governments on financial transfers from the center.The inability of the Indian states to finance their current expenditure with their own resources has become more acute over time (BAGCHI, 2004;RAO and SINGH, 2005).Increasingly with the growing reliance of sub-national governments on transfers, the central government becomes more influential in determining sub-national expenditure (PRASAD, 2004).Moreover, the central government has increasingly made inroads in constitutional assignments that were concurrent and exclusive to the states (KRISHNASWAMY et. al., 1992;BAGCHI, 2004).The vertical fiscal imbalances are even more acute at the local level.Local governments have very small sources of revenue and cannot enforce certain taxes.This situation is more severe in the rural areas where rural governments have in comparative terms even smaller revenues.The little leverage of the sub-national governments has done little to alter this situation.Central transfers to sub-national government have remained politically motivated (RAO and SINGH, 2005).Recent data reviews that in India, higher political competition at the state-level measured by the vote margin among competing parties, will lead to higher government expenditure (GHOSH, 2010).
Fiscal federalism in Malaysia is a reflection of the centralism that dominated intergovernmental relations in the country.The Malaysian central government has acquired important taxation powers leaving state governments with a small tax base.It is interesting to note that the states that obtained a -privileged‖ status in the Malaysian fiscal system, e.g.Sabah and Sarawak, have experienced difficulties in maintaining their resource base.The centralism of the intergovernmental fiscal arrangements in Malaysia has jeopardized the principle of equitable distribution in the country to the extent that the discretion with which intergovernmental fiscal transfers are made creates disparities among the states (WILSON, 1996).Although the states in Malaysia on the whole are able to finance most of their expenditures with their own revenues, the lack of an equitable distribution of resources across the federated states is obvious.The case of Malaysia demonstrates that the central government has displayed a pattern of intergovernmental distribution of resources characterized by larger resources given to states that are politically aligned to the central government.As fiscal transfers under the State Reserve Fund are given to the states based on the recommendations of the National Finance Council, where the Malaysian primeminister has important saying, the central government has an upper hand in the final transfer decisions.And as a result, states that were not politically aligned with the center, e.g.Kelantam and Terengganu, apart from receiving lower fiscal transfers, saw the reduction of petroleum royalties and the imposition of export taxes on certain products (JOMO and HUI, 2003).
Mexico is a fiscally centralized country that has relied on the party hegemony to distribute financial resources.In Mexico, there is a close relationship between the growing hegemony of the PRI and the lack of fiscal resources at the sub-national level of government.With the consolidation of the PRI as a hegemonic party, the central state increased its tax base while the states and municipalities relied on central government transfers.Since the Mexican Revolution, despite the variation in the revenues among levels of government, one trend can be identified: the decline of tax revenues of the states and municipalities and an increase of tax collection of the central government (DIAZ-CAYEROS, 2006).In the post-revolutionary fiscal system in Mexico the states managed to preserve some fiscal authority while the central government centralized revenue collection and distribution.This centralization made sub-national governments particularly reliant on central government transfers.In the context of Mexican fiscal federalism, revenue sharing was subject to the discretion of the central government.In Mexico the statutory transfers, which should follow a predetermined formula, do not correspond to the actual transfers from the central government to state governments (DIAZ-CAYEROS, 2006).This centralization is replicated at the sub-national level as states in similar fashion to the central government, distributed resources among municipalities with high discretion.

Concluding Remarks and Extensions
A central question raised in this article has been to understand why dominant party systems decide to resort to democratic mechanisms to survive politically.This question was addressed using the intuitional and the agency approach.By combining these two approaches, the article has shown that federalism and the electoral institution have determined the choices of the dominant party to increment their role as a relevant political actor.
On this note, in India the INC responded with tactical cooperation with the opposition at the national and sub-national levels of government.In Mexico the PRI slowly reformed the electoral institutions in the country while trying to control the federal dynamics, a tactic that showed to be contributing in the longrun to power alternation.In Malaysia, the transformations brought about by the Barisian coalition have created a momentum for change, although the signs of democratic changes are yet to be seen.
The role of parties as agents of transformation in India, Mexico and Malaysia is uncontroversial.A strong indication that parties are still very relevant despite the demise of the dominant parties in regional elections and the fall of their public support is the transformative impact that parties have in the party system.And as far as democracy is concerned, these cases also show against all the odds that dominant parties can foster democratic changes.This argument should be read with caution as democratic means are often used by dominant parties echoing the claim that Gandhi and Przeworski (2007) put forward to show how democracy can be used by autocrats as a mechanisms for survival.This is certainly the case of Malaysia, where electoral dynamics, together with territorial dynamics, have reinforced the dominant party's power.However, this claim does not hold true to the case of Mexico.Differently, the Mexican case highlights the limited powers of electoral manipulation under autocracy.In effect, the pattern of power alternation of the two main political parties shows signs of democratic consolidation in Mexico.In a similar vein, India has proceeded with democratic deepening through a strategic choice to counter the advancement of opposition parties at the state level and at the national level.

Figure 1 -
Figure 1 -India, Malaysia and Mexico under party-systems typology

Figure 2 -
Figure 2 -Effective Number of Parties in India, Malaysia, and Mexico

Table 1 -
Dominant and Opposition Parties in Parliamentary elections (% of vote)

Table 2 -
Interplay between Territorial Strategy and Electoral Strategy Source: Own elaboration.

Table 3 -
Conditions producing change Source: Own elaboration.