The Indian party system has undergone a dramatic transformation in the last decade. This essay is a preliminary attempt at making sense of this transformation. It delineates the major trends in the party system in the period 1989-99, empirically details them, and relates these developments to the theoretical literature on the determinants of party systems. Finally, it outlines some of the consequences of party system change for policy outcomes.

A quick outline of major trends in the party system in the last decade must include the following developments:

1. The single-most important development in this period has been the decline of the Congress, for long the inevitable nodal reference point of both political practice and theoretical reflection on Indian politics. It is true that the Congress has, in purely electoral terms, shown that it is capable of rebounding successfully – witness the results of a number of Assembly elections. But it has been definitively dislodged from the position of the centre around which all political calculations must of necessity revolve.

2. The trend towards federalisation of the party system,¹ a trend already strong in the 1980s, has been substantially deepened. It made its presence felt at the Centre for the first time with the National Front government in 1989, and has become a trend that shows every sign of enduring. Consequently, despite coalition and/or minority governments and related cabinet instability at the Centre, the state level has seen the maturing of bipolar party systems in a majority of states.

3. This decade saw a sharp rise in political mobilization on the basis of social cleavages based on ascriptive identities, in particular of religion and caste.

4. Central to parliamentary government is the process of government formation and the constitution of the cabinet. In this decade, this process resulted in, variously, majority coalition, minority coalition and single-party minority governments. In addition, by rational anticipation of the verdict of a hung Parliament, a number of parties have veered towards what can well be labelled an ‘alliance culture’. Alliances have become part of the accepted rules of the game, rather than something to be resorted to in exceptional moments.

We now detail selected empirical facts about the party system in the 1989-99 period to examine whether these patterns can be put into an analytically coherent framework.² Since explanation and description can hardly be separated, it is best to make clear what theoretical issues the description is meant to foreground.

The issues are two. The first is the influence of institutional variables in shaping the evolution of the party system. There are two institutions that need to be considered: the electoral system and the federal character of the state. They need to be considered jointly because, in ways to be spelled out, it is not each in isolation but both working in conjunction that produces their effect: a party system which produces coalition and/or minority governments at the national level, but at the state level yields two-party or more generally bipolar systems. The second issue relates to the role of social cleavages, or more precisely how these cleavages are ‘particized’ through the dynamic of party competition. The trajectory of the party system in India, I argue, can best be made sense of in terms of these two factors. But for now it is necessary to move to the empirical details.

 

 

A particularly interesting trend in this decade is the emergence of a bifurcated or two level pattern in the party system in which the state pattern is significantly different from the national pattern. The following provides evidence for this claim.

At the national level, the first five general elections, 1952 to 1971, gave rise to a one-party dominant system in which Congress received over 40% of the vote. With the important exception of the 1967 elections, the pattern in the states was not dissimilar to that at the Centre, replicating the pattern of Congress dominance. By contrast, the period from 1977 to the 1998 elections shows a pattern of growing competitiveness in the party system (again, with one exception: the 1984 election, held in the shadow of Indira Gandhi’s assassination, which resulted in a large victory margin for the Congress). The vote margin between the first and second parties tended to decrease.

In particular, general elections from 1989 onwards reveal an unambiguous trend of Congress decline, accompanied by the rise of two other formations: the BJP and its allies and a coalition centred around the Janata Dal. In 1989, a minority coalition led by the National Front assumed office. Its main component was the Janata Dal, with outside support from the BJP and the Left Front. The 1991 general elections brought to power a minority Congress government, which during the course of its term transformed itself into majority status by carefully orchestrated defections.

In 1996, a 13-party coalition, christened the United Front, formed the government with external support from both the Congress and the Left Front. Finally, the 1998 election resulted in a coalition government led by the BJP, with a number of the partners being strongly anchored regional parties. Government formation at the national level thus revealed a definite move away from one-party majority rule; multiparty coalition and/or minority government seems here to stay.

 

 

However, the state level tells a significantly different story. The pattern revealed by state assembly election results is as follows:

(i) In a number of states, the decline of the Congress was caused by the growth of one other party, resulting in a two-party system in which both Congress and the other party have been able to form governments on their own. This is the case in M.P., Rajasthan, H.P., Gujarat, A.P., Orissa and Assam.

(ii) Yet another set of states reveal a pattern that is not quite a two-party system, but analogous to it, in the sense that there exist two poles in the party system, with one or more parties clustered at each pole. This bipolar pattern³ occurs in Maharashtra, Haryana, Tamil Nadu, Kerala, West Bengal and Tripura. The last named three states have evolved a stable bipolar system comprising of a Left Front coalition opposed by a Congress or Congress-led coalition. Some change in this pattern has occurred in Bengal by the breaking away of Mamata Banerjee’s Trinamul Congress from the Congress and her subsequent aligning at the national level with the BJP; but the consequent weakening of the Congress has left the party system in the state essentially unaltered, as a bipolar structure.

(iii) A multiparty system without a clear bipolar party system exists in U.P., Karnataka and Bihar. Recent developments in Karnataka in the run-up to the forthcoming 1999 general elections, however, augur a serious weakening of the Janata Dal, and the two-party pattern seems likely to be established there as well.

 

 

The upshot of the above empirical patterns is that till the late ’60s/early ’70s , there existed a one-party dominant system in which a dominant party, the Congress, overwhelmed a fragmented opposition at the state level in both the Lok Sabha and Assembly elections. The only exception to this in a limited number of states was the 1967 election. This pattern has now disappeared in all states, for both Lok Sabha and Assembly elections. It is reasonable to conclude that, leaving aside some important exceptions, the state level party system has evolved towards either a straightforward two-party system, or a multiparty system which is bipolar in terms of the pattern of party competition.

At the national level there is no visible tendency towards a two-party system. Rather, what has emerged is a multiparty system with three loose alliance structures, and a number of small parties that stand independent of these. This was only to be expected because the state level is characterised by a multiplicity of bipolarities. In one set of states it is Congress vs. BJP and allies; in another set it is Congress vs. Left parties; while a Congress vs. regional party pattern prevails in Punjab, Haryana, A.P., Assam and the Northeastern states (except Tripura, where it is Congress vs. Left).

 

 

The literature on the determinants of party systems offers a number of contending theories which help to make sense of the above patterns. They may be classified under two broad heads: theories that relate party systems to institutions, and theories that relate them to social cleavages. Among the former, most attention has been devoted to the effects of one particular institution, viz. the electoral laws in place in the country concerned. However, a second institutional feature, the federal structure of the state, while it has received much less analytical attention, is more significant in the Indian context.

Duverger’s law asserts a strong connection between the type of electoral laws and the party system prevalent in a polity.⁴ The claim is that single-member-district simple plurality (SMSP) electoral systems favour two-party systems. Conversely, proportional representation (PR) systems encourage multi-partyism.

This is at best only an inductive generalization in need of an explanation which would anchor it in the purposive behaviour of voters and candidates. Subsequent research has pointed to two such explanatory mechanisms whose joint operation is thought to underpin this law: strategic entry by party/candidates, which tends to reduce the number of candidates who actually enter to two (given the expectation of failure of third parties); and strategic voting by the electorate, which reduces the effective number of candidates to two, in case more than two have entered the race.

Vis-a-vis India, Duverger’s law raises questions as to the reason for the existence of a multiparty system at the national level despite a simple plurality electoral system. The key lies in recognizing that the law essentially operates only at the level of the electoral district.⁵ Starting out from this level, the overall number of parties in a polity will be decided by the linkages effected between three levels: district, state and national. Under some conditions, linkages will be successfully effected across all three levels, thus leading to a national two-party system. Other circumstances may produce a different equilibrium where aggregation stops short of the national level. In the Indian case, the question can be reformulated as: what is the character of the linkage process that results in a bifurcated party system? This is where federalism, the second institutional feature, needs to be introduced to complete the explanation.

 

 

Intuitively, it is not hard to see that the ‘political opportunity structure’⁶ offered by the federal nature of the Indian state provides strong incentives for ambitious politicians to set up new parties at the province level of aggregation, or to establish powerful regional branches of national parties, in both case leading to an effective federalization of the party system. In fact, there exist both office-seeking and policy-seeking attractions for politicians at the state level. Most policy in India, including specifically the key domains of agricultural policy and job reservations policy, is both constitutionally assigned to the state assemblies, and/or their effective implementation is controlled at the state level (e.g. land reforms).

As for office-seeking motives, the chances of capturing positions at the Centre are far more remote compared to the opportunity of state-level office. While empirical evidence on this has yet to be systematically collected, it is reasonable to hypothesize that a regional politician would find it far easier to cobble together a support coalition with the promise, to be plausibly encashed in the short-to-medium run, of the distribution as spoils of the offices of the many parastatal enterprises and organizations under state control, appointment powers to which comes with occupying governmental power.

 

 

As for the demand side of the political equation, not much is required in terms of ideological originality on the part of these leaders, insofar as they can tap an available vein of regional developmental cum cultural sentiment in the electorate. Diverse instances of this logic at work include the Telugu Desam Party (TDP), Asom Gana Parishad (AGP), Haryana Vikas Party (HVP) – parties which by their very names announce their regional bases. What is equally interesting, however, is the effective regionalization of parties with an avowedly national perspective, viz. the Samajwadi Party (confined to U.P. and to a small extent Maharashtra), the Rashtriya Janata Dal (Bihar), and the Bahujan Samaj Party (U.P.). The CPI(m) has become regionalised in the states of Kerala, Tripura and West Bengal while suffering an almost total elimination from the Hindi heartland, even in constituencies where it once had a strong electoral record, such as Kanpur.

The above account on institutional influences is incomplete because little has been said regarding the substantive reasons for the discontent which prompted both voters and ambitious politicians to abandon the Congress and move towards other parties, either strengthening pre-existing ones or floating new parties altogether. This is where social cleavages enter. It has been widely commented that a number of cleavages have been ‘particized’ in the 1990s. The two most prominent forms this has taken are the growth of the Hindutva movement and caste based mobilization in at least two distinct forms (OBC-based and SC/ST-based).

 

 

One finds two distinct types of theories on offer in the literature on social cleavages. The first originated in a very influential volume edited by Lipset and Rokkan.⁷ The links between social cleavages and party systems were seen as emerging from the national and industrial revolutions that characterized Europe at the beginning of the 20th century. This argument relied on a form of sociological determinism whereby parties are influenced by social change, but not the other way around. What the Lipset and Rokkan conceptualization leaves out of consideration is the role of politicians as entrepreneurs, who actively exploit the lines of division in a society.

Precisely this aspect is the strength of an alternative account of cleavages offered by E.E. Schattschneider in The Semisovereign People. This book rejects the view that, because there are a variety of cleavages within any society, parties must necessarily arise to reflect those cleavages, and that the resulting party system will be simply a reflection of these social divisions. Instead, Schattschneider argues that among the many divisions within a society only some will become dominant and form the focus of division between parties. Other cleavages become overridden and largely irrelevant to political competition. Party-political entrepreneurs try to define politicize cleavages in such a way as to build up a coalition that will best give them access to power.

 

 

What resonates about this perspective vis-à-vis Indian politics in the post-Congress dominance years is the denial of any natural order of conflicts in a society which must necessarily be reflected in a party system. Rather, it is ambitious political leaders who impose that order themselves through their efforts to develop cleavages that will advantage them.⁸ To quote Schattschneider: to understand the nature of party conflict it is necessary to consider the function of the cleavages exploited by the parties in their struggle for supremacy. Since the development of cleavages is a prime instrument of power, the party which is able to make its definition of the issues prevail is likely to take over the government.⁹

This seems a prescient description of, for instance, the so-called Mandal vs. Mandir struggle that erupted in 1990 between the votaries of reservation for OBCs and the BJP, and which has been a defining feature of this decade of politics in diverse ways ever since.¹⁰

A normative worry about the role of social cleavages in politics that deserves to be addressed is whether such politics may lead to a hardening of ascriptive identities and generally a decline in the quality of civil society. It is true that a threat to democracy arises if some group is converted into a permanent majority which cannot conceivably be overturned; this undermines democracy as it offers no option for excluded groups but to seek recourse to extremist measures. But this does not appear to be the direction in which Indian politics is moving. This can be exemplified by the situation in U.P. which has witnessed the most intensely divisive forms of cleavage politics in this decade.

 

 

In U.P. none of the three major contenders – the bjp, the Samajwadi Party (sp), and the Bahujan Samaj party (BSP), by itself has the strength to build a sustained majority. The logic of numbers inherent in electoral democracy exerts a centripetal effect whereby each party is forced to appeal to groups not previously part of its core constituency. The net result is that no group is permanently excluded from partisan support coalitions. Backward Castes are perhaps in the best situation, being appealed to by all three parties. Scheduled Castes, who form the core base of the BSP, find themselves in addition being wooed by the bjp and whatever is left of the Congress. Upper caste groups, who in general have moved over to the bjp, are also targeted by the sp, and (to a much lesser extent) the BSP. Mobilization of Muslims is attempted by both SP and BSP.

A final point on the pattern of social cleavages in India is the relative weakness of cleavage development on explicitly economic lines. The oddity of this in one of the most economically unequal countries in the world deserves attention. A major reason is the absence of permanent consolidated organizations based on the organization of productive life. It is not as if such organizations were never mooted as a possibility in Indian party-political history, even leaving aside the Communist parties. The historian D.A. Low¹¹ provides a careful reconstruction of a remarkable moment in history, in 1936-37, when the Congress party was faced with a proposal initiated by its left-wing to grant functional representation to workers and peasants as economic categories within the body of the Congress. The proposal was defeated. It is interesting to speculate whether the later pattern of cleavage mobilizations in post-1947 India might have been infused with more explicit economic-class content had this moment yielded a different choice.

 

 

A democratic order must serve two major functions: first, it must ensure participation (and this, in modern representative democracies, will imply contestation); but second, it must also deliver public policies that yield satisfactory outcomes. A purely procedural democracy is unlikely to be a robust democracy. The recent transformations of the party system have delivered little that is encouraging in this respect.

A generic characteristic of the public policies adopted in India over the last decade is the short time horizons embodied in them. This is so not because of some axiomatic logic written into the very character of democracy; after all, a platform advocating longer time horizons and collective patience may also democratically succeed, witness a number of (non-party, social movement based) environmental movements. Policies instantiating short time horizons are, rather, simply an empirically more likely result in conditions such as those prevailing in India – of simultaneous extreme economic-cum-social inequity and an increasingly competitive party system committed to providing short-run benefits to constituents.

 

 

One important facet of this is the well-known crisis of public finance plaguing most state governments in India, in particular the larger states. No mean contribution to this crisis has been made by political incentives for short-run mismanagement of the exchequer. Examples include writing off interest on loans, sometimes even the full amount of the original loans. But overshadowing all else is the single largest scandal of public resource mobilization in the country, namely the state’s inability to tax the rural well-off. The rural rich have thereby managed to exempt themselves from the core responsibility of citizenship of the Indian state. The rise of regional parties with strong constituencies of the rural wealthy strata has helped the prospect of such taxation recede even further.

Such negative examples could be multiplied. It may be more appropriate, however, to end with a more limited question: in terms of policy outcomes, are we necessarily doing much worse than in the days of massive one-party majorities? Possibly not. Whether this offers any reason for hope must remain a question of personal judgement for the observer.

 

 

Footnotes

1. I borrow this phrase from Aseema Sinha’s ongoing Ph.D dissertation, Department of Government, Cornell University.

2. This section is indebted to E. Sridharan, ‘Duverger’s Law, its Reformulations and the Evolution of the Indian Party System’, IRIS India Working Paper No. 35, IRIS, University of Maryland; and Yogendra Yadav, ‘Reconfigu-ration in Indian Politics: State Assembly Elections 1993-95’, Economic and Political Weekly 31(2-3), 1996, pp. 95-104.

3. Cf. E. Sridharan, op cit.

4. Cf. Maurice Duverger, Political Parties. Wiley, New York, 1954.

5. Cf. Gary Cox, Making Votes Count. Cambridge University Press, New York, 1997.

6. The term comes from Joseph Schlesinger, Political Parties and the Winning of Office. University of Chicago Press, Chicago, 1991.

7. S.M. Lipset and S. Rokkan, Party Systems and Voter Alignments. Free Press, New York, 1967.

8. One implication flowing from the above characterization is that any equilibrium achieved by a party system will always be subject to the possibility of disequilibrium. Specifically, this disequilibrium is not of the type attributable to long-run changes – in, say, the state of the economy – but to more short-run factors directly traceable to the mobilizing and cleavage development activity undertaken by party-political entrepreneurs.

9. E.E. Schattschneider, The Semisovereign People. Holt, Rinehart and Winston, New York, 1960, p. 63.

10. This point is elaborated in my ongoing Ph.D dissertation, Department of Government, Cornell University.

11. D. A. Low, ‘Congress and "Mass Contacts", 1936-1937: Ideology, Interest, and Conflict Over The Basis of Party Representation’, in R. Sisson and S. Wolpert (eds.), Congress and Indian Nationalism: The Pre-Independence Phase. Oxford University Press, Delhi, 1988.