The dynamics of federal (in)stability and negotiated cooperation under single-party dominance: insights from Modi’s India

ABSTRACT Many studies have demonstrated a trend towards greater centralization of the federal system in India since 2014. However, the causal mechanisms that underpin this centralization remain underexplored. This article systematically explores the processes involved in tipping the delicate federal balance in favour of greater central control, and in doing so pays particular attention to the role of the party system therein. Our analytical narrative is structured around the assertion that ‘negotiated cooperation’ is contingent upon: (a) the interaction between constitutional structure and federal politics; (b) the capacity and willingness of subnational incumbents to safeguard their policy space and (c) the nature of party organization and the ideological disposition of the polity-wide parties towards territorial power-sharing. Drawing from India’s experience since 1952, we demonstrate that a dominant party system does not necessarily entail encroachment on subnational authority. At the most, it is a necessary but insufficient condition for federal imbalance. However, the problem of federal encroachment arises only when the dominant party is internally centralized and personalized, advocates unitarism and operates in a context of weak constitutional, political, procedural and judicial safeguards for self-and shared-rule. India under the leadership of Indira Gandhi, but more recently of Narendra Modi, comes close to meeting these conditions.


Introduction
Many studies have demonstrated a trend towards greater centralization of powers in India since Prime Minister Narendra Modi entered office in 2014. However, the causal mechanisms that underpin this centralization of the federal system under Modi remain underexplored. This study seeks to systematically explore the processes involved in tipping the federal balance of powers in favour of greater central control over the states, and in doing so pays particular attention to the role of the party system therein. Drawing from India's federal experience since 1952, we demonstrate that the presence of a dominant party does not necessarily entail federal centralization or the encroachment on subnational autonomy. At the most, it is a necessary but insufficient condition for federal imbalance and instability. In this article, we advance three propositions to demonstrate that the problem of federal encroachment under one party dominance arises only when the dominant party is internally centralized and personalized, advocates a unitarist ideology, and operates in a context of weak constitutional, political, procedural and judicial safeguards for protecting state autonomy. India under the earlier leadership of Indira Gandhi, but more recently that of Narendra Modi, comes close to meeting these conditions. Beyond providing some insights into the state of federalism under different forms of party dominance throughout India's history, we seek to contribute to the broader debate on the relationship between the party system, parties and the organization of the state in two ways.
Firstly, an analysis of the Indian case may help us understand the direction of this relationship. Scholars of federalism and territorial party politics have long recognized the importance of political parties and the structures of the party system in determining federal dynamics and policy outcomes. For instance, Elazar (1987), citing the examples of the United States and Canada, argued that a non-centralized federal system requires the existence of a non-centralized party system. Moreover, Elazar, echoing Riker (1964) and Grodzins (1960), emphasized the significance of internally decentralized parties for maintaining a non-centralized federal system. Hence, the decentralized structure of American parties (despite the US nationalized two party system), protected the integrity and power of the states in the U.S. federal system. It follows that while in the USSR and Mexico highly centralized parties in structure and ideology have indeed limited the power of constituent polities (or subnational governments, hereafter SNGs or states) in direct proportion to the extent of their dominance, this was not the case for the League of Communists in erstwhile Yugoslavia because of its internal decentralization. Furthermore, dominant parties can be ideologically committed to the federal principle (that is, respect the vertical distribution of power in the state and the organization of the latter based on a combination of self-rule and shared rule; Burgess 2012) as we will demonstrate with reference to Congress in India under Nehru. For some scholars though the direction of the relationship between parties and the multi-level structure of the state runs in the opposite direction (Chhibber and Kollman 2004;Truman 2011). A centralized allocation of governmental authority (a) increases the capacity of national parties to discipline subnational co-partisans, (b) generates incentives for the coordination of party activity across constituent boundaries, and (c) strengthens the coattail effects running from national to subnational co-partisans. Conversely, federal decentralization is associated with party system dissimilarity across jurisdictions (Thorlakson 2003), lower levels of party system nationalization (Schakel 2013) and weak national coattail effects (León 2014). Disproving this assumption, the Indian experience demonstrates that what by and large has been a centralized federal system throughoutat least in constitutional, legal and fiscal termsdid not stop the emergence of a very pluralized or decentralized party system . But equally, querying the reverse causality, the Indian state did not decentralize nearly as much as might have been assumed in the context of a pluralized party system, just as it did not necessarily centralize as much under certain phases of one party dominance. Therefore, our article shows that the relationship between centralization and the party system is complex and bidirectional (Watts 2004) andcruciallythat it is mediated by the organizational properties and ideological commitment to the federal principle by the parties in the multi-level party system.
Secondly, our analysis contributes to the literature on federal stability. We define federal stability as a condition in which both, the federal government and the subnational governments (hereafter states), acquiesce to each other's legitimate authority. In a 'stable' federal system, this relationship is charactised as a form of 'negotiated cooperation' in which the national government does not overawe the states by imposing its policy preferences on the latter, and the latter do not violate federative constraints either by pursuing the narrow interests of their constituents, or by free-riding on the common national pool of resources. Observing a positive correlation between federal centralization and federal stability, Riker (1987) held the view that the centralization of the political party system can effectively limit intergovernmental conflict. However, Filippov, Ordeshook, and Shvetsova (2004) took a more nuanced view by linking federal stability to the dominance of integrated parties in the party system in which a national party label or leader may help the election of subnational co-partisans (when the national candidate or party is popular), while the prospect of career advancement incentivises subnational co-partisans further to integrate their campaigns with national co-partisans. This symbiotic relationship, which is built on a seamless partnership between national and subnational incumbents is considered to facilitate the coordination, success and stability of a federal system. Therefore, the second objective of this article is to unravel the causal mechanisms and temporal pathways underpinning the (in)stability of federal systems under dominant party systems, using India as a reference case. Our article shows that in India, especially given the context of a comparatively centralized federal design, one party dominance generates imposed cooperation for most of the time due to the centralized nature of the dominant party in organization and ideology. Only the Congress Party under Jawaharlal Nehru defied this characterization.
In the remainder of this article, we first set out our theoretical framework in more detail and introduce three theoretically informed propositions. Next, we apply this framework to the case of India, by contrasting India's federal experience post-2014 with the period before. The conclusion summarizes our main argument and identifies avenues for further research.
Cooperative federalism in a dominant-party federal system: conceptual overview and three theoretical propositions We first develop a typology of cooperation to account for different types of cooperation under different structures of party organizations, types of party systems, and multi-level polities. We consider three types of multi-level politiescentripetal/unionized, federalized and centrifugal which correspond to three ideologies, namely unitarism, federalism and autonomism respectively. Political parties may be organized around these ideologies ( Figure 1). Autonomists are wary of the homogenizing and uniformizing tendencies of federalism (Lluch 2012), whereas unitarists are wary of federalism's centrifugal and disintegrating tendencies. Both positions are extreme. On the other hand, federalists support polities that combine principles of 'shared rule' and 'self-rule' in their territorial designs (Elazar 1987). However, federal systems vary in the degree of self-rule, in the manner in which self-rule and shared rule are combined, and in the extent to which some unitary features are incorporated into the federal constitution (Watts 2008). A federal polity may allow subnational incumbents to participate in policymaking 'as junior partners' (exhibiting limited interdependency), or it may empower them to collaborate as equal partners (exhibiting a high degree of interdependency). In the former case, which may be termed a centralized federal politystates agree to surrender their powers to the central government, either in the wider national interest or in exchange for policy benefits and resources. In the latter casewhich may aptly be called a non-centralized federal politystates acquire a stronger say in determining how the centre is run.
To distinguish between the various approaches to intergovernmental cooperation, we borrow from international theories of cooperation among nations (Young 1989) and adapt them to intergovernmental interactions in multi-level polities (Sharma 2022).
. Tacit cooperation happens without communication because the expectations of the actors converge in a superior-subordinate type relationship. . Imposed cooperation happens when the stronger party (the central government in a federal polity) can force other parties (SNGs) to alter their policies. This generally happens under a 'dominant party equilibrium', whereby one political party dominates the federal political landscape. Mechanisms for centre-state interaction exist, but SNGs adjust their policies to match the national policy stance. . Negotiated cooperation involves an explicit bargaining federal process, whereby both parties adjust their policies to realize mutual gains. . Contingent cooperation happens when regional governments are strong and can choose to exercise restraint or oppose or extend active cooperation with the national government.
Negotiated cooperation under a dominant-party federal system?
In a system where polity-wide parties control the central government and state-based parties control SNGs but neither relies on the other for electoral success 'negotiated cooperation' is the only way to establish intergovernmental partisan harmony and federal stability (Cairns 1977, using Canada as an example). Moreover, as the American example illustrates, 'negotiated cooperation' is also encouraged by a system of decentralized national political parties, or by a decentralized party system in which national incumbents are dependent upon subnational party organizations for their political fortunes and hence respect substate sovereignty (Grodzins 1960). We argue here that in a dominant party system (in which a single party controls both levels of government), 'negotiated cooperation' occurs when the internal structure of the dominant ruling party is relatively decentralized. In this situation, two-way electoral 'coattail effects' between the national and subnational incumbents create incentives for cooperation and coordination across and between levels of government. However, when the dominant polity-wide party is highly disciplined, hierarchical, and centralized, the scope for negotiated cooperation will depend on (a) the extent to which the dominant party is restrained by the procedural and political safeguards inherent in federal institutional structures and the role of courts in safeguarding federalism; (b) the ability and willingness of the subnational incumbents to resist their national counterparts, especially when the centre's policy stance is perceived as transgressive and conflicts with the electoral and policy interests of co-partisans at the subnational level; (c) the existence of specific policy areas in which the central government depends on SNGs for their implementation or where strong protections for SNGs are constitutionally embedded or require a consultative process; and finally (d) the extent to which the national executive is committed to the 'spirit of federalism' (Burgess 2012). When all these conditions are absent, an internally centralized and hierarchical dominant party, can 'impose' cooperation or generate 'tacit cooperation' simply by controlling the political careers of subnational co-partisans, making them compete for nominations, retentions, and promotions etc. The hierarchical control by a dominant party uses vertical carrots and sticks to generate intraparty discipline and intergovernmental partisan harmony. The dangers of central imposition or intense centralization of power are evident from the experiences of several countries such as Venezuela, Mexico, Argentina, India (under Indira Gandhi and Narendra Modi's leadership), and Nigeria.
Conversely, when the national party is weakeither because its survival at the national level depends on other parties' support or because the partisan alignment of most of the states differs from that of the national ruling partycentre-state interaction can produce what we call 'contingent cooperation'.

Three propositions
We tie the new conceptual framework on intergovernmental relations and our appreciation of the theoretical principles which underlie federal stability into a network of three interrelated propositions: Proposition 1: Negotiated cooperation requires 'constitutionally guaranteed' shared rule (to enable states to shape national policy and influence decision-making) and self-rule (to enable states to exercise authority within their jurisdictions). However, the extent to which states can engage in independent or interdependent policymaking is determined not only by the formal constitutional structure but also by the nature of informal federal politics and processesparticularly, the type of party system and the territorial organization of parties. While neither of them alone can predict the extent of subnational autonomy or codetermination, the outcome is dependent on their interaction. The centralization of all threeconstitutional structure, party system and dominant partyis incompatible with negotiated cooperation.
Proposition 2: Subnational authority is a necessary but insufficient condition for exercising two distinctive domains of authority: self-rule and shared rule. Both require capacity and a willingness at the subnational level to exercise that authority. Without these, the central government will have no incentive to incorporate states into its decision-making procedures, and centre-state relations resemble a hierarchical (tacit or imposed cooperation) rather than an equal partnership (negotiated cooperation). Subnational capacity is a function of the incentive structure (rewards and punishments) which motivates incumbents to exercise authority. This incentive structure is a product of (a) the political environment within which SNGs operate, particularly the nature of the party system (b) their strength at the state level (strong leaders can assert their autonomy in relation to the central leadership irrespective of their party affiliation) and (c) subnational financial autonomy.
Proposition 3: Irrespective of the party system, the conditions for 'negotiated cooperation' are more favourable when the main polity-wide parties and their leaders are organizationally and ideologically disposed towards territorial power-sharing or autonomy.
Institutions, incentives, and ideology: analysing the three propositions underlying the theory of negotiated cooperation with reference to India Proposition 1: Negotiated Cooperation is contingent upon Interactions between the formal constitutional structureparticularly, the extent of institutionalization of shared rule and self ruleand the nature of informal federal politics and processes.
The interaction between constitutional structure and federal politics  The constitution of India has a sovereigntist, statist bias whereby the interests of India's sovereignty and integrity are upheld as sacrosanct. Despite this, some scholars such as Granville Austin see in its actual working an example of 'cooperative federalism' (Austin 1966, 187) in the sense of interdependence and cooperation between the centre and the states. Interdependency, as argued by Austin (1966) and Appleby (1953), is reflected in the fact that states are financially dependent on the national government, while the centre depends on states to administer national policies and social action programmes. Although the Indian Constitution gives the states significant powers within their jurisdiction, in comparative constitutional terms, Indian federalism is rather centralized (Swenden 2016). This is paralleled by the comparative weakness of 'intra-state federalism', i.e., of channels enabling the states to influence central policy on issues affecting their autonomy (Watts 2019). For instance, the powers of the Rajya Sabha, the national second chamber are subordinate to the lower house. On the other hand, the institutions of executive federalism are either almost moribund (for example, the Inter-State Council), advisory in nature only (such as the Governing Council of the NITI Aayog) or sectoral and hence limited in scope (for example, the GST Council, which is an executive forum for deal making on GST matters only). Despite their role in the administration of central legislative policy, states are mandated to follow the instructions of the central government. Furthermore, the Indian Administrative, Forest, and Police Services provide an integrated but also centralized elite administrative structure even though senior civil servants are politically accountable to the state government under which they serve. Nevertheless, the architecture of the Indian constitution has made both centralization and decentralization possible, given its 'flexible' underpinnings. Indeed, under this 'flexible' model, India has experienced periods of high centralization as well as episodes of relative decentralization. However, by overemphasizing the need for flexibility, the founding fathers enabled centralization and authoritarianism. For instance, the Congress Party, under the leadership of Indira Gandhi reduced the system almost to a unitary government. In contrast, India's experience during the coalition era  illustrates how states acquired a greater role in shaping public policies and their outcomes when (a) regional parties participated in a national governing coalition and (b) national parties (such as Congress or the BJP) depended on them to maintain the cohesiveness of the coalition. In fact, a pluralized party system maximizes opportunities for regional agency while one-party dominance minimizes it.
The relationship between the party system and federalism, as discussed in the introduction, indicates that ceteris paribus, the likelihood of central intrusion or transgression is high under a dominant-party system because such attempts can be repelled only if all, or most of the states, challenge the central government. Under a dominant party system, opposition-ruled states, which are likely to mount a challenge, are outnumbered by affiliated states, which are likely to acquiesce (Sharma 2022). In such a situation, the practice of coercive federalism (Kincaid 1990) or the central government's abuse of its executive and financial spending powers, may become a norm. The claim to cooperation ends when the states are not involved in the formulation of national policies.
The system becomes coercive when the central government influences subnational government policies and micromanages their affairs. In federations across the developing world, strong national incumbents often use their control over public resources to mete out rewards and punishments (González 2015). These assumptions are corroborated in the case of India under Indira Gandhi between 1967 and 1989 (Sharma 2017). The party system was most favourable for 'collaborative federalism' between 1989 and 2014 (Sharma 2015;Swenden 2016). This happened not because of the constitutional strengthening of state autonomy during this period, of which there is no evidence (Singh 2019), but because of the compulsions of coalition politics. The difference between the dominant party era and the coalition era in Indian politics in terms of the nature of centre-state interaction brings to the fore the significance of the interaction between constitutional structure and federal politics.
Although previous experiences with one-party dominance (1952-1966 and 1966-1989) also have been associated with the centralization of powers, there is a notable difference. While Nehru's India exhibited the traits of negotiated cooperation and partnership between a strong centre and strong states (Brass 1994;Kochanek 1968), coercion became the norm under Indira Gandhi's rule (Maiorano 2015). She 'captured' independent institutions and brought about a dangerous conflation of party and state (Kochanek 1976). The difference between the two phases of the Congress dominance (under Nehru and Indira) shows that whether a dominant party practices cooperative federalism or not depends on the politics of the period. For instance, under Nehru's leadership, the Congress party never had serious doubts about its electoral future. Therefore, as Chatterjee (1997, 11) points out, an interesting characteristic of the one-party dominant system in the Nehru era was the large degree of autonomy that the provincial party units were able to assert in relation to the central leadership … . the post-independence Congress party was federal in character with a wide organizational network.
On the other hand, state politics during the Indira era, from 1967 onwards, became highly competitive. The theory that 'the future belongs to provincial parties' appeared on the political horizon. Facing defections and the rise of opposition parties at the state level, Indira Gandhi centralized the Congress party, restructured the State legislative elites from above and used President's rule (Art 356) to destabilize opposition party governments.

The interaction between constitutional structure and federal politics (2014 onwards): towards unilateralism
Modi's government is facing a scenario similar to the Indira era in Indian politics. There is stiff political competition at the state level, which has resulted in partisan polarization among states on the one hand, and a blending of cooperative and coercive tactics at the centre (Sharma and Swenden 2018). We argue that one-party dominance at the central level of government does not necessarily trigger a more centralized polity if it coincides with a constitutional system in which (a) the powers of the states are constitutionally entrenched, thereby guaranteeing state autonomy or self-rule, (b) mechanisms of shared rule, such as a federal second chamber, complement the self-rule attributes of a federal system and (c) the Supreme Court acts as a 'federal safeguard' against the encroachment on state rights. However, the founders of the Indian constitution left the nature of Indian federalism highly vulnerable to the compulsion of politics. Subsequent constitutional amendments have not altered this fact.
Therefore, Modi has used the BJP's dominance to strengthen the capacity of the party to exert its constitutional authority and to weaken the shared and self-rule attributes of Indian federalism. Unilateralism marks the Modi government's administrative style in policy formulation. At times, the policy decisions are straightaway announced on TV without much space for consultation with stakeholders such as state governments, opposition parties, interest groups or civil society organizations. Demonetization, the response to the COVID-19 pandemic, and the de facto abrogation of Article 370 illustrate this approach.
Demonetization was announced in an unscheduled live national TV address on 8 November 2016. The unilateral decision to withdraw 500 and 1000 rupee notes as legal tender (accounting for 86% of currency in circulation) was taken without any prior consultation with state governments. This was so despite its severe impact on their activities (e.g., access to liquidity and rising demands for MGNREGA-funded labour in the face of sudden, especially rural, unemployment).
The Indian approach to handling of the COVID-19 pandemic, at least at its outset has been marked by unilateralism and non-coordination. During the first wave of the COVID-19 pandemic in 2020, the centre and the states initially operated under (subsequent) iterations of the Epidemic Act (1897). On that basis, the centre issued travel guidance, regulated air, and port traffic and oversaw surveillance and quarantine systems. In light of rising cases (with high inter-state variability), the Modi government subsequently announced one of the world's strictest lockdowns in a nationalized TV address (Gettleman and Schultz 2020). This was done without prior consultation or planning with either the relevant central ministries or the state Chief Ministers (Swenden 2020;Tremblay and George 2021, 182). Once the announcement was made, it was legally backed by invoking Section 6 (2)(i) of the National Disaster Management Act, 2005 which effectively led to the take-over of the management of the pandemic and health by the centre. A central taskforce supported by the central bureaucracy and eleven empowered groups with the participation of 15 union ministers was to oversee the planning and execution of decisions. State officials were not involved other than through a requirement to report directly and daily to the National Disaster Management Authority (chaired by the PM) and the Home Ministry (ibid. 182). States could still invoke provisions of Section 2 of the Epidemic Disease Act, 1897, but only to enforce centrally directed restrictions on the movement of people. Seeking reports on matters such as law or health, both lying in the exclusive domain of the states, is unusual, but it is largely consistent with a centralized approach to centrestate relations. After the initial strict lockdown, the centre held frequent meetings with state Chief Ministers through videoconferencing, especially in the process of gradual unlocking. However, it was selective in only allowing NDA Chief Ministers to speak, and by that time already lost the good-will of some opposition-ruled states, who boycotted at least some of the meetings. This suggests a return to a more confrontational form of centre-state interactions.
Centralization also marks the provision of financial support during the pandemic. The central government actively discouraged donations to the Chief Minister's Relief Fund and incentivised contributions to the PM-CARESwhich was created overnight on 27 March 2020 without parliamentary approval or consultations with the opposition parties or the state Chief Ministers. It is unclear why this fund was created when PMNRF (Prime Minister's National Relief Fund) already existed for this purpose. Furthermore, PM-CARES does not come under the audit of the comptroller and auditor general of India nor is it open to public scrutiny because it has not been declared as a public authority under the ambit of section 2(h) of the RTI (Right to Information) Act. On 26 May, the Companies Act was changed retrospectively with effect from March 28, to make PM-CARES eligible for corporate social responsibility donations. Overall, the crowding out of donations to the states was achieved by allowing uncapped corporate donations to PM-CARES and making the same tax-exempt as well as countable against a company's corporate social responsibility obligationsa double benefit of tax exemptionnot available under either the state's Chief Minister's Fund or the pre-existing PMNRF. In addition, the central government's unilateral decision to suspend the MPLAD (MP Local Area Development) Scheme limited the ability of MPs to use these funds to address the pandemic within their constituencies and districts. During the deadly second wave of the pandemic (March-June 2021), the centre attempted to attribute failure to contain infections to the states, but did little to prevent large religious or state election gatherings which contributed to contagion, or to support the free and universal procurement of COVID-vaccines, despite its larger financial capacity. In a Suo Motu case on COVID-19 vaccination policy, the Supreme Courtwhich otherwise sanctioned the Modi government's centralized approach (Swenden and Saxena 2022)found the centre's vaccination policy 'prima facie arbitrary and irrational'. When the Centre submitted that the Court had limited power on policy issues, the Court replied, 'You can't just say that you're the Centre and you know what's right. We have a strong arm to come down on this'. (Suo Motu Writ Petition (CIVIL) No. 3/2021).
A final example of unilateralism is the Modi government's decision to downgrade the constitutionally recognized special status of the state of Jammu and Kashmir (Article 370; 35). Whether replacing the state of Jammu and Kashmir with the union territories of Ladakh and Jammu and Kashmir may mitigate security concerns falls beyond the scope of this article; but we take issue with the unilateral way in which this was decided (Nair 2019). Given that the state was under Governor's rule (the equivalent for President's Rule for Jammu and Kashmir), the government first amended article 367, using the powers of the President under Article 370 (1)(d), and then issued a presidential order under Article 370(3) to declare all clauses of article 370 inoperative except the one which states that all provisions of the Constitution of India shall apply to the state of Jammu and Kashmir. At no point were members of the JK Assembly (suspended under Governor's rule) or the leaders of the most prominent JK parties informed of the decision.
Proposition 2: Negotiated Cooperation is contingent upon the capacity and willingness of subnational incumbents to safeguard their policy space and domains of authority.
States' incentives to engage in independent and interdependent political power  Negotiated cooperationwhereby both parties adjust their policies to realize mutual gainsis possible only when state governments are strong enough to withhold cooperation if they perceive central encroachment on their competencies (Sharma 2022). State governments derive their incentives primarily from the constitution, which guarantees their authority, and provides the states with appropriate institutional channels through which they can voice their concerns and exercise shared rule or intergovernmental mediation. Additionally and more importantly, states also derive their incentives from a belief in their 'political power' which is exercised outside the constitution (Hutchcroft 2001).
As discussed in the previous section, constitutional shared rule provisions in India are comparatively weak. The only formal dimension of shared rule in which states state play a role is their capacity to co-determine constitutional amendments seeking to alter the constitutional division of powers. Weak institutionalization of shared rule mechanisms does not coincide with greater self-rule (as in Canada). In fact, the scheme for the distribution of powers in the Indian constitution does not confer autonomous sovereignty to the states, but rather a functional devolution for administrative convenience. Even so, the states have legal supremacy in the 'limited sphere' left to them by the Seventh Schedule and have full power to act in the fields under their jurisdiction. The capacity and willingness of states to exercise this constitutionally assigned authority is determined by the informal dimensions of political power (Mueller 2014).
As argued in the previous section, Nehru ensured that the regional branches of the Congress party were politically empowered, administratively autonomous, and financially self-sufficient. The autonomy of the state units (Pradesh Congress Committees) however was not a gift from above but was earned by the powerful chief ministers who had impressive grassroot following and were closely involved in local politics.
Similarly, during the coalition era, the SNGs could enjoy measures of independent and interdependent policymaking because the longevity of national ruling coalitions depended on regional parties who either shared executive power as coalition partners or extended issue-based support from outside. Thus, we argue that the incentives for the states to engage in independent and interdependent political power and the incentives of the centre to incorporate SNGs into its decisionmaking procedures were strong during the periods of Congress dominance under Nehru/Shastri (1952)(1953)(1954)(1955)(1956)(1957)(1958)(1959)(1960)(1961)(1962)(1963)(1964)(1965)(1966) and multi-party coalition governments under NDA (1999)(2000)(2001)(2002)(2003)(2004) and UPA (2004UPA ( -2014. This was not so during the Indira-Rajiv Gandhi era . Scholars of Indian politics such as Brass (1994) and Frankel (2006) have noted that Indira Gandhi's experience with 'Congress rule through powerful Chief Ministers' (as president of the Congress Party during the premiership of her father J.L. Nehru) was bitter. She had witnessed her father's dependency on powerful chief ministers when he attempted to implement national policies at the State level. Her suspicion was strengthened in the course of the power struggle with those leaders after Nehru's demise. The struggle culminated in the Congress party's split in 1969. Thereafter, Indira Gandhi over-centralized party-politics, tightly regulated state legislative bodies and established direct control over state units of the party. She bypassed state party leaders, appealed directly to the electorate to vote in her name, and forged direct ties with local elites. In fact, Indira Gandhi developed a system where Chief Ministers were cherry-picked by her and remained in office at her pleasure. She frequently used Article 356 (President's rule) to assume direct control over those state governments in which factional conflict prevented the establishment of a stable government or in which non-Congress parties held sway or when she grew suspicious of her own party's Chief Ministers.
Overall, we argue that that the capacity of states (collectively as well as individually) to engage in independent or interdependent policymaking has varied considerably from one ruling dispensation to the othereven if there has been no change in the range of constitutional competences. The difference between Congress rule under Nehru and Indira Gandhi shows that a dominant party might practice negotiated cooperation if it has a consensual structure and the national party's subnational co-partisans have an independent base of local grassroots support.

States' incentives to engage in independent and interdependent political power (2014-)
The politics under Narendra Modi shows great structural and functional similarity to the Indira era. Like Indira Gandhi, Modi has also turned elections into populist referendums. Like Indira, Modi encourages voters to focus on his policies and leadership (rather than state level leadership) and uses direct, plebiscitary appeals to circumvent potential rivals among the regional elites. Interestingly, in most of the assembly elections contested since 2014, the BJP has campaigned around the central rather than state leadership. The party uses the phrase 'double engine government' during campaigns which is an appeal to voters to choose the same party in these states as the one ruling at the Centrethat is the BJP.
Furthermore, the ruling party has established an equilibrium where affiliated chief ministers are unwilling to exercise independent or interdependent political power while the opposition chief ministers are incapable of doing so. This happens because, in this system, the centre's attempt to encroach on states' powers is not challenged by the affiliated chief ministers who rule most of the states. They rather provide tacit support to the central government's encroachment on state powers. This system works for the BJP because its chief ministers depend on the central leadership for enjoying the fruits of the office, but the central leadership does not depend on them for winning elections. Therefore, only the opposition-ruled states challenge the centre's attempt to impose pan-India policies that constrict their jurisdiction; however, their challenge fails because they are outnumbered by the affiliated states who acquiesce. The net result is a mix of tacit cooperation and imposed cooperation. This system is again similar to the dominant party equilibrium associated with Indira Gandhi.
In addition to rendering the states politically incapable of creating conditions for negotiated cooperation, the Modi government has also attempted to diminish the fiscal capacity of the states. As we show below the purpose of the Modi government is not to financially empower state governments but rather to increase its own fiscal space while making states financially dependent on the centre. This fiscal centralization has happened despite the XIVth's Finance Commission's (hereafter FCa body considering the global revenue expenditure requirements of the states-) recommendation in 2015 to increase the states' share in the divisible pool of central tax revenue from 32% to 42%. Although the Modi government continued the convention of accepting its recommendations, it fostered the perception that the XIVth FC's award would shrink the fiscal space available to the union. In reality the recommendations represented 'a compositional shift in transfers from grants to tax devolution' without imposing 'an additional fiscal burden on the Union government' (XIV FC report, para 8.13, 90). 1 Interestingly, the XIV FC's compositional shift from grants to tax devolutionsignifying an increase in the decentralizing componenthas been effectively reversed by 2020-2021 (see Figure 2). In fact, throughout the XIV FC award period, the Modi governmentacting on the perception that the commission's recommendations would deplete the central government's resourcesproceeded to expand its fiscal space, at the cost of the states. The Union government has accomplished this in four steps which are set out in detail below. Crucially the BJP-ruled states were largely acquiescent in this change.
Step 1: Reduction of Central Contributory Share in Centrally Sponsored Schemes (CSS) The XIVth Finance Commission (FC), in its 2015 report, had recommended that CSS (cost-sharing grants)which limit the ability of States to legislate on subjects within their competencebe rationalized, and more flexibility be given to the states in their implementation. Although the Commission had identified over 30 CSS to be delinked from central support (giving states full autonomy to (dis)continue them), the central government delinked only eight, citing national priorities and legal obligations (Sharma and Swenden 2018). However, the centre reduced its cost-sharing of CSS from 75% to 60%, forcing states to contribute moreeven more than what the XIVth Finance Panel had factored into its recommendations. Despite this reduction, the total outlay of these schemes has increased fivefold from 5% in 2016-2017 of total transfers to 25% in 2019-2020 (see Figure 3). This implies that CSS have continued to proliferate throughout the fourteenth FC award period. Although the XVth FC, like its predecessor has recommended that the number of CSS, as well as outlays for them, be reduced, the number of umbrella schemes have continued to go up from 29 in the 2019-2020 budget to 35 in the budget 2021-2022. 2 Coupled with the obligation to adhere to fiscal discipline targets, this led to a fall in the level of state development expenditure from 12% of GDP in 2016-2017 to 11% in 2017-2018 while their capital expenditure decreased from 3.4% to 2.5% (RBI 2020). The resulting financial strain on state budgets prompted state Chief Ministers to protest against these 'under-funded mandates' at the fifth meeting of the NITI Aayog's Governing Council on 15 June 2019, but to no avail.
Step 2: Increase central discretion in the disbursement of central revenue to the states The central government has increased the level of discretion (and therefore the scope for political patronage) in the disbursement of central revenue to the states. Apart from CSS, until 2015 there used to be a specific fund, earmarked as 'Normal Central Assistance for State Plans' which was formula-based (known as the Gadgil-Mukherjee formula), and therefore predictable in its allocation from one 5-year Plan to the next. The Modi-led government, alongside the scrapping of five-year and annual plans beyond 2017 also dropped Normal Assistance for State Plans (State Finance: a study of Budgets, RBI, 2019). This increased the latitude of the central Ministry of Finance to devolve (or deny) the equivalent funds to the states in whichever manner preferable. Consequently, as Figure 4 shows, during the period between 2015 and 2019, the average annual growth rate of discretionary ad hoc grants (in absolute totals) disbursed to the states grew exponentially to 46% compared with just 6% between 2010 and 2015.
Step 3: Minimize the relative size of the divisible pool To offset the implications of the FC recommended 10% increase in the states' share in the divisible pool, the centre has sought to minimize its relative size, primarily by increasing revenue bases that credit the central government alone. Hence, the centre has increased its reliance on non-shareable cesses and surcharges (as Figure 5 attests), while reducing the rates of shareable taxes. For instance, in the 2015-2016 budget, the wealth tax was abolished, and in the 2016-2017 budget, the tax rate on income between 2.5 and 5 lakh rupees (0.25-0.5 million) was reduced (both within the divisible pool); however, the surcharges on taxable income (not within the divisible pool) were increased by an equivalent amount for individuals with an income below 5 lakh, as well as for the so-called highnet-worth individuals. Similarly, in the 2018-2019 budget, the excise duty on petrol was reduced by Rs. 9 per litre (in the divisible pool), while the road cess (not within the divisible pool) was hiked by a comparable amount. The XVth FC, like its predecessor, highlighted the problem of growing cesses and surcharges (XVth FC report, 67-68), and recommended a higher level of grants as a counterbalance to the states' dwindling fiscal space over the years. However, the central government has declined to accept the FC's recommendations on state specific and sector specific grants. At the same time, the Union Finance Minister in her Budget speech on 1 February 2021 announced yet another cess to boost agriculture infrastructure (see Budget 2021-2022).
Step 4: Diminish States' Fiscal Autonomy without expanding untied revenue resources Going by the criteria developed by Hooghe et al. (2016), the fiscal autonomy score (as an expression of self-rule) of India has dropped from 4 (the regional government sets the base and rate of at least one major tax) to 2 (the regional government sets the base and rate of minor taxes), after the implementation of GST. However, when the state governments ratified the GST bill in 2016, they expected the expansion of an independent revenue source ('shared taxes' and untied grants) coupled with a generous compensation package to offset their loss of control over own-source revenue (sales tax) in the wider fiscal federal structure. The reality has turned out somewhat differently, especially since the BJP increased its majority in the 2019 general elections. Rather than offsetting the states' loss of fiscal autonomy by increasing their access to untied resources through tax sharing, the centre moved to shrink the divisible pool by overtly relying on cesses and surcharges and reducing its contributory share to CSS without making the latter voluntary. Furthermore, as the tax-to-GDP ratio declined after the implementation of the GST due to a shortfall in GST collections (Govt. of India 2020), the union government defaulted on the bi-monthly instalment in October 2019 and yet another one in December 2019. In the 41st meeting of the GST council (held on 27 August 2020), the centre refused to compensate the states, suggesting they should borrow instead. However, when the opposition-ruled states joined forces to approach the Supreme Court against what they called a 'discriminatory and illegal action of Centre regarding GST Compensation', the centre agreed to borrow and pass it on to the states (Sharma 2022).
Proposition 3: Negotiated Cooperation is contingent upon the nature of party organization and the ideology of its leader or party.
Even under one party dominance, negotiated cooperation is possible, but it requires an ideological openness of the dominant party to federalism and an internally decentralized party which leaves considerable room to state party branches in the process of candidate selection, state policy formulation and finance (Fabre and Swenden 2013;Shrimankar 2020). As we have shown above, the Nehru and Indira Gandhi periods of Congress dominance are differentiated from each other because the autonomy and participation of sub-state branches was tolerated under Nehru only (Brass 1994;Chatterjee 1997;Frankel 2006;Kothari 1970;Lijphart 1996). This factor accounts for the different types of cooperation under the two leadersnegotiated cooperation under Nehru and imposed cooperation under Indira. Narendra Modi's politics is reminiscent of the territorial dynamics under the Indira Gandhi era, but he goes further in seeking to establishing an electorally legitimized majoritarian state based on a unified Hindu nation (Jaffrelot 2021). Pandit Deendayal Upadhyaya, an RSS ideologue and progenitor of the doctrine of integral humanism, after whom the Modi government has named many central schemes, explicitly demanded a strong, monolithic, and unitary form of government in which minorities would be assimilated into a dominant Hindu culture (defined as the 'Indian' culture) and be made to pay allegiance to the motto of 'Hindi, Hindu, Hindustan' (Upadhyaya 1955). However, constitutional veto points, such as the FC (e.g., the XVth FC did not proceed in the manner suggested by the Modi government), Supreme Court (e.g., in the policing of President's Rule in relation to Arunachal Pradesh and Uttarakhand) can slow down the pace at which centralization is achieved. To some extent, these rule-based institutions have been emboldened by the coalition-era, and in this at least, they have generated some degree of 'stickiness' which the BJP cannot always undo. On balance, our Note: GST compensation cess has been excluded because its proceeds are shared with states.1 Crore denotes 10 million.
analysis in this section reinforces that under Modi's BJP rule, the spirit of federalism has been in short supply, despite declarations to the contrary.

Modi's political unitarism and one nation nationalism
Unitarism refers to 'a political system where the national government is sovereign relative to its territorial units' (Gerring and Thacker 2004, 303). Gerring and Thacker further note that unitarists believe in non-federalism and in unicameralism, or at best weak bicameralism. BJP ministers have often expressed their ideological position against strong bicameralism by stating that the lower house should not be questioned by an indirectly elected second chamber in a democratic republic.
Indeed, evidence supports the view that the BJP, using its absolute majority in the Lok Sabha, is pushing India's parliamentary federalism towards centralized unitarism. If not for the Supreme Court, the Modi government would have imposed President's Rule in states more frequently to unseat or keep the opposition out of power (Khaitan 2020). In terms of undermining the legislative process, the Modi government undid the Pre-Legislative Consultation Policy (PLCP), introduced by the previous UPA-coalition. Most legislative bills are no longer routed through Parliamentary committees, bypassing important mechanisms of legislative scrutiny. While the UPA-coalition government (15th Lok Sabha) referred 68 bills to parliamentary committees, the BJP-dominated 16th and 17th Lok Sabha did so for only 16 and 7 bills respectively (loksabha.nic.in). PRS Legislative Research (2021) shows that under the UPA-coalition governments (2004)(2005)(2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014), 60 (UPA I 2004(UPA I -2009) and 71% (UPA II 2009-2014) of bills were referred to parliamentary committees. This share has dropped to 25% (2014-2019) and just 11% (2019-present) in the 16th and 17th Lok Sabha respectively.
In addition, many bills are marked as 'money bills' to prevent their more detailed scrutiny by the upper house. Important bills treated in this way during the current legislative term are the Citizenship Amendment Bill, 2019 and the so-called farmer's bills (2021withdrawn after farmer's protests in 2022). In its first term in office, when political resistance to the BJP in a more fractious Rajya Sabha was more in evidence, the government also sought approval for its policies by executive ordinance. A case in point was the Land Acquisition, Rehabilitation and Resettlement (Amendment) Bill, 2015, which sought to exempt certain provisions of the 2013 adopted bill from consent and social impact requirements and to widen the remit of projects not residing under LARR provisions. The data available at the website of the Ministry of Law and Justice, Government of India (legislative.gov.in) shows that under the BJP-led National Democratic Alliance (2000 and 2004) 40 ordinances were promulgated. During UPA-I and UPA-II the numbers were 23 and 28 respectively. The Modi government has issued 60 ordinances during its first term alone (2014-2019).
Another unitarist trend is the centre's attempt to circumvent state bureaucracies by creating direct chains of command with districts. In doing so, state governments have ceded ground to the centre in holding senior civil servants to account even though they carry the financial and administrative brunt of running intergovernmental welfare schemes (Aiyar 2019). For instance, in seeking to curb resistance from opposition-ruled states to the implementation of the Citizenship Amendment Bill, the central government inserted a clause (section 6B) enabling it to redraft implementation rules so that citizenship applications of persecuted minorities can be dealt with by a central government officer in lieu of a district collector who is normally accountable to the state government. Aiyar and Tillin (2020, 130-131) also observed a general tendency to centralized monitoring in the implementation of welfare programmes, many of which rely on significant (and increasing) levels of state cost-sharing. The PRAGATI process (Pro-Active Governance and Timely Implementation) involves a monthly review of the implementation of centre-state projects but without the participation of state Chief Ministers or cabinet members.
The Modi government's focus on integration and political centrism is further reflected in the framing of certain policies and demands. Aiyar and Tillin (2020) and Kailash (2021) have commented on the ruling BJP's tendency to push the party's ideological world view by prefixing 'One Nation' to policies, laws, and actions. Modi's 'one nation' framing is far from 'one nationism' as some have argued (see Aiyar and Tillin 2020, 132). The term 'one nationism' is associated with the social cohesion politics of Benjamin Disraeli, the former prime minister of the United Kingdomarguably the first 'one nationist' (Cole and Deighan 2012). Disraeli feared that widening social inequalities (the two nations of rich and poor) would threaten the established order. Therefore, he advocated social welfare programmes and committed to full employment. Thus 'one nationism' is a strand of political thinking which has little to do with nationalism or an ideology of uniformity. In contrast, Modi's One Nation framing, couched in terms of enhancing efficiency, actually seeks to promote the governing BJP's populist and nationalist agenda. As it is the opposite of Disraeli's 'one nationism', it may aptly be termed as 'one nation nationalism' which is neither 'national federalism' (Beer 1993) nor 'onenation federalism' (Forsyth 1981) because it violates the very spirit of federalism byemphasizing a unitarist ideology. Modi's nationalist appeals drive the justification behind the push for concurrent national (and at least half of) state assembly elections (Debroy and Desai 2017). It propels the introduction of the GST (One Nation, One Market, One Tax), or a single Ration Card in the administration of the Public Distribution System or One Nation, One Grid. One nation nationalism unquestionably widens the legitimacy of centralization and diminishes support for federalism.
This 'one nation nationalism', also drives the so-called JAM-trinity (Jan Dhan-Aadhaar-Mobile) underpinning the delivery of government welfare schemes. These link unique identification numbers with mobile phone technology and the opening of individual bank accounts. Although the roll-out of digital and mobile technology may reduce leakage or state shirking, it risks centralizing delivery, especially when states may not have the same level of access to information stored in digital databases for the roll-out of similar state-level programmes. More widely, although the centre cut the central cost-sharing of these schemes at the expense of the states, it also widened their remit. Contrary to expectations, Centrally Sponsored Schemes or federal cost-sharing programmes have continued to proliferate under Modi. Yet, to enhance, their 'federal' credit-taking, each of them has had to carry the prefix 'Pradhan Mantri' (Prime Minister), facilitating their association with the centre. In this, CSS which benefited the incumbent state governments during the coalition-era (Sharma 2017;Sharma and Swenden 2020), now appear to boost the electoral performance of the central government (Deshpande, Tillin, and Kailash 2019). The undercutting of the states in the administration of central or centrally sponsored schemes could be offset if the states were involved in co-designing these programmes (shared rule). The creation of the NITI (National Institution for Transforming India) Aayog offered some promise in this regard (Sharma 2015;Swenden and Saxena 2017). However, as shown above, there is little evidence for increasing shared rule in the design or delivery of these programmes. For instance, the NITI Aayog did not increase the co-operative attributes of Indian federalism compared with the Planning Commission (PC) and National Development Council (NDC) which it replaced (Reddy and Reddy 2019; Sharma and Swenden 2018;Swenden 2019;Swenden and Saxena 2017).

Conclusion
In this article, we have used the case of India to underscore that while one-party dominance (or party system centralization) is often associated with centralization, there is no direct causal relationship. One party dominance can coincide with non-centralism and federal balancea situation where both levels of government acquiesce to each other's legitimate authority. Similarly, a relatively centralized federal design may foster party centralization, but can coincide with party system pluralization (as during the coalition-era). This suggests therefore that the organizational properties and ideology of the dominant party or parties play an important intermediate role in influencing federal relations.
We show that disruptive centralization or unilateralism occurs when all three factors, namely, constitutional structure, party system and party-as-a-system, are centralized. A dominant party that is hierarchical, centralized and disciplined may ensure a high level of partisan harmony between national and subnational levels and limit intergovernmental conflict, but is unlikely to guarantee what we call 'federal balance' or negotiated cooperation. At best, it will produce a tacit or imposed cooperation because the transgressive behaviour of the centre encounters little resistance from the affiliated states, reflecting their unwillingness to exercise their authority. Meanwhile, in a dominant party system any challenge by the opposition-ruled states (which are outnumbered by the affiliated states) is likely to fail, reflecting their inability to block or repel, individually or collectively, central policies designed to limit subnational authority. The centre-state relations during the Indira era and during the Modi period bear testimony to this contention. However, centralization during the Modi era is more disruptive than during the Indira era because the BJP is comparatively more inclined towards the ideology of unitarism. Conversely, when the dominant party is internally decentralized, 'negotiated cooperation' arises. Centre-state relations during the Nehru era constitute an example of this phenomenon.