Narendra Modi was sworn in today for a third consecutive term as prime minister of India, but he begins that term under a political structure very different from the one that governed India for the previous decade. In the 4 June Lok Sabha results, the Bharatiya Janata Party won 240 of 543 seats, down from 303 in 2019 and below the 272 required for a majority on its own. The broader National Democratic Alliance still secured a majority with 293 seats, allowing Modi to form government, while the opposition INDIA alliance won around 230 seats and Congress itself increased its representation to 99 seats. The result is therefore neither the sweeping rejection of Modi that some opponents initially celebrated nor the overwhelming mandate the BJP expected after campaigning for more than 400 seats for its alliance. It is something institutionally more interesting: Indian voters returned Modi to office while simultaneously reducing his ability to govern without bargaining with other political actors.
That distinction matters because elections do not simply answer the binary question of who governs. They determine the distribution of bargaining power inside government. Modi remains prime minister, and the BJP remains by far the largest party in Parliament. Yet for the first time since 2014, the BJP cannot pass ordinary legislation or sustain the government entirely through its own Lok Sabha numbers. It now depends on coalition partners, especially regional parties such as the Telugu Desam Party and Janata Dal (United), whose support gives them leverage over appointments, policy priorities and potentially the pace or form of controversial reforms. The same electorate that authorised a third Modi government has therefore imposed a stronger set of internal political constraints upon it.
This is why describing the election simply as a Modi “victory” or “defeat” removes much of the information voters produced. Electoral outcomes are multidimensional. The BJP won enough seats within a successful alliance to keep power. It lost enough seats individually to lose unilateral control. Congress remained far smaller than the BJP but nearly doubled its parliamentary representation. Regional parties regained influence that two consecutive BJP majorities had reduced. The result redistributed power without changing the identity of the prime minister, which is exactly the kind of outcome parliamentary systems can produce.
The election therefore offers a useful correction to the concept of a political mandate. Governments often speak as though victory produces one unified instruction from voters. In reality, hundreds of millions of people vote for different reasons. One citizen supports Modi because of infrastructure. Another values Hindu nationalism. Another prioritises welfare programmes or foreign policy. A voter supporting a regional NDA ally may prefer Modi as prime minister while opposing specific BJP policies. The final seat total aggregates these different motivations without telling the government that every component of its programme has been endorsed equally.
This is particularly important because the BJP had explicitly campaigned around a much larger parliamentary objective. Its slogan of “400 paar” for the NDA created expectations of an overwhelming mandate capable of enabling ambitious policy change. Instead, the BJP’s own tally fell to 240. The institutional meaning is not that voters rejected every BJP policy. It is that the party now has weaker grounds for interpreting the election as authorisation for measures its coalition partners or Parliament strongly resist. A reduced majority does not remove legitimacy; it changes the price of exercising power.
Coalition government raises that price through bargaining. A BJP leader considering legislation now needs to ask not only whether party MPs support it but whether enough allies do. Those allies possess different electoral constituencies and ideological priorities, particularly because several are regional parties whose political survival depends on demonstrating independence from the national party they support in Delhi. Coalition partners therefore act as additional veto players.
Veto players create costs. Decision-making becomes slower. Policies can become diluted. Ministries may be allocated partly according to coalition arithmetic rather than administrative efficiency. Governments can become vulnerable if small partners threaten withdrawal. Economists and investors therefore often prefer clear majorities because predictable governments can implement reforms more decisively.
Yet veto players also create insurance against policy error and over-centralisation. A government possessing a large majority can coordinate quickly but can also implement a mistaken policy quickly. Coalition bargaining forces additional constituencies to examine the proposal before implementation. This does not guarantee better policy, because coalition partners can demand inefficient subsidies or patronage. It increases the number of actors whose agreement is required, trading decisiveness for restraint.
India spent the previous decade operating under unusually strong single-party parliamentary control by the standards of its post-1989 politics. The BJP’s 2014 victory produced India’s first single-party Lok Sabha majority in thirty years, and the 2019 result strengthened it further. Before that, coalition government was normal for decades. The 2024 result therefore represents less a completely new political order than a return to coalition constraint after an exceptional period of centralised party dominance.
This could affect policy substantially even if broad economic strategy remains stable. Moody’s and Fitch have already suggested that the weaker parliamentary position could complicate some difficult fiscal or structural reforms, even while infrastructure spending and broad fiscal discipline are expected to continue. The reason is political economy rather than ideology. Reforms frequently create concentrated losers before generating dispersed future benefits. Governments with large majorities can absorb more short-term political pain. Coalitions need to consider whether one partner’s constituency bears disproportionate costs.
Suppose the central government wants labour-market reforms that increase national efficiency but threaten organised workers in a state where an allied regional party is electorally vulnerable. The regional party evaluates the reform differently from the BJP’s national leadership. National GDP gains are dispersed across India; electoral losses may be concentrated in one state’s constituencies. Coalition government therefore internalises geographic distributional effects that a centralised majority can more easily overlook.
This is one of federal democracy’s recurring tensions: what is optimal for the national aggregate may not be optimal for every political unit composing the nation. Regional parties exist partly to make those distributional differences politically visible. They can obstruct nationally beneficial reforms. They can also prevent national governments from treating regional costs as irrelevant merely because the aggregate calculation is positive.
The election results in Uttar Pradesh demonstrate why the BJP may need to reassess precisely those distributional politics. The state sends 80 MPs to Parliament and has been central to Modi’s national majorities. Yet the BJP suffered major losses there, while the Samajwadi Party and Congress performed far more strongly than anticipated. Even in Faizabad, the constituency containing Ayodhya and the recently inaugurated Ram temple, the BJP candidate lost. Reuters reporting from the area found voters emphasising issues including jobs, inflation and local economic concerns alongside the political importance of the temple.
The Ayodhya result does not prove that Hindu identity ceased to matter. That would be an equally crude interpretation. Voters possess several preferences simultaneously. A citizen can support the construction of the Ram temple and still vote against the BJP because of employment, candidate quality, caste politics or local dissatisfaction. The result therefore demonstrates a broader principle: symbolic politics does not eliminate material politics. Identity can structure voter loyalties without making those loyalties unconditional.
This is important because political analysis frequently searches for one variable explaining electoral behaviour. Religion. Caste. Welfare. Inflation. Leadership. Employment. Voters do not operate through one-variable models. Elections are aggregation processes in which citizens assign different weights to several considerations. The weight of each factor can also change over time. A cultural achievement that dominates politics in January may become one factor among many by June as households return to evaluating prices, employment and local representatives.
The BJP’s campaign itself appeared to shift during the election. Early messaging emphasised Modi’s leadership, development, welfare and the ambition of a large majority. Later phases included increasingly sharp religious rhetoric, with Modi making controversial statements concerning Muslims and opposition redistribution proposals. Reuters identified this escalation as one of the significant features of the campaign. The final result provides no simple statistical way to isolate whether that rhetoric helped, hurt or merely mobilised existing constituencies. What it does show is that a campaign containing intense majoritarian appeals did not produce the overwhelming parliamentary result many exit polls predicted.
This matters institutionally because it places a limit on deterministic arguments about identity politics. Political entrepreneurs can mobilise religion effectively, but voters remain capable of updating on other issues. India is not moving through an irreversible process where one identity automatically determines every election. Electoral coalitions remain contingent.
The opposition’s performance demonstrates the same thing from the other direction. Congress won 99 seats, far below the BJP’s 240 but almost twice its 2019 tally. The broader INDIA alliance improved partly because opposition parties coordinated more effectively in several states rather than dividing anti-BJP votes across multiple candidates. This demonstrates how much electoral outcomes depend on coordination rather than raw popularity alone.
India’s first-past-the-post system rewards whichever candidate receives the largest number of votes in each constituency, even without a majority. If two opposition parties each receive 25% while the BJP receives 40%, the BJP wins despite 60% voting elsewhere. If opposition parties coordinate behind one candidate receiving 50%, the same underlying anti-BJP electorate produces a completely different seat. Political organisation therefore transforms preferences into power.
This is why the election result should not be interpreted solely as millions of individuals independently changing ideology. Some of the parliamentary shift reflects changed voting behaviour; some reflects better opposition coordination; some reflects state-specific alliances and candidate effects. Electoral systems are conversion technologies. The same distribution of public preference can produce very different parliamentary outcomes depending on how parties organise around it.
That makes the revival of regional parties especially significant. National narratives often treat Indian politics as BJP versus Congress because those parties dominate international coverage. In practice, regional parties remain crucial to India’s federal political architecture. Andhra Pradesh, Bihar, Tamil Nadu, West Bengal, Uttar Pradesh and Maharashtra each contain party systems that cannot be understood simply through a national left-right division.
The return of coalition dependence therefore gives federalism greater leverage inside the Union government. A regional party supporting Modi can negotiate not only over ideological policy but over infrastructure, cabinet portfolios, fiscal transfers and state priorities. This can look transactional. It is also one way India’s national system incorporates geographical diversity.
India is not a culturally homogeneous electorate choosing one national programme. It is a federation containing radically different languages, economies, histories and party systems. A political architecture requiring national governments periodically to bargain with regional parties may therefore create inefficiency while reflecting the country’s actual structure more accurately than complete centralisation would.
The economic consequences should not automatically be pessimistic. India’s coalition governments during the 1990s and 2000s implemented important economic reforms despite fragmented politics. The landmark 1991 liberalisation began under a minority Congress government dependent on political support it did not entirely control. Later coalition governments expanded infrastructure, telecommunications and integration with the global economy. Political fragmentation therefore does not mechanically prevent reform.
What changes is the type of reform that becomes feasible. Policies requiring broad consensus can survive. Measures producing large concentrated political losses become harder. Reform therefore shifts from command towards negotiation. Whether that produces better outcomes depends on the quality of the institutions conducting the negotiation.
A strong coalition government needs clear agreements, predictable consultation and enough cabinet discipline that every disagreement does not threaten collapse. A weak coalition can become a permanent auction in which partners exchange support for narrow benefits. The difference is institutionalisation.
Modi himself now faces an important leadership transition because his political brand was built partly around decisiveness and centralised authority. He spent a decade governing with BJP majorities and frequently presented strong executive capacity as an advantage over the fragmented coalition governments preceding him. The political system has now required him to govern through exactly the type of coalition management his rise helped displace.
This creates a test of adaptive leadership. A leader effective under majority government does not automatically remain equally effective under coalition government. The skill set changes. Centralised decision-making gives way partly to consultation. Regional allies need public credit for government achievements. Policy announcements may require negotiation before presentation. Political authority becomes less vertical.
If Modi adapts successfully, coalition constraint could make the government broader without making it ineffective. If the BJP treats allies as temporary arithmetic rather than genuine partners, tensions could emerge quickly.
The coalition partners face their own problem. Supporting Modi gives them access to national government and resources but risks allowing the BJP to expand into their political bases. Regional parties therefore need to cooperate nationally while preserving differentiation locally. This is another principal-agent problem: the NDA functions collectively at the Union level while members compete for political market share.
Their bargaining behaviour should therefore be expected, not treated automatically as opportunistic. A party such as TDP needs voters in Andhra Pradesh to believe national participation produces state benefits. Without visible gains, joining the coalition becomes electorally costly.
This is democratic accountability operating through federal negotiation.
The election also provides evidence against the idea that institutional concerns raised during the campaign made meaningful opposition impossible. Arvind Kejriwal was arrested before the election, Congress complained of frozen funds, critics raised concerns about media asymmetry and electoral institutions, and the BJP possessed enormous financial and organisational advantages. Those issues remain legitimate subjects of scrutiny. Yet voters still delivered an outcome dramatically weaker for the BJP than exit polls predicted and strengthened opposition parties substantially.
That does not prove every concern was unfounded. It demonstrates why democratic quality exists on several dimensions simultaneously. A political system can contain significant inequalities while retaining enough contestability that incumbents suffer meaningful electoral losses. The correct conclusion is therefore neither that India’s democracy is perfectly healthy nor that opposition competition is meaningless.
The election provided real uncertainty and real consequence.
That matters enormously.
A government expecting 350 or 400 coalition seats received 293.
A party accustomed to governing alone received 240.
A weakened opposition returned with far greater parliamentary strength.
The electorate altered the distribution of political power substantially without changing government completely.
That is meaningful democratic feedback.
It may also strengthen institutions indirectly because a larger opposition can scrutinise government more effectively in Parliament. Parliamentary committees, debates and procedural challenges become more consequential when the governing party cannot assume numerical dominance in every situation. The government retains a majority through the NDA, but opposition speech gains greater political weight because alternative coalitions are no longer mathematically irrelevant.
This is information value through opposition. Governments benefit from having institutions capable of identifying mistakes before those mistakes become crises, even when leaders dislike criticism. Strong oppositions act partly as external auditors. They investigate, question and provide alternative narratives. Some criticism will be partisan or inaccurate. The aggregate system benefits from having more than one institution attempting to interpret government performance.
The same applies to media and civil society. Concentrated political power can make negative information harder for leadership to hear because subordinates become reluctant to challenge dominant figures. Electoral losses break through that filter. The 2024 result gives the BJP information it could not receive as clearly through internal party structures: some constituencies that supported it previously are dissatisfied enough to vote elsewhere.
The rational response is diagnosis rather than denial.
Which voters moved?
Why?
Was unemployment decisive?
Inflation?
Candidate selection?
Caste coalitions?
Local governance?
Agrarian concerns?
Opposition coordination?
Different states will produce different answers.
This is why national electoral analysis needs heterogeneity rather than one grand story. BJP losses in Uttar Pradesh may have different causes from Maharashtra. Gains in Odisha reflect another political dynamic entirely. Strong performance in Madhya Pradesh or Gujarat cannot be explained through the same mechanism as losses elsewhere.
India’s scale makes national averages particularly misleading.
The same applies to turnout. More than 640 million votes were cast across the election, making it an extraordinary exercise of mass participation. But those votes were distributed across hundreds of locally specific contests. National democracy is produced through local political systems.
That decentralised complexity may be one of India’s strongest safeguards against complete political monopolisation. A party can dominate the national conversation without controlling every state, caste coalition, linguistic identity or regional organisation. Political power therefore encounters multiple social structures outside the central government’s direct control.
This is societal redundancy.
Institutions matter, but so do independent centres of political organisation embedded in society.
Trade unions.
Religious organisations.
State parties.
Caste associations.
Business groups.
Civil society.
Federal governments.
A national party needs to bargain with some and compete with others.
Plural societies can be difficult to govern precisely because authority is fragmented.
That fragmentation also makes them difficult to monopolise.
The 2024 election has demonstrated the point.
The BJP remains the strongest national political organisation by a wide margin.
It is not omnipotent.
That distinction matters particularly after a decade when Modi’s political dominance encouraged increasingly personalised interpretations of Indian government. Elections remind leaders that personal popularity translates into authority only through institutional arithmetic. Modi remains extremely popular. He still needs 272 parliamentary votes to sustain a government.
That is the value of constitutional procedure.
Popularity does not substitute for seats.
Seats do not substitute for coalition support when one party lacks enough.
The prime minister’s authority therefore remains mediated through Parliament even when political campaigns revolve heavily around his individual identity.
This is personalised politics inside impersonal rules.
Strong leaders naturally attempt to turn elections into direct plebiscites on themselves because personal brands simplify political coalitions. Parliamentary constitutions convert that plebiscitary energy back into constituency results and legislative numbers.
The result can disappoint leaders who believe national popularity should produce direct governing authority.
That disappointment is the institution doing its job.
The economic reform debate should therefore be recalibrated rather than abandoned. India still needs productive employment at scale, more efficient urbanisation, better educational outcomes, manufacturing expansion and sustained infrastructure investment. None of those constraints disappeared because the BJP lost its majority. If anything, coalition government forces the question of which reforms possess enough distributional legitimacy to survive broader consultation.
There is a danger that political competition produces excessive welfare promises or subsidies as parties attempt to protect constituencies. Moody’s has already noted that a smaller mandate could slow fiscal consolidation or encourage more politically responsive spending. Yet fiscal restraint itself has distributional consequences. A government reducing deficits through cuts that fall disproportionately on poorer households can be economically tidy and politically unsustainable.
The relevant objective is not reform regardless of politics.
It is reform capable of surviving politics.
Durable economic policy needs constituencies benefiting enough from it that future governments do not reverse it immediately. Coalition negotiation can weaken technical purity while strengthening political durability.
India’s 1991 reforms endured partly because subsequent governments of different political orientations ultimately accepted much of the new economic framework. Policy became institutional rather than belonging entirely to one party.
The next generation of reform may need the same cross-party or cross-coalition foundation.
This is particularly important for reforms affecting agriculture, labour and federal finance, where distributional effects are large and regional variation enormous. Imposing one national solution quickly may produce resistance. Negotiated reform can be slower while more adaptable.
The result therefore need not be interpreted as voters rejecting development or market reform. It can be interpreted as voters demanding that development remain more politically accountable to its distributional consequences.
The Uttar Pradesh results strengthen that interpretation. Economic growth, infrastructure and religious symbolism all existed simultaneously. Voters still moved substantially. That suggests performance legitimacy remains conditional. Governments cannot assume national achievements automatically outweigh local experience.
This may be the most important political lesson for Modi’s third term. The BJP spent a decade demonstrating that centralised leadership could deliver large-scale policy and transform India’s national political narrative. The electorate has now reminded it that centralisation operates under periodic consent.
Consent can be renewed.
It can also be narrowed.
The result has narrowed it without withdrawing it.
That is more sophisticated than the landslide exit polls expected.
Indian voters have not returned the country to the political instability of the 1990s automatically. They have returned bargaining to the centre of government.
The difference matters.
Narendra Modi remains prime minister.
The BJP remains dominant.
The opposition is stronger.
Regional parties matter again.
Parliament matters more.
All of these facts can coexist.
In fact, that coexistence may be the most democratic interpretation of the result.
A healthy election does not need to destroy the incumbent to constrain them. Sometimes voters produce a more precise instruction: continue, but not on exactly the same terms.
India has done that.
The third Modi government begins today not because the BJP won a majority but because several parties representing different electorates have agreed to sustain one.
That makes the government less unilateral.
Whether it makes the government less effective will depend on whether India’s political institutions can convert bargaining into policy rather than paralysis.
The next five years will test that distinction.
For now, the election has already demonstrated one thing clearly.
Political dominance can survive a democratic correction.
It just does not survive unchanged.
