The Bargain and the Crossing · Chapter 12

Madani in Office: The Reformasi Government Governs (2022–2026)

A note on this draft. This chapter is built entirely from four of the corpus's own working documents — the mid-tenure consolidation record (MY-E-02), the second-phase cabinet-and-coalition record (MY-D-06), the year-three fiscal-and-foreign-policy record (MY-D-07), and the biography of the man at its centre, Anwar Ibrahim (MY-H-PM-09). It carries the heaviest evidentiary caution in the book. Three of its four sources are the most recent and most heavily flagged drafts in the entire Malaysia archive — MY-D-07 alone wraps eighty-eight separate claims in a [TBD-VERIFY] tag, MY-D-06 fifty-four, MY-E-02 thirty-seven — because they chronicle events of 2025 and 2026 that were still unfolding as they were written. Every figure, date, and dollar value in the pages that follow should therefore be read as the sources state it, reported rather than settled: the budget envelopes, the deficit numbers, the fuel prices, the by-election majorities, the ringgit's trading band, the tariff rates. Where a source could not confirm a thing it left a gap, and this chapter honours those gaps rather than filling them. The chapter is also, by the corpus's own framing, contested at its core. Its governing question — was this Reformasi delivered, or Reformasi compromised? — is one the sources pose but deliberately refuse to answer, setting the competing evaluations side by side and attributing each. This chapter does the same. It does not adjudicate; it reports what the reform generation did with power, and the two irreconcilable verdicts that power has drawn.

The Arithmetic of the Improbable

At five o'clock in the evening on 24 November 2022, at the Istana Negara in Kuala Lumpur, Anwar Ibrahim was sworn in as the tenth Prime Minister of Malaysia — an office twenty-four years in the reaching, after a journey through sacking, two imprisonments, a pardon, and the Sheraton Move that had displaced him from a near-certain premiership in 2020. That he reached the office at all is the story of the chapter before this one. What he did with it is this one — and the first thing to understand is that the government he formed should not, by the ordinary logic of Malaysian politics, have existed.

The general election of 19 November 2022 had produced the first formally hung parliament in the country's federal history. No coalition came near the 112 seats a majority required in the 222-seat Dewan Rakyat, the lower house: Anwar's Pakatan Harapan, PH — the reform coalition — led with 82; the Islamist-nationalist Perikatan Nasional, PN, took 74; Barisan Nasional, BN, the alliance that had governed Malaysia without interruption from 1957 to 2018, was reduced to 30, the worst result in its history; the Sarawak bloc Gabungan Parti Sarawak, GPS, held 23 and the Sabah bloc Gabungan Rakyat Sabah, GRS, six. For five days nobody could form a government, and the resolution rested where it had twice rested before — on the Yang di-Pertuan Agong, the King, then Sultan Abdullah of Pahang, who determined under his constitutional discretion that Anwar commanded the most members and appointed him, the third time in thirty-three months the monarchy had chosen who would govern the federation. To clear the threshold, Anwar's Reformasi coalition had to govern alongside UMNO — the United Malays National Organisation, BN's dominant party, and the party that had imprisoned him twice; PKR, Anwar's own Parti Keadilan Rakyat, born out of his 1998 jailing, was now to sit in cabinet with the machine that had jailed him. With the East Malaysian blocs and smaller allies the unity government commanded something in the order of 145 to 152 votes depending on the issue — a working majority assembled from coalitions that had spent a generation as enemies. The corpus's phrase for the formula is plain: it combined historically opposing coalitions under royal-constitutional sponsorship. Nobody had governed Malaysia this way before.

The arithmetic carried its price from the first day, and the price is the spine of this chapter: a government stitched from rivals cannot pursue any one rival's programme. PKR's Reformasi commitments — electoral reform, civil liberties, an independent judiciary, the dismantling of the old machine's habits — now had to survive a cabinet at which UMNO held the historically UMNO-coded portfolios of Defence and Rural Development, and at which the Deputy Prime Ministership went to UMNO president Ahmad Zahid Hamidi, himself under dozens of corruption charges. On 19 December the Dewan Rakyat affirmed the government by a confidence motion carried with 148 votes, the PN opposition declining to divide. The footing was secure. The question the next three and a half years would test was not whether the formula could hold a vote, but what a Reformasi movement could deliver from inside a government it had to share with the establishment it was built to oppose — and whether what it built would be Reformasi delivered, or Reformasi compromised. We begin where the government began: with the word it chose for itself.

Madani: The Word and the Work

On 27 January 2023, two months into office, Anwar delivered at the Securities Commission auditorium in Kuala Lumpur the framework address that gave the government its name. Ekonomi Madani — the Madani Economy, "Empowering the People" — set out six pillars under the acronym MADANI: keberlanjutan, sustainability; kesejahteraan, prosperity; daya cipta, innovation; hormat, respect; keyakinan, confidence; and ihsan, compassion. The vocabulary drew on Anwar's own decades-old project — the civilisational discourse of his 1996 book The Asian Renaissance and its argument for an Islamic-democratic synthesis grounded in Asian cultural pluralism — and paired the pillars with seven medium-term aspirations, among them top-thirty competitiveness and a lift in labour's share of national income from around 32 per cent toward 45. They were, the corpus is careful to note, aspirations rather than commitments: no year-by-year pathway was attached to most, and the baseline year was left unfixed.

What the framework did was as much political as economic, and here the contest begins early. Madani's deliberate breadth, the sources record, let four coalitions that agreed on little subscribe to one programme without abandoning their own creeds — PH, BN, GPS, and GRS could all sign the same vocabulary precisely because it committed none of them to a fixed ideology. To the government's defenders this was the right register for a multi-party arrangement, coalition-managerial language doing the work it must do. To its critics — drawn both from the PN opposition and from the Reformasi purists in Anwar's own base — the vagueness was the point: a framing capacious enough to mean anything meant, in policy terms, very little. The corpus sets the two readings down without choosing, and the disagreement is worth holding onto, because it will attach in sharper form to everything the government actually did. Its first instrument was a budget: Ismail Sabri's outgoing government had tabled a Budget 2023 just before the election, and Anwar, holding the Finance portfolio himself, re-tabled it entirely on 24 February 2023 at RM 388.1 billion, launching the Sumbangan Tunai Rahmah, STR — a cash-transfer scheme reaching some 8.7 million recipients — and committing, for the first time in a federal budget, to phased subsidy rationalisation, naming diesel and electricity as the first targets while carefully attaching no date. That single sentence of intent would become the hardest, costliest work of the premiership.

The Hard Reforms

To see why fuel subsidies were so dangerous to touch, one must see what they had become. Malaysia had run blanket fuel subsidies since the 1980s, and by the early 2020s they were among the largest items in the federal budget — the total bill reaching roughly RM 81 billion at the 2022 energy-price peak, the corpus records. The orthodox critique came from every quarter — the IMF, the World Bank, the domestic institutes: the blanket subsidy was regressive, since higher-income households and commercial fleets captured a disproportionate share; it fed cross-border smuggling; and it crowded out targeted spending. Every government for a generation — Najib's, Mahathir's second, Muhyiddin's, Ismail Sabri's — had named the problem and deferred it. Anwar's did not. On 10 June 2024 the blanket diesel subsidy was removed in Peninsular Malaysia, the pump price floating from a subsidised RM 2.15 a litre to around RM 3.35 — an immediate rise of some 56 per cent, as the sources state it. The reform was deliberately not total: Sabah and Sarawak kept their subsidy, in deference to East Malaysia's higher transport costs and the federal-state sensitivities that run through the relationship; commercial fleets kept subsidised access through a fleet-card scheme; and vulnerable households were meant to be cushioned by a targeted transfer, Budi Madani — Aid for Diesel under the Madani framework — paying eligible smallholders, fishermen, and lower-income groups around RM 200 a month. The Finance Ministry projected savings of roughly RM 4 billion a year, and Bank Negara Malaysia, the central bank, later judged the inflationary hit contained.

The reform's design was also its first controversy, along three lines the corpus keeps separate. The government's technocratic reading was that this was overdue discipline, well targeted, its inflation contained, its savings freed for higher priorities — the unavoidable cost of necessary work. The opposition reading, advanced most sharply by PN and PAS and by rural-Malay constituencies, was that it was regressive pressure falling hardest on those least able to bear it, the Budi Madani cushion too thin and too late — the portal opened only in late May 2024 and the first payments landed in late June, so the most exposed absorbed the full rise for weeks unaided. And a third reading the corpus marks as distinct: the hybrid, that what was built — partial geography, commercial carve-out, targeted cushion — was neither the textbook reform the orthodox wanted nor the untouchable status quo, but the maximum achievable inside the unity government's constraints. Each reading points to something real; none is allowed to absorb the others.

Diesel was the breach in the wall. The far larger prize, and the far larger danger, was RON95 — the petrol the Malaysian middle class puts in its cars, the single biggest fuel subsidy at something like RM 30 billion a year — and here the government's nerve met its limit. Anwar committed to RON95 reform repeatedly, in the Budget 2024 and Budget 2025 speeches and in answer after answer, and reform after reform was deferred. The stated obstacles were three: the targeting database was incomplete; the eligibility design was unsettled; and, above all, the politics, the August 2023 state elections having shown a Malay electorate drifting toward the opposition that a botched petrol shock might send further still. The database the whole architecture waited on was Pangkalan Data Utama, PADU — a central registry launched in January 2024 by Economy Minister Rafizi Ramli to hold the income and household data that means-tested targeting would run on, contentious from the start over data privacy and inter-agency data-sharing, its registration below target and its deadlines slipping through 2024 and into 2025, so that a database's enrolment rate had become, in effect, the pace-setter of national fiscal policy. By the close of the mid-tenure period the deferred RON95 reform stood as the government's most prominent unkept major promise, and the clearest illustration of the bind this chapter describes: sound on the technocrats' terms, ruinous on the politicians', and the government could not be both at once.

Alongside the fuel reforms ran a quieter consolidation of the revenue base. The Sales and Service Tax — the SST, which the first Pakatan Harapan government had restored after abolishing the broader Goods and Services Tax, the GST, in 2018 — was broadened in stages: the service-tax rate on selected categories rose from 6 to 8 per cent from 1 March 2024, with further base-broadening waves following in July 2025 and, by the dates the corpus carries, March 2026. The cumulative effect reopened a durable argument — whether to simply bring the GST back. The orthodox consensus, from the IMF to the domestic institutes, held that a properly designed GST with input-tax credits would raise more revenue with less distortion than an ever-broadening SST; the government's answer was political memory, anti-GST sentiment having helped sink the old government in 2018, and the SST could be stretched to comparable yield, a GST revival left to "when conditions allow" — a phrase the commentariat read as code for "after the election." The opposition called the broadening a "stealth GST," all the burden and none of the input-credit relief; the Reformasi purists, from the other side, complained that a well-built GST would have been the more progressive instrument, and the government had chosen the less progressive path for safety's sake. Beneath both the subsidy and tax reforms sat a piece of architecture meant to outlast any one budget: the Public Finance and Fiscal Responsibility Act of 2023, the first formal fiscal-rule statute in Malaysian history, setting a deficit anchor descending toward 3.0 per cent of GDP and debt ceilings the corpus puts around 60 and 65 per cent. The budget cycle was built to its logic, the envelope climbing across the three Madani budgets to the order of RM 430 to 440 billion by Budget 2026 as the sources state it, along an explicit deficit glidepath from 4.3 toward 3.8 per cent of GDP. And in the year-three period the long-deferred reform finally moved: RON95 targeted rationalisation entered implementation, the subsidised RM 2.05 base preserved only for eligible Malaysians verified through PADU while everyone else paid a market-reflective price, for a saving the corpus puts at around RM 8 billion a year — reported, as with every figure in these final stretches, as the source states it and not as settled fact. The fiscal architecture the November 2022 government had promised was, by the chronicle's end, substantially in place.

The Tools of Delivery and the Map of Discontent

If subsidy reform was the hardest act, the machinery built to soften it mattered nearly as much. The scattered cash transfers were drawn together over 2025 into a single PADU-anchored digital-welfare system, MyKasih, enrolling by early 2026, the sources state, somewhere between five and eight million households — one of the largest such systems in Southeast Asia by reach. Its supporters pointed to the leakage it closed against a pipeline that had bled to lost cheques and fraud; its critics, to who fell through it — the elderly without smartphones, the rural districts of East Malaysia with thin connectivity, the households whose eligibility was disputed on the day a subsidy was removed. The architecture that made the welfare state more efficient made it, for those it could not reach, harder to enter.

The political ground beneath all this was shifting, and the state elections of 12 August 2023 were where the government first read the new map. Six assemblies went to the polls at once — Selangor, Negeri Sembilan, and Penang, held by the PH-BN side; Kedah, Kelantan, and Terengganu, held by PN — and the headline was a draw, each side keeping the states it governed. But the movement underneath was the story: in Selangor, the federation's richest state, PH-BN held on with a sharply reduced majority while PN surged from a single seat in 2018 to twenty-two, concentrated in the Malay-majority districts, and in Kelantan and Terengganu PAS swept all but a handful. The interpretation converged across the analytical spectrum — Bridget Welsh's "green wave," after PAS's colour, and James Chin's "structural challenge": not the loss of the parliamentary majority, untouched, but the visible erosion of the Malay base beneath the coalition's PH and BN components, which the Merdeka Center put down to cost-of-living strain, identity-and-religion appeals casting the unity government as compromised, rural-Malay frustration over the unresolved Najib question, and a generational shift among younger Malay voters. The by-elections that followed confirmed the pattern more than they disturbed it: at Sungai Bakap, a Penang Malay-belt seat, on 6 July 2024 the PN-PAS candidate not only held but expanded the 2023 majority — evidence both sides accepted that PH-BN coordination did not automatically translate into combined votes in Malay-majority seats — while Nenggiri and Mahkota went to BN-UMNO on factors analysts read as seat-specific. The composite reading was threefold and stable: the Malay consolidation behind PN was now a structural feature, not a one-cycle event; the federal arithmetic translated unevenly to the seat level; and the cycle ahead — Sabah due around the end of 2025 and Sarawak by the end of 2026 by the timing the corpus carries, with the Sarawak bloc GPS expected to keep its supermajority — would carry the same problem into the general election the constitution placed in 2027. That problem stayed fixed throughout: a government able to point to a completed fiscal architecture, a flagship economic zone, and a regional chairmanship discharged still had to convert that record into Malay votes it was visibly losing.

Inside the coalition, the strain showed first as a quarrel within the reform coalition's own flagship party. On 11 April 2024 the Economy Minister, Rafizi Ramli — PKR deputy president and longtime Anwar ally — delivered an address titled "Performance Indicator: How Successful is the Madani Government," the most substantial internal critique of the government's delivery, calibrated across six dimensions to credit successes while flagging under-performance on institutional reform and the unfinished PADU database. The corpus keeps the two readings it drew: principled internal accountability in the Reformasi tradition, by Rafizi's supporters; and positioning for the party's deputy-presidency contest, by Anwar loyalists who heard destabilisation. The sequence that followed — the speech, his decision on 28 April not to seek re-election, then his departure from the cabinet — read less like a resignation than a coordinated exit, and the deputy presidency passed in time to Nurul Izzah Anwar, the Prime Minister's daughter, drawing the charge of dynasty alongside the fact of renewal. The reshuffle of 14 May 2024 that resolved it brought Amir Hamzah Azizan, the Treasury Secretary-General, directly into the cabinet as Minister of Finance II — a constitutionally unusual move, crossing the civil-service neutrality convention, that placed at the head of fiscal policy a man with three decades of sovereign-investment and energy-sector experience while folding the Economy Ministry Rafizi had left into the expanded Finance Ministry. The reading across the spectrum agreed on one thing even where it disagreed on all else — that the government had reached an operational, stabilising phase rather than one of fresh reform ambition.

No single episode tested the reform coalition's identity more sharply than the partial pardon of Najib Razak, on 29 January 2024, when the Federal Territories Pardons Board reduced his sentence in the 1MDB-linked SRC International case from twelve years to six and his fine from RM 210 million to RM 50 million. An earlier chapter carries the pardon and the disputed "Royal Addendum" that followed it in full; here it is enough to record the three irreconcilable readings the corpus sets down by name — the government's, that the board had acted within its authority and the executive could not substitute its judgment; the reform purists' and civil society's, that it was a Reformasi betrayal, the undoing of the anti-corruption agenda the 2018 movement was built on; and the pro-Najib UMNO stream's, that it was a bounded vindication of a prosecution they held excessive — and to note that the addendum litigation remained unresolved at the close of the chronicle, with Najib's own status, still in Kajang or moved to house arrest, among the things the sources could not confirm, and so not asserted here. Nor did the pardon stand alone: Anwar's own Deputy Prime Minister, Ahmad Zahid Hamidi, carried into office a contested "discharge not amounting to acquittal" from September 2023 — the fact that most crystallised the unease of the whole arrangement, an anti-corruption coalition governing alongside men it had once prosecuted, or might yet. Whether the system was operating with genuine independence produced, the corpus concludes, no single answer.

Governing at home meant, through these same years, governing in a turbulent world, and the foreign record — treated at length in later chapters — became part of the domestic argument. Malaysia took the rotating ASEAN chairmanship for 2025, its fourth, and accepted BRICS partner-country status, framed within its long doctrine of non-alignment; then, on 2 April 2025, the returning Trump administration's "Liberation Day" tariffs landed a 24 per cent reciprocal levy on Malaysian exports, by the rate the corpus carries, a mid-chairmanship shock to the semiconductor base that a reported August 2025 framework eased toward the high teens. The Johor–Singapore Special Economic Zone signed on 7 January 2025 and the data-centre boom it anchored belong to a later chapter too; here they enter the record only as part of what the government would carry into the election — a flagship bilateral instrument and a visible engine of growth, set against a Malay electorate it was still losing.

Reformasi Delivered, or Reformasi Compromised

By the close of the chronicle, in the middle of 2026, the unity government had governed for three and a half years and had not fragmented. It had survived a state-election cycle that exposed a widening fault-line, a resignation from inside its flagship party, a ringgit trough near the 1998 lows before a recovery into a band the corpus puts around RM 4.20 to 4.50, a charged pardon, and a mid-chairmanship tariff shock; it had completed, or substantially completed, the fiscal architecture it had promised — diesel rationalisation, SST broadening, the Fiscal Responsibility Act, the digital-welfare consolidation, and at last the deferred RON95 reform. By every measure of operational survival, the improbable formula had worked. And yet the question the corpus poses at the outset is not whether the government survived but what it was — and on that the sources offer not a verdict but two, set deliberately against each other and attributed to their holders. This is the chapter's contested core, and the discipline, here as in the corpus, is to present each by name and let neither absorb the other.

The first frame is Reformasi delivered. On this reading — advanced by the Prime Minister's office, the reform coalition's own commentary, and parts of the technocratic-reform community — the government did, within binding constraints, what the movement had promised. It executed fiscal reforms that four prior governments had named and dodged; it passed Malaysia's first fiscal-rule statute and put the deficit on a credible glidepath; it built one of the region's largest targeted-welfare systems; it signed a flagship cross-border economic zone, chaired a regional bloc through a year of acute external pressure, and moved the criminal-justice and institutional-reform processes forward where they had stalled. On this account the unity-government formula is itself the achievement: stable, reforming government in a country where no single coalition could command a majority, held together by a constitutional monarchy doing real and not merely ceremonial work.

The second frame is Reformasi compromised, and it comes from two directions that diverge on everything except the diagnosis. From within the reform tradition — the PKR technocratic critique Rafizi gave voice to, the civil-society reading of groups like Bersih and the human-rights monitors — the charge is that the pace fell short of the 2018 promise: that company-and-financing reform, electoral-reform legislation, and the repeal of the old coercive laws moved slowly or not at all, that the Official Secrets Act and the Sedition Act remained in use, and that the partial pardon, the deferred petrol reform, and the limited institutional change together repurposed Reformasi's political capital for the management of a coalition. From the opposite direction — the PN-PAS-Bersatu critique — the charge is inverse in content but identical in shape: that the government neglected the Bumiputera framework and the Islamic-policy advancement its Malay-Muslim constituents wanted, Madani's vocabulary being cosmetic cover for the absence of conservative substance. The two strands prescribe opposite remedies; they converge, the corpus notes, on a single conclusion — that delivery fell short of ambition, whichever ambition one measured it against.

Between the two frames the corpus records a third — the structural account of the academic and think-tank analysts, on which the record was less a matter of will than of constraint: the hung-parliament arithmetic of 2022 forced the formula regardless of any party's preference, the ethnoreligious cleavage imposed an electoral logic any government would have had to accommodate, and the Madani government was an attempt to govern within those constraints rather than against them — its significance not the substance of its policy output but the demonstration that the formula could operate without breaking, its durability hostage to a cleavage the cabinet could not itself engineer. The corpus does not choose. It locates the empirical record — on reform, on subsidy, on the rule of law alike — between the readings: more institutional reform than under the governments before, less than the 2018 movement had envisaged; targeting progressive in design but with gaps in implementation; a criminal-justice independence that produced no single answer. And it points, finally, to the test that would settle what argument could not. The seventeenth general election, due by 2027, would be the binding trial of whether the reform generation's three and a half years had converted into a mandate for a second term — or whether the Malay-electorate consolidation the August 2023 elections had mapped would carry the opposition to power and reopen the post-2018 cycle of realignment this whole stretch of the book has traced. The year-three record, in the Madani vocabulary the government chose for itself, was an ihsan — a careful, structurally considered, incomplete-but-concrete contribution to the governance of post-2018 Malaysia. Whether it amounted to Reformasi delivered or Reformasi compromised was, by the corpus's own insistence, a question for an election not yet held and a verdict this chapter cannot pronounce.

What can be said is what the reform generation did with power, which is the question this chapter set out to answer. It governed — improbably, under sponsorship of a crown it had not expected to need, alongside enemies it had not expected to govern with, through reforms it had long demanded and now had to pay the political price of enacting. The economic architecture beneath that governing — the sovereign funds and the national oil company that underwrote the fiscal room the Madani budgets spent, and the federal-state resource tensions that strained it — is the subject of the chapter that follows; for the machinery of the bargain runs deeper than any one government's programme, and to understand what the Madani years could and could not afford, one has to descend into the sovereign machine that has financed the Malaysian state since 1974.