Pakistan’s economy and the balance-of-payments trap

What this chapter is for

This chapter explains why Pakistan’s economy keeps stopping. It covers what the economy is made of, what it buys and sells abroad, how it pays for the difference, what the state owes and collects, and why the country has gone to the IMF about two dozen times. By the end you should be able to explain the trap in plain words, give the current figures for each part of it with their year and source, judge CPEC, privatisation and the mineral deposits on the record rather than the promise, and say what would have to change for the trap to end.

The chapter is long, so here is its map. It begins with the argument and the history. The first part then covers the real economy, meaning the three sectors, the land, industry and energy, and the people who work and the people who are poor. The second part covers money and the outside world, meaning prices and the State Bank, the rupee, trade, the newer exports and remittances. The third part covers the state’s own finances, meaning debt, tax and the budget, the enterprises it owns, and the IMF. The last part covers what is being built, meaning CPEC and the minerals, and ends with where things stand in October 2026.

The economy is among the most-asked subjects in the paper. The 2023 question tells you what the examiner wants most. It asked candidates to Critically discuss why Pakistan’s economic growth continues to be “Balance-of-Payments” constrained with a focus on the long-term challenge to Pakistan’s export performance under pressure.¹ Learn the mechanism that question names and you can answer the debt question, the export question, the stability question and the IMF question from the same material.

Before you start

The words you need

The mechanism, drawn out

The one thing to understand is a loop, not a list, and Figure 1 draws it. The economy grows, so people and firms buy more. Much of what Pakistan needs for growth comes from abroad, including fuel, machinery and industrial inputs, so imports rise quickly. Exports do not rise as fast, because Pakistan’s factories make a narrow range of products and cannot expand quickly. So the country needs more foreign currency than it earns, and its reserves fall. To protect the currency, the government slows the economy, usually with an IMF loan and its conditions.

Imports fall, the crisis eases, growth resumes, and the loop begins again.

Growth is not being interrupted by bad luck. Growth is what brings on the interruption, and that is why this is called a trap rather than a downturn.

The story in plain words

1. Pakistan buys far more from the rest of the world than it sells to it.
2. What it does sell is mostly one thing, textiles, and mostly in a low-value form.
3. The difference is covered mainly by money sent home by Pakistanis working abroad, and by borrowing.
4. When the gap grows too large, the country runs short of foreign currency and turns to the International Monetary Fund for a loan.
5. That loan comes with conditions, such as collecting more tax, cutting subsidies, raising energy prices and letting the currency fall.
6. Those conditions steady the economy but also slow it, and they have been agreed about two dozen times since 1958.
7. The cost of steadying it falls on ordinary households, so poverty can rise in a year when the economy grows.
8. One new thing is now being sold abroad in quantity, computer services and freelance work, and it is growing fast though still small.
9. Nothing changes for good unless what Pakistan sells abroad changes, and that is the argument of this chapter.

1. Growth itself brings on the crisis

Most candidates write this subject as a list of problems, such as debt, inflation, a narrow tax base and corruption. Then they recommend “export promotion” and “fiscal discipline”, which are labels rather than proposals. A list cannot explain why the same crisis keeps coming back. A mechanism can.

The argument of this chapter is that Pakistan does not have a growth problem. It has an economy in which growth itself brings on the crisis. When output rises, imports rise at once, because Pakistan does not make most of what growth needs, above all fuel and machinery. Exports do not rise with them, because they rest on a few low-value products that cannot expand quickly even when buyers exist. So the current account moves into deficit, reserves fall, and the state must slow the economy to protect the currency. The recovery is ended by the thing that caused it.

That is what “balance-of-payments constrained” means, and it is why Pakistan has repeatedly been steadied without becoming steady. The figures for 2025-26 show the pattern again in miniature. Growth rose to 3.70 per cent, the fastest in four years, and in the same year goods imports rose by 8 per cent while goods exports fell by 6 per cent.²³ The current account, which had been in surplus the year before, slipped back into a small deficit.⁴

So the claim worth defending is that the binding constraint is what Pakistan sells, not what it owes. Debt is the symptom that appears when exports cannot pay for imports. Every IMF programme treats that symptom, which is why each one works and why each has to be repeated. Judge any economic proposal by whether it changes the export base, and you will be arguing about the cause rather than listing the effects.

2. How the economy reached this point

Pakistan began with almost no industry. The areas that became Pakistan grew cotton and jute and sent them elsewhere to be made into cloth, so the new country exported raw material and imported the finished article. That shape is what the rest of this chapter is about. Its first IMF arrangement, a standby for SDR 25 million signed on 8 December 1958, was never drawn on.⁵ Figure 2 sets out the turns that followed.

The 1960s under Ayub Khan were the fastest sustained growth in the country’s history, at about 5.82 per cent a year from 1958 to 1969.⁶ They proved Pakistan could grow quickly, and they concentrated that growth in few hands, which became a political grievance. In the 1970s Zulfikar Ali Bhutto nationalised banks, insurance and large industry, and four IMF standby arrangements were taken between 1972 and 1978.⁶ Much of what the state still owns and loses money on dates from then. In the 1980s growth averaged about 6.6 per cent under Zia ul Haq, carried by American aid during the Afghan war and by wages sent home from the Gulf.⁶ Both were somebody else’s money, and neither changed what Pakistan sold.

The one true default came in 1998. After the nuclear tests of 28 May, sanctions followed. Reserves were near $1 billion, against foreign currency deposits in Pakistani banks of about $11 billion, so the government froze those accounts to stop the money leaving, and the country went into technical default on its external debt.⁷ There was no cushion, so the state reached into its own citizens’ deposits. A country with a strong export base could have absorbed the shock, and Pakistan could not, for the same reason as now. If you use one date from this history in an answer, use 1998.

The pattern repeated in the 2000s, when inflows after 11 September 2001 closed the external gap for a while, and again with CPEC from 2015, which ended the long power cuts.⁸ In 2022 floods put about a tenth of the country under water and affected 33 million people, causing $14.9 billion of damage and a further $15.2 billion of economic losses.¹⁰¹¹ In 2023 the country came close to default and was carried through by an IMF standby, and in September 2024 it began a $7 billion programme under which it has stabilised.¹²¹³

Read the history for one lesson. Every turn fixed a real constraint and left the export base where it was. Ayub built industry, Zia’s decade had foreign money, CPEC removed the power cuts, and the 2024 programme rebuilt the reserves. Each was a genuine achievement, and none of them changed what Pakistan sells to the world. That is why the same crisis returns.

3. What the economy is made of

An economy is the sum of what a country grows, makes and does, and Pakistan’s three parts behave very differently. Agriculture is the land, meaning wheat, rice, cotton, sugarcane and livestock. Industry is manufacturing, construction, mining and electricity, and the part watched most closely is large-scale manufacturing, the big registered factories, measured every month. Services are everything else, from shops and transport to banking, telecommunications, government and the fast-growing export of software. Services are the largest part of Pakistan’s output.

In 2025-26 the economy grew by 3.70 per cent, up from 3.18 per cent the year before, and reached about $452.1 billion, with income per person of $1,901.² Agriculture grew 2.89 per cent, industry 3.51 per cent and services 4.09 per cent.² That is a recovery from the crisis years, and it is not fast enough. The population grows by over two per cent a year, so growth of 3.7 per cent leaves little more for each person. That is why “growth returned” and “people felt better off” are not the same statement.

The shape matters more than the size. Most of Pakistan’s output is services, but most of its exports are goods, and mainly one kind of goods. Most of its workers are in farming and in informal services, which are the hardest activities for a tax office to see. So the structure of the economy explains both the export problem and the revenue problem at once. A candidate who opens with structure rather than with a list of troubles has already framed the answer better than most.

4. Agriculture and the land

Agriculture is 23.4 per cent of what Pakistan produces and employs 33.1 per cent of the people who work.² Hold those two shares together, because the gap between them is the point. A third of the workforce produces under a quarter of the output, so each farm worker produces much less than a worker elsewhere, and most Pakistani workers are in the least productive part of the economy. Moving people out of low-productivity work into better-paid work is what development is, and it has been slow here.

The Economic Survey tracks five major crops, which are cotton, rice, sugarcane, maize and wheat. Together they grew 0.65 per cent in 2025-26, after shrinking by 13.19 per cent the year before.² A swing that size in one part of the economy is enough to move national growth by a point on its own, and it is why Pakistan’s growth is less steady than a services economy’s. In 2025-26 wheat rose 4.3 per cent to 29.61 million tonnes, sugarcane 6.2 per cent to 89.45 million tonnes and rice 2.8 per cent to 9.99 million tonnes, while maize fell 2.7 per cent to 8.79 million tonnes after floods and cotton fell 0.5 per cent to 7.05 million bales.²

The cotton line looks the smallest and matters most. Cotton is the raw material of the textile industry, and textiles are about three-fifths of everything Pakistan sells abroad. So a weak cotton crop becomes weak exports about a year later, and the mills then buy cotton from abroad, which raises imports at the same time. One bad season pushes both sides of the trade gap the wrong way at once.

Notice also where cotton lost ground. Its sown area shrank 1.5 per cent to 2.01 million hectares, in a year when sugarcane and wheat grew.² Farmers plant whatever pays best, so the country’s export crop is losing land to crops grown for the home market.

Pakistani farming is irrigated farming, fed by the Indus and its tributaries, which depend on snowmelt and the monsoon, so the climate reaches it directly. The 2022 floods did $3.7 billion of damage to agriculture and livestock alone. The World Bank put the loss to the economy at about 2.2 per cent of that year’s GDP and projected that poverty could rise by 3.7 to 4.0 percentage points, pushing 8.4 to 9.1 million more people below the poverty line.¹¹ That is the convergence the 2026 paper asked about in one event. A climate shock hit the sector that employs a third of the workforce, pushed millions into poverty, raised the import bill for food and forced borrowing to rebuild.

One more fact belongs here. Agricultural income is largely untaxed, and agriculture is a provincial subject under the Constitution, so the federal government cannot simply tax it, and the provincial assemblies contain many of the people who would pay. All four provinces passed laws on agricultural income tax between November 2024 and February 2025, while the IMF was reviewing the programme, and the federalism chapter follows what they collect.¹⁴ For this chapter the point is that a quarter of the economy has long paid little direct tax, which is a governance problem dressed as an accounting one.

5. Industry, energy and the circular debt

From too little power to too little payment

Industry grew 3.51 per cent in 2025-26, with large-scale manufacturing up 6.11 per cent and construction up 5.73 per cent. Electricity, gas and water supply, however, shrank by 10.63 per cent.² Power generation fell in a year when factories produced more. The reason is not a shortage of power stations, and it shows how Pakistan’s energy problem has changed.

Until about 2015 the constraint on industry was electricity itself. Power cuts ran to about 18 hours a day in the worst periods, and factories could not run even when they had orders.¹⁵ That was fixed, and an answer should give Pakistan credit for it. Completed CPEC energy projects worth over $15 billion added 9,504 megawatts from coal, wind, water and solar power.⁹

Then notice what did not happen. The electricity constraint was removed and the export structure did not change. That is the strongest single piece of evidence for this chapter’s argument, and it comes from Pakistan’s own success. Electricity was necessary, and it was not enough.

The problem now is that the power sector does not collect enough money to pay for the electricity it produces, and the shortfall piles up as the circular debt. Figure 3 traces the chain. Power is generated by companies, many of them private, under contracts that guarantee them payment. It is sold to consumers by distribution companies, most of them owned by the state, which lose much of it to old wires, theft and bills never paid. So they cannot pay the generators in full, the government promises to cover the difference, and the unpaid promises accumulate. It is called circular because each party in the chain is owed by the one in front and owes the one behind.

The figures, and who pays them

The circular debt stood at Rs 1.614 trillion in June 2025. The government told the IMF it would add nothing in the following year and hold the stock at that level by June 2026.¹⁶¹⁷ It rose instead, to Rs 1.837 trillion by February 2026 and Rs 1.924 trillion by the end of May. By 30 June 2026 it stood at Rs 1.835 trillion, Rs 221 billion above the agreed ceiling.¹⁶¹⁷ Officials blamed mainly about Rs 200 billion of unpaid dues from K-Electric, the Karachi utility, and the weak performance of several distribution companies.¹⁷

To clear the older arrears, the government arranged Rs 1.225 trillion of loans from 18 commercial banks in September 2025, repayable over six years in 24 quarterly instalments. They are paid for by a surcharge of Rs 3.23 on every unit of electricity consumers use.¹⁶ Read that slowly, because it is the argument of the chapter in miniature. The cost of losses and unpaid bills in the distribution system is being collected from the households and factories that do pay.

Three consequences follow. Electricity becomes dearer for exporters, at exactly the moment this chapter says export competitiveness is the binding constraint, and part of that price is other people’s unpaid bills. Money lent by banks to clear the power sector’s arrears is money not lent to a business that might export something. The underlying fault is administrative, not economic, since meters, billing, collection and enforcement are tasks of administration. A state that struggles to collect a tax also struggles to collect an electricity bill, which is the same weakness showing up in a different ledger.

So if a question asks about energy, do not describe the load-shedding of 2012 as current. Say the shortage of power was largely solved and give the megawatts. Then say the problem moved from supply to price and payment, give the circular debt with its date, and name the remedy in specific terms, such as cutting distribution losses, metering every connection and enforcing payment.

6. Work, wages and poverty

Who works

The Pakistan Bureau of Statistics records a labour force participation rate of 46.3 per cent, unemployment of 7.1 per cent and an average monthly wage of Rs 39,042.¹⁸ The first number matters most and is the one candidates rarely use. Participation of 46 per cent means fewer than half the people old enough to work are in the labour force at all. The unemployment rate counts only those looking for work, so a country where most adults are outside the labour market has a problem the unemployment rate does not describe.

A narrow working population carries a large dependent one. That produces the conditions this chapter keeps meeting, which are low savings, low tax revenue and pressure on the state to subsidise consumption. It also means the largest single opportunity is bringing more people into paid work, which is why questions on women’s work and on education are economic questions too. Much of that work is informal, with no registration, no written contract and no records. Informal work is not idle work. It is most of the retail trade, transport and construction of the country, and it is invisible to the tax office, to labour law and to the statistics.

How to handle the poverty number

Poverty figures for Pakistan seem to contradict each other, and a candidate who quotes one without its definition can be corrected in a line. A poverty rate counts people below a line, and the answer depends on where the line is drawn. On the national poverty line, poverty rose from 21.9 per cent in 2018-19 to 28.9 per cent in 2024-25.¹⁹ On the World Bank’s extreme poverty line of $3.00 a day, 23.0 per cent of Pakistanis were poor in 2024-25, and on its line for lower-middle-income countries of $4.20 a day, 47.9 per cent were.²⁰ All three are correct, so always name the line and the year.

The level is arguable, and the direction is not. Poverty on the national measure rose by seven points between the two surveys, and the later survey fell in a year of recovery.¹⁹² Three reasons explain it, all from this chapter. The population grows fast, so 3.7 per cent growth leaves little per person. The adjustment fell on consumption, since energy priced at cost, taxes on what people buy and withdrawn subsidies all reach a household directly. The prices that rose did not come back down, and in 2026 households faced a second rise on top of the first, with inflation back above 10 per cent.²¹

The state’s main answer is the Benazir Income Support Programme, whose first budget year was 2008-09. It pays cash directly to poor women, and its Kafaalat payment reached 10.20 million of them in 2025-26.¹⁹ The 2026-27 budget raised its allocation to Rs 838 billion.²²

Cash paid straight to a household is the least leaky help a weak state can give, because it needs no working school, clinic or ration shop to reach the person. It is also compensation, not development. A transfer keeps a family above a line, and it does not create the job that would keep them there without help. The chapter on poverty and social protection explains the programme in full.

Write this part as a policy sequence with a cost attached. The adjustment was necessary, its cost landed on households with the least room, the state’s answer compensates without changing their position, and the remedy is whatever raises paid employment. That is criticism, it is specific, and it proposes something, which is the register of a future officer.

7. Money and prices

The State Bank and the policy rate

The State Bank of Pakistan is the central bank. It issues the currency, holds the country’s foreign exchange reserves, regulates the banks and sets the policy rate, the interest rate that steers what banks charge each other and so what everyone pays to borrow. Raising the rate makes borrowing dearer, which slows spending and cools prices, and cutting it does the opposite. The Bank raises rates against inflation and cuts them to support growth, and it cannot fully do both at once.

The rate had come down from the crisis years to 11.50 per cent, and on 14 September 2026 the Monetary Policy Committee held it there by seven votes of ten.²³ It held rather than cut because inflation had come back. Prices rose 9.2 per cent in the year to July 2026, 11.1 per cent to August and 10.3 per cent to September, so prices in the first quarter of 2026-27 averaged 10.2 per cent higher than a year before, against 4.3 per cent a year earlier.²³²¹ Figure 4 sets those months against the State Bank’s target band of 5 to 7 per cent.

The Bank names two causes. Food prices rose, led by wheat, and energy stayed dear because the conflict in the Middle East kept oil prices high, which raised transport costs.²³ Read this against the stabilisation, not instead of it. Inflation had fallen from the tens of per cent in the crisis years to single figures, and that fall was real. The 2026 figures show how exposed the recovery is to an import it cannot control. Pakistan buys its oil abroad, so a war it is not fighting sets its prices.

Hold one more distinction. A falling inflation rate means prices are rising more slowly, not that prices have fallen. A household that saw flour and electricity double between 2022 and 2024 never got those prices back, and it now faces another rise on top. Say that in an answer on the cost of living and you are describing what people feel rather than what an index says.

Reserves

Foreign exchange reserves are the foreign currency the country holds to pay for imports and meet debts falling due. In mid-September 2026 the State Bank’s own reserves rose to about $21.4 billion, helped by a $3 billion Eurobond sold that month and by the Bank’s purchases of dollars in the market.²³ The Bank says it bought $20 billion from the market over three years to rebuild them, and it expects them to approach three months of imports by June 2027.²⁴²³ The figure is watched because it answers a blunt question. If no new money came in, how long could the country keep paying for what it buys abroad?

8. The rupee, and the gap that matters more than the rate

A section on money that leaves out the rupee has left out the price that sets every other price. Pakistan imports its oil, its machinery and much of its food, and it owes a large part of its debt in dollars. So when the rupee falls, the import bill rises and inflation rises with it, and the rupee cost of servicing the same dollar debt also rises, although nobody has borrowed a further rupee. The rupee was near 105 to the dollar in 2017 and about 150 by 2019. It fell hard through 2022 and 2023 to an all-time low of 307.75 in September 2023, and stood at about 277.5 in September 2026, about 2 per cent stronger than a year earlier.²⁵

There are two rates, and the difference between them is the lesson. The interbank rate is the rate banks deal at, under the State Bank’s supervision. The open market rate is what licensed exchange companies charge a person buying dollars over a counter. In a working market the two sit close together. In 2023 they came apart, with the dollar at Rs 340 in the open market while the interbank rate was held at Rs 307.¹²

A gap like that is an offer, not a statistic. An overseas Pakistani sending money home through a bank was paid the official rate, while sending it through an informal hundi or hawala dealer paid the higher one. So the money left the banking system, and with it went foreign currency the State Bank could have counted as reserves. The Exchange Companies Association of Pakistan put the cost at about $4 billion of remittances lost in FY23, the year Pakistan reached the edge of default.¹²

What closed the gap was enforcement rather than a rate decision. A crackdown in 2023 on currency smuggling and informal transfer networks narrowed it, and recorded remittances rose from $27.33 billion in FY23 to $30.25 billion in FY24.¹² Much of that increase was not new money. It was the same money changing route, from the informal channel back into the banks, and an answer that reads it as a surge in earnings has misread it.

The grey market reappeared during 2024 at about Rs 284 against an open market rate of Rs 280, drawing an estimated $500 million a month.¹² So the number to watch is the gap, not the rate. A steady rupee with a wide gap is worse than a weaker rupee with none, because the first is losing the country its reserves while looking calm.

9. What Pakistan buys and sells

In 2025-26 Pakistan’s goods exports were $30.13 billion and its goods imports $69.60 billion, a gap of $39.47 billion.³ Exports fell by 6 per cent on the year before, while imports rose by 8 per cent. Imports were about 2.3 times exports, which is the whole problem stated as a ratio. The pattern carried into the new year. In July and August 2026 exports rose 7 per cent to $5.46 billion, but imports rose 13 per cent to $12.58 billion, and the trade gap widened by 18 per cent.²⁶

These are customs figures from the Bureau of Statistics. The State Bank counts trade a little differently in the balance of payments, by when payment is made, and on its basis the goods gap for 2025-26 was $33.6 billion.⁴ Both are correct, so say which you are using.

Now look inside the export total, because this is the part most scripts never reach. Figure 5 shows it. Of the 2025-26 exports, textiles and clothing earned $17.93 billion and food $5.02 billion, and everything else together earned about $7.2 billion.²⁷²⁸³ Food fell by almost 30 per cent in one year, mainly because rice earnings dropped from $3.35 billion to $2.29 billion.²⁸

Textiles alone were 59.5 per cent of everything Pakistan sold abroad in 2025-26, and their share rose only because other exports fell, since textiles themselves grew by a quarter of one per cent.²⁷³ That single fact answers the 2023 question better than any general talk of competitiveness. Pakistan’s export earnings rest on one sector. So it cannot raise those earnings quickly, cannot absorb a bad cotton year, and cannot raise prices, because the buyer can go to Bangladesh or Vietnam next season. Concentration is the long-term challenge to export performance, and naming it as concentration rather than as “low exports” turns description into analysis.

10. The newer exports: technology and services

If the trap ends when the mix of exports changes, the honest test of the argument is whether anything new is being sold. Something is. Exports of information technology and related services reached a record $4.6 billion in 2025-26, up 21 per cent on the year before, and June 2026 alone brought in $416 million, the best month on record.²⁹ Freelancers, people selling their own work to foreign clients, earned $1.76 billion on State Bank figures, up 78 per cent. Of that, $1.16 billion came from computer work and $592 million from other kinds of freelancing, which nearly tripled. Industry estimates put Pakistan’s freelancers at nearly three million, the fourth-largest freelance workforce in the world.³⁰

This kind of export differs from textiles in the way that matters most for the trap. A textile mill must import machinery, dyes and in a bad year the cotton itself, so more textile exports pull in more imports behind them. A software service exports a person’s time, so the foreign currency arrives without a matching import bill. It does not depend on electricity prices, water or a cotton crop either. That is exactly the property the balance-of-payments trap needs.

Now be honest about the size, because a candidate who stops at the good news will be marked down. IT exports of $4.6 billion sit beside goods exports of $30.13 billion and textiles of $17.93 billion, as Figure 6 shows.³²⁷²⁹ Technology earns about 15 per cent as much as all goods and about a quarter as much as textiles. It is not yet the answer, and it is the first real evidence that the answer is possible.

It also carries risks the textile sector does not. Much freelance earning is individual and informal, so it is easily undercounted and easily moved. It depends on internet access and on being able to receive foreign payments, and both are policy choices, so a restriction on the internet reaches this sector faster than any tariff. It is concentrated in a small, educated, English-speaking group, which is why the education chapter and this one make the same argument from two ends. The 2026-27 budget extended the low tax rate on IT exports to 2029.²²

The government’s own plan sets a test this sector can be measured by. Uraan Pakistan, the National Economic Transformation Plan for 2024 to 2029, aims to double annual exports to about $60 billion by 2029, counting services as well as goods.³¹ If technology and services carry a large part of that increase, the structure will have changed. If the increase is textiles, Pakistan will have built a bigger version of the same problem. A proposal with an instrument and an actor might run like this. Keep foreign payment channels open and simple for individual earners, treat internet access as export infrastructure rather than as a security question, and put training money where the sector already grows at 21 per cent a year.

11. Remittances and the current account

A country, like a household, has money coming in and money going out. The current account counts exports and other earnings coming in, and imports and other payments going out. If more goes out than comes in, the country must cover the difference by borrowing or by spending reserves. Pakistan’s goods trade is heavily in deficit, so the question that decides whether it can pay its way in any year is what else comes in.

The answer is money sent home by Pakistanis working abroad. Over 15 million Pakistanis have registered for work overseas since 1971, and Saudi Arabia alone accounts for more than half of them, with the United Arab Emirates second.³² What they send back is the largest inflow in the external accounts, larger than foreign investment. Remittances reached a record $41.6 billion in 2025-26, up 8.6 per cent from $38.3 billion the year before.³⁴ The largest sources were Saudi Arabia at $9.783 billion, the UAE at $8.8 billion, the United Kingdom at $6.325 billion and the European Union at $5.226 billion.³³ They kept rising in the new year, to $7.29 billion in July and August 2026, up 14.7 per cent, and the State Bank expects $44 billion in 2026-27.³⁵²⁴

Figure 7 sets the year’s accounts side by side, and it is the arithmetic of Pakistan’s solvency in one picture. On the State Bank’s basis the goods gap was $33.6 billion and services and income payments abroad added more. Remittances of $41.6 billion covered almost all of it, and the current account ended 2025-26 with a deficit of only $139 million, about 0.03 per cent of GDP, after a surplus of $1.84 billion the year before.⁴ Money sent home by workers abroad, not exports, is what keeps the account close to balance.

That strength has two weaknesses built into it. Remittances are concentrated by destination, with most coming from a few Gulf economies, so a downturn in Gulf construction or a change in one country’s visa rules reaches Pakistani households directly. Registrations for the UAE have already fallen on tighter immigration rules.³² They are also payment for exported labour rather than exported goods.

Remittances close the gap and do not narrow it, so Pakistan pays for its imports with the wages of workers who had to leave. Say this in measured terms. It is a serious analytical point, not contempt for the workers who are holding up the account.

Remittance totals change every month, and reports quote different periods. So do not carry a remembered number into the hall. The State Bank publishes a press release on workers’ remittances every month, and the Bureau of Emigration publishes the emigration figures.³⁶³² Take the latest figure from one of them a few days before you sit, and quote it with its month.

12. Debt, and what it costs to carry

Two kinds of debt, and how much

Domestic debt is owed in rupees, mostly to Pakistani banks. The state can always pay it, because it controls the currency, but paying it takes a large share of the budget. External debt is owed in foreign currency, mostly to other governments and to bodies like the IMF and the World Bank. This one is different in kind, because rupees cannot pay it. It has to be met in dollars, which must be earned by exporting, borrowed, or received as remittances.

In the government’s own figures, total public debt at the end of June 2025 was Rs 80,518 billion, which was 70.7 per cent of GDP. Of it, Rs 54,472 billion was owed at home and Rs 26,047 billion abroad. By the end of March 2026 the total had reached Rs 83,285 billion, and the government’s external debt was $92.2 billion.³⁷ Ten years earlier, in June 2016, total public debt had been Rs 19,677 billion, so it more than quadrupled in a decade.³⁷ Who Pakistan owes is concentrated too. China holds about $29 billion of Pakistani loans, roughly 22 per cent of the external debt and the largest share of any single country.³⁸

Interest has to be paid before anything else, so when much of the revenue goes on interest, what is left decides how much reaches schools, hospitals and development. In the first nine months of 2025-26 the government paid Rs 4,948 billion in interest, a third of all the revenue it collected in those months.³⁷ That was 23 per cent less than a year earlier, which the Finance Ministry puts down mainly to lower interest rates.³⁷ So the burden eases when rates fall, and it does not go away. The 2026-27 budget sets aside Rs 8.054 trillion for debt service, more than any other item.²²

The law that caps the debt

Pakistan has a law that limits its own borrowing, the Fiscal Responsibility and Debt Limitation Act, 2005. Section 3, as amended in 2016 and 2017, says debt should have come down to 60 per cent of GDP by 2017-18, then fall by half a point a year for five years, and after that by three-quarters of a point a year until it reaches 50 per cent in 2032-33.³⁹ On that path the ceiling for 2024-25 was 56 per cent. The Act counts debt in its own way, taking total public debt and subtracting the money federal and provincial governments keep on deposit in banks.³⁹ On that measure the government owed 64.4 per cent of GDP in June 2025, which is 8.4 points above the law’s own ceiling, as Figure 8 shows.³⁷

The law allows an exit, and it has been changed rather than kept. Section 3(4) lets the government depart from the limits because of “unforeseen demands on the finances of the Government”, provided the departure is temporary and the Finance Minister gives the reasons.³⁹ The Act also caps new government guarantees at 2 per cent of GDP a year and the total stock at 10 per cent, and the stock of Rs 4,322 billion in March 2026 was about 3.4 per cent of the GDP estimated for 2025-26, well inside that cap.³⁹³⁷ Parliament amended the Act again in June 2026, leaving the ceiling alone and changing only how the Debt Management Office is staffed.⁴⁰

So use this as a governance point in an answer. Pakistan does not lack a rule on debt. It has one it does not keep.

13. Taxes and the budget

Revenue is the domestic half of the trap

The OECD publishes a tax ratio measured the same way for every country. On it, Pakistan collected 10.5 per cent of GDP in tax in 2023, against an Asia-Pacific average of 19.5 per cent, as Figure 9 shows.⁴¹ A state that collects a tenth of national income has to borrow to do ordinary things, and borrowing abroad adds to the external bill that exports cannot cover. That is how the domestic half of the trap feeds the external half. Pakistan’s own figures for later years are higher because they are counted differently, so give the year and the source with any tax ratio.

The Federal Board of Revenue collects the main federal taxes, of which the largest are income tax and a general sales tax on goods. Much of the revenue is indirect, collected on what people buy rather than on what they earn. Indirect taxes are easier to collect, because the shop or the importer pays them, but they fall hardest on people with the least, since a poor household spends nearly all of its income. Broadening the tax base means bringing in the activity now outside it, above all retail trade, property and agricultural income. Every serious plan says so. It has been slow because each of those bases is politically organised and partly provincial.

Where the budget goes

The federal budget for 2026-27 was presented on 12 June 2026 by the Finance Minister, Muhammad Aurangzeb, with spending of Rs 18.77 trillion.⁴² It aims for growth of 4 per cent, inflation of 8.2 per cent, a primary surplus of 2 per cent of GDP and an overall deficit of 3.6 per cent. The Federal Board of Revenue must collect Rs 15.26 trillion, up about 8 per cent.⁴²²² Figure 10 shows where the money goes before any choice is made.

Debt service takes Rs 8.054 trillion, defence Rs 3 trillion, up 18 per cent after the conflict with India in 2025, and pensions Rs 1.169 trillion.⁴²²² The provinces receive Rs 8,848 billion from federal revenue under the NFC Award, and the federal development programme gets Rs 1 trillion.²² So a finance minister’s room to choose is far smaller than a campaign speech implies. The budget also cut income tax for salaried people in several brackets, brought in a fixed tax of 1 per cent of sales for small shopkeepers, and warned that dearer oil could add 0.8 per cent of GDP to the deficit.²²⁴² Say these things in any answer that asks what the government should spend on.

14. The state as an owner

Pakistan’s government does not only tax and spend. It owns airlines, power distributors, steel, railways and banks, much of it acquired when Bhutto nationalised banks, insurance and large industry in the 1970s.⁶ Later governments sold some and kept the rest, and the enterprises that were hardest to sell, because they lost the most, are the ones still owned. A loss-making public enterprise costs the budget twice. It needs money to keep running, and it pays no tax on profits it does not make. That money comes from the same budget that pays interest, defence and the provinces, so it competes directly with development.

Pakistan International Airlines is the clearest case. It built up more than $2.8 billion of losses and was losing about Rs 35 billion a year before it was restructured.⁴³⁴⁴ In December 2025 it was privatised after competitive bidding, valued at Rs 180 billion, about $643 million. A consortium led by the Arif Habib Group took 75 per cent for Rs 135 billion, and the government kept the other quarter.⁴³ The IMF’s resident representative said the sale met a primary commitment under the $7 billion programme, so it stopped the losses and kept the programme on track.⁴⁵

The handover came on 29 June 2026, two months late, after regulatory approvals, lessors’ consent, tax restructuring and financing took longer than planned.⁴⁶ At this first closing the consortium took management control, paid Rs 10 billion to the government and put Rs 80 billion of fresh equity into the airline. It has committed a further Rs 45 billion within twelve months and holds an option to buy the government’s last quarter for another Rs 45 billion.⁴⁶ Notice where the money goes. Most of it goes into the airline, not to the Treasury, so the gain to the budget is the end of the losses, not a large cheque.

A question on privatisation wants both sides in measured terms.

For selling, the state was funding a loss of tens of billions of rupees a year from a budget that cannot pay for schools, the buyer now carries the risk of the turnaround, and a firm that competes has reason to improve its service. Against it, a national airline serves places private operators may drop, jobs are lost, and a sale made under programme conditions and against a deadline is made from weakness, which affects the price. The defensible position is narrower than either. The case for selling a loss-maker is strong, because it ends a recurring loss. The case for privatisation as a way to raise money is weak, because selling an asset closes this year’s gap and cannot close next year’s. Judge each sale by whether it ends a loss, not by what it raises.

15. Pakistan and the IMF

What a programme is, and how many there have been

The International Monetary Fund lends foreign currency to a member that cannot pay its external bills. The money arrives in tranches, each released only after a review confirms that agreed conditions were met, so the country receives the loan gradually and the Fund keeps the power to stop. The Fund is not an aid agency and does not claim to be. It lends to be repaid, so its conditions are about the borrower’s ability to repay, which means collecting more revenue, spending less, pricing energy at cost and letting the exchange rate move. Understand that before criticising it, because it explains why the list never changes.

Pakistan’s first arrangement was the standby of 8 December 1958, from which it drew nothing. The IMF’s own table lists 26 lending arrangements between then and May 2025.⁵ By the Economic Affairs Division’s count, Pakistan had taken part in 24 programmes by the 2023 standby, which makes the $7 billion, 37-month Extended Fund Facility approved on 25 September 2024 the twenty-fifth.⁴⁷¹³ Across all of them Pakistan had drawn SDR 21.8 billion, about $31 billion.⁴⁷ The largest single arrangement was the standby of November 2008, agreed at $7.6 billion.⁴⁸ Counts vary between twenty-three and twenty-six depending on whether short arrangements, emergency loans and paired facilities are counted separately, so say “about two dozen since 1958” or give the source with the number.

Watch the unit too. The IMF keeps its books in SDRs, its own reserve unit, not in dollars. A widely copied table gives the 2008 drawings as “$4,936,035”, which is really the IMF’s figure in thousands of SDRs. It means SDR 4.94 billion, a good deal more than $4.9 billion.⁴⁸

Why the list repeats, and the counter-example

Most of those programmes did not deliver their reforms. The Pakistan Institute of Development Economics describes the pattern as “patchy, piecemeal implementation with some progress followed by regression”.⁴⁹ The 2008 standby is the clearest example, since its central reform, a general sales tax on value added, “died a spectacular death on the floor of parliament in 2010”.⁴⁹ The Fund’s own figures tell the same story. Of the nine closed Extended arrangements, Pakistan drew the full amount only once, in 2013, and drew 60.9 per cent of the 2019 one before it ended.⁵

That is why the list of conditions is the same every time. The same reforms keep being agreed and not finished, so they are still outstanding at the next crisis. It also tells you what the real constraint is. Pakistan knows what to do, and every programme names the tax base, energy pricing and the state enterprises. Each reform takes something from an organised group that can resist, and the pressure to do it disappears the moment the crisis eases. The binding constraint on reform is political, not intellectual.

Do not let that become a claim that Pakistan cannot reform. Pakistan was placed on the Financial Action Task Force’s grey list in 2018 and left it on 21 October 2022, after completing 34 action items across two action plans and an on-site visit, from 29 August to 2 September 2022, to check them.⁵⁰ Look at what was different. The items were specific and could be checked one by one, an outside body did check each one, and failure carried a large and immediate cost, the loss of easy access to international banking.

Under those conditions Pakistan finished a demanding programme on time. So the lesson is about design, not character. Reform that is specific, checked and penalised gets done, and reform that is general, self-assessed and postponable does not.

An answer that presents the IMF as the cause of Pakistan’s difficulties has the direction of cause backwards, and an examiner from the service will see it at once. Pakistan goes to the Fund because it has run out of foreign currency, so the shortage is the cause and the programme is the consequence. Criticism is still available and it scores. Criticise the design of a specific condition, for example raising energy prices without protecting poorer consumers, and say what you would do instead.

16. CPEC, and what it did

The China-Pakistan Economic Corridor is a programme of Chinese-financed investment in Pakistani infrastructure, agreed in 2015. It covers roads, power stations, the port of Gwadar in Balochistan, and industrial areas called special economic zones. By December 2024, the government told the National Assembly, forty-three projects worth about $24.7 billion had been completed, with eight more worth about $760 million under way.⁸ Energy projects worth over $15 billion have added 9,504 megawatts, and a further $6.71 billion of energy projects would add 3,544 megawatts.⁹ Gwadar port is operating and handles transit cargo for Afghanistan, and an international airport, a hospital and desalination plants have been built around it.⁸⁹ The special economic zones at Rashakai and Bostan and the second phase of the Gwadar free zone were still in progress.⁸

A candidate who takes a simple side, triumph or trap, writes a weaker answer than one who separates the parts. On energy, CPEC delivered, and Pakistan should say so, since power cuts of about eighteen hours a day ended.¹⁵ On exports it has not yet delivered. The special economic zones were the part meant to change what Pakistan makes and sells, and they are the least finished part, because roads and power stations were built faster than the factories to use them.

On debt, be precise rather than alarmed. China holds about $29 billion, roughly 22 per cent of external debt.³⁸ That is a real concentration worth stating plainly, and it is not the majority, so the claim that CPEC alone caused the debt does not survive the arithmetic in the section on debt.

Put in this chapter’s terms, CPEC removed a real constraint and left the export base where it was, like every other turn in the history. That is not an argument against CPEC. It is an argument about what infrastructure can and cannot do. Electricity lets a factory run, and it does not decide what the factory makes or whether anyone abroad wants to buy it.

The stated aim of the next phase is industry, agriculture and Gwadar rather than more power stations.⁸ So the useful thing to propose is a test. The next phase succeeds when a special economic zone exports something Pakistan did not export before, which is the same test of composition the rest of this chapter applies.

17. What is under the ground

Pakistan holds large mineral deposits, mostly in Balochistan and Sindh, and mostly not yet mined at scale. Reko Diq, in Balochistan, is one of the world’s largest undeveloped copper and gold deposits. Its first phase was costed at $7.7 billion, half of it to be borrowed, and in November 2025 the US Export-Import Bank approved $1.25 billion of that loan.⁵² It was planned to produce about 200,000 tonnes of copper a year, plus gold, from 2028, with Barrick Gold owning 50 per cent, federal state companies 25 per cent and the Balochistan government 25 per cent.⁵²⁵¹ Saindak, in Chaghi district of Balochistan, has been worked since 2002 by a Chinese state company, Metallurgical Corporation of China, whose lease was extended for fifteen years in 2022.⁵³ The Thar coalfield in Sindh holds an estimated 175 billion tonnes of lignite, among the largest coal deposits in the world.⁵⁴

The 2026 paper asked whether natural resources can become “a source of conflict and instability”, and the idea it points at is known as the resource curse. Countries with large mineral or oil wealth often grow more slowly than countries without it, and there are four reasons to explain. Mineral exports bring in foreign currency, which raises the value of the currency and makes every other export dearer abroad, so a mining boom can shrink manufacturing. The revenue is unstable, because copper and coal prices move a great deal. A tonne of copper sold is gone, so unless the proceeds are turned into something lasting the country ends the boom poorer than it looked. A mine is one site with one stream of revenue, which a small group can capture, and that is why mineral wealth is so often linked with weak accountability and with conflict.

The deposits are largely in Balochistan, the poorest province, with the smallest population and so the smallest share under the population-weighted federal formula. The province where the copper is has the strongest claim to the proceeds and the least power to win an argument by numbers. Write this as a design question about federal arrangements. Reko Diq gives the Balochistan government a 25 per cent shareholding, a direct stake rather than a transfer negotiated afterwards.⁵¹ Name that arrangement approvingly and then test it, by asking whether the revenue reaches provincial services and whether citizens can see what was received and spent. A candidate who proposes a transparent, checkable provincial share reads as a future officer, while one who frames the same facts as a quarrel between the province and the state does not.

Four lessons follow, and the first is already visible. Mining is slow, so do not budget for it early. In April 2026 Barrick slowed development at Reko Diq and extended its review to mid-2027, warning that the cost and timetable could rise significantly.⁵⁵

Save the receipts rather than spending them, as countries that handled resource wealth well did, spending only the return on a fund. Protect the other exports, so that mineral earnings do not price textiles and technology out of foreign markets. Publish everything, meaning what was produced, sold, received by the state and received by the province, because openness is the one cheap measure that addresses the accountability problem directly. The way forward fits in one line. Treat mineral revenue as capital to be converted, not income to be spent.

18. Where it stands now

The stabilisation is real and should be credited. On 8 May 2026 the IMF Executive Board completed the third review of the Extended Fund Facility and the second of the Resilience and Sustainability Facility, releasing about $1.1 billion and $220 million.⁵⁶ The Fund expected a primary surplus of 1.6 per cent of GDP for 2025-26, growth of 3.6 per cent and average inflation of 7.2 per cent. Gross reserves had risen to $16 billion at the end of December 2025 from $14.5 billion at the end of June 2025.⁵⁶ Since then Moody’s has raised Pakistan’s rating to B3 with a stable outlook, and Pakistan sold $3 billion of Eurobonds in September 2026.²³ Both show that lenders expect to be repaid, though B3 is still well below investment grade, so it measures recovery rather than safety.

The fourth review shows the pattern this chapter predicts. An IMF mission arrived in late September 2026, and talks formally opened on 29 September, with about $1.2 billion across the two facilities at stake and about $4.8 billion received so far.⁵⁷ The government briefed the mission on how the Gulf war and disruption in the Strait of Hormuz had hit activity and revenue in the first quarter.⁵⁷ Pakistan met its numerical targets and missed structural ones. Spending on health and education fell about Rs 370 billion short of the agreed floor, the sugar policy promised as a condition was not adopted, and reform of state enterprises slowed, while the condition to adjust gas and electricity prices on time was met.⁵⁸ The circular debt ended the year Rs 221 billion over its ceiling.¹⁷

The external account has held, narrowly. The current account ended 2025-26 with a deficit of $139 million after a surplus the year before, and ran a deficit of $543 million in July and August 2026, carried by remittances while the trade gap widened.⁴³⁵ Inflation, at 10.3 per cent in September 2026, was well above the target band because of oil.²¹ The Governor of the State Bank called the rising trade deficit a “big concern”.²⁴

What the Fund asks for has not changed. Broaden the tax net, improve compliance, and keep fuel, electricity and gas prices in line with costs while protecting poorer consumers.⁵⁶ Those are all measures to reduce demand or raise revenue, and none of them makes Pakistan sell more abroad. Stabilisation and transformation are different tasks, and only the first is clearly under way.

The government’s own plan names the right target. Uraan Pakistan, launched by Prime Minister Shehbaz Sharif on 31 December 2024, is built on five headings, exports, digitalisation, energy and infrastructure, climate, and equity, and aims to double exports to about $60 billion by 2029.³¹ In 2025-26 goods exports were $30.13 billion and falling, and IT exports $4.6 billion, about $35 billion together with three years left.³²⁹ Apply the test to the mix, not only the total. Reaching $60 billion by selling more textiles would repeat the structure at a larger scale, and reaching it by selling different things would end the trap.

Much new investment is routed through one body, so name it. The Prime Minister set up the Special Investment Facilitation Council in June 2023 to attract investment from the Gulf states in defence, agriculture, minerals, information technology and energy, as a single window for investors.⁵⁹ Its Apex Committee is chaired by the Prime Minister and includes nine federal ministers and the four provincial chief ministers, with the Chief of the Army Staff by special invitation and a National Coordinator from the Pakistan Army.⁵⁹

Two things are true at once. The problem it answers is real, since an investor once needed approvals from several ministries and a province, any of which could stop a project and none deliver it. The objection is real too, because the army now sits in the body that routes investment decisions, which works around the civil service rather than repairing it. The remedy that keeps the speed and restores accountability is to put the Council on a statutory footing, with a defined term and a duty to report to Parliament.

19. What each side wants

An answer that treats “the government” as one actor cannot explain why the same reforms are announced again and again and delivered slowly. The Finance Ministry wants the external account to hold and the programme to stay on track. Its tools are taxes, subsidies and interest rates, which all work on demand, so it can stop a crisis and cannot build an export sector. Exporters want a predictable exchange rate, refunds paid on time, energy at a price comparable to competitors’ and stable policy, and their complaint is rarely about tariffs.

The IMF wants repayment and stability, so its conditions are about revenue, subsidies and prices.⁵⁶ It is not in the business of changing what a country makes, and seeing that the Fund’s programme and Pakistan’s export problem aim at different things is one of the most useful distinctions in this chapter. The provincial governments hold much of what would raise the tax ratio, such as agriculture, property and services, and much of what would raise productivity, since education and health went to them in 2010. So the revenue problem and the human-capital problem are both partly provincial.

Consumers, and the poor above all, carry the cost of the adjustment. When energy is priced at cost the bill lands on households, and the IMF’s own language pairs that with protecting vulnerable consumers.⁵⁶ Remittance-sending workers want a cheap, quick and trusted way to send money home, which is why the gap between two exchange rates matters to them. The state as a borrower wants creditors who roll over their loans, and with about 22 per cent of external debt owed to one country, that relationship is an economic fact as well as a diplomatic one.³⁸

20. The strongest case against this chapter’s argument

The first objection is that the constraint is energy and productivity, and exports follow them. Firms do not export because they cannot produce at a competitive cost, so fixing input costs and reliability would make exports respond. On this reading “grow exports” is a result, not a policy. The second objection is that stabilisation is working, so the model may be wrong. Growth recovered, reserves were rebuilt and the current account came close to balance, so perhaps the trap is being escaped by ordinary means rather than by transformation.⁵⁶⁴

The third objection is that remittances change the arithmetic. They covered almost the entire trade and income gap in 2025-26, so a model built only on goods trade understates the country’s real capacity to pay.⁴³³ The fourth is that the energy fix was real and takes a decade to show, since industry responds slowly, so judging the export base now may be premature.¹⁵

The objections sharpen the argument rather than overturn it. Exports are the constraint, and “raise exports” is not a policy, so the policy is whatever changes the mix. That means energy at a competitive and predictable price, refunds paid on time, and one or two new sectors taken seriously enough to build them. The best evidence for the objections also supports the argument. The electricity constraint was lifted and exports stayed concentrated, which suggests electricity was necessary and not enough. A position held up by remittances is steady, not solved, because it leaves the constraint in place for the next expansion, and the widening trade gap of 2026 is the first sign of it.²⁶

21. How the debate stands among people who study it

Four views run through the writing on Pakistan’s economy. The structuralist view holds that growth is limited by the external account, which is limited by what a country can sell, and its evidence is the ratio of imports to exports and the concentration of exports.³²⁷ The stabilisation view holds that the problem is repeated fiscal and monetary indiscipline, so holding the line long enough will bring investment, and its evidence is the improvement under the programme.⁵⁶

The governance-and-revenue view holds that the state cannot fund what growth needs because it collects too little, so tax administration is the first reform, and its evidence is a tax ratio nine points below the regional average.⁴¹ The productivity view holds that Pakistani firms do not compete on cost or quality, so the answer is skills, technology and better firms rather than macroeconomics. This chapter draws on the first and the fourth. Say which you take and support it, because an examiner rewards a candidate who knows the subject is argued about.

22. What the examiners have asked

The 2023 question continues in the paper by asking whether export performance “remains the most relevant long-term challenge” and for “the radical structural reforms to improve export performance”.¹ CPEC was also asked as a regional question in 2016 and 2023, and the paper set IT’s impact in 2000.

Three habits stand out. This is a prescription paper, since most of these ask outright for measures, mechanisms or a way forward, so half your answer should be the remedy, specific enough to argue with. Debt and exports are set together, in 2011, 2016 and 2023, which tells you the link the examiner wants, that Pakistan borrows because it does not earn enough abroad. The recent papers are converging questions, bundling the economy with climate, population and governance, so be ready to write the economy as one strand of a combined answer. Areas taught here that have not been set whole, and could be, include export concentration, Uraan Pakistan’s target, IT exports, the circular debt, privatisation after PIA, and the 2026-27 budget.

23. Facts to take into the examination hall

• Goods trade, 2025-26. Exports $30.13 billion, down 6 per cent; imports $69.60 billion, up 8 per cent; gap $39.47 billion; imports 2.3 times exports.³ July and August 2026: gap up 18 per cent.²⁶
• What Pakistan sells, 2025-26. Textiles and clothing $17.93 billion, 59.5 per cent of the total; food $5.02 billion, with rice down from $3.35 to $2.29 billion; everything else about $7.2 billion.²⁷²⁸
• IT exports, 2025-26. $4.6 billion, up 21 per cent; freelancers $1.76 billion, up 78 per cent, of which computer work $1.16 billion.²⁹³⁰
• Remittances. A record $41.6 billion in 2025-26; Saudi Arabia $9.783 billion, the UAE $8.8 billion.³⁴³³ $7.29 billion in July and August 2026, up 14.7 per cent; $44 billion expected for 2026-27.³⁵²⁴
• Current account. Deficit of $139 million in 2025-26, after a surplus of $1.84 billion in 2024-25.⁴
• Growth, 2025-26. 3.70 per cent, up from 3.18; economy about $452.1 billion; income per person $1,901. Agriculture 23.4 per cent of GDP and 33.1 per cent of jobs.²
• Prices and money. Inflation 9.2 per cent in July, 11.1 in August and 10.3 in September 2026, against a target of 5 to 7. Policy rate held at 11.50 per cent on 14 September 2026. State Bank reserves about $21.4 billion after a $3 billion Eurobond; Moody’s B3, stable.²³²¹
• The rupee. Low of 307.75 to the dollar in September 2023; about 277.5 in September 2026. In 2023 the open market rate reached Rs 340 against an interbank Rs 307, and about $4 billion of remittances went informal in FY23.²⁵¹²
• Circular debt. Rs 1.614 trillion in June 2025; Rs 1.835 trillion in June 2026, Rs 221 billion over the IMF ceiling. Rs 1.225 trillion of bank loans repaid through a surcharge of Rs 3.23 a unit.¹⁶¹⁷
• Debt. Rs 80,518 billion, 70.7 per cent of GDP, in June 2025; external public debt $92.2 billion in March 2026; interest a third of revenue in July to March 2025-26. The Act’s own measure, 64.4 per cent, against a legal ceiling of 56.³⁷³⁹ China about 22 per cent of external debt.³⁸
• Tax. 10.5 per cent of GDP in 2023 on the OECD’s measure, against 19.5 in Asia-Pacific.⁴¹
• Budget 2026-27. Rs 18.77 trillion; tax target Rs 15.26 trillion; debt service Rs 8.054 trillion; defence Rs 3 trillion; primary surplus 2 per cent of GDP; deficit 3.6 per cent.⁴²²²
• Poverty. National line 28.9 per cent in 2024-25, up from 21.9 in 2018-19; World Bank 23.0 per cent at $3.00 a day and 47.9 per cent at $4.20.¹⁹²⁰
• The IMF. First arrangement 8 December 1958; the $7 billion facility of 25 September 2024 the twenty-fifth by the Economic Affairs Division’s count. Third review completed 8 May 2026; fourth review talks opened 29 September 2026.^547135657
• FATF. Grey list from 2018; out on 21 October 2022 after 34 action items.⁵⁰
• PIA. Privatised December 2025, 75 per cent for Rs 135 billion; handed over 29 June 2026.⁴³⁴⁶
• CPEC. 43 projects worth about $24.7 billion completed by December 2024; energy projects added 9,504 megawatts; the special economic zones the least finished part.⁸⁹
• Minerals. Reko Diq’s first phase costed at $7.7 billion, with $1.25 billion from the US Export-Import Bank; Balochistan holds 25 per cent; development slowed in April 2026 with a review to mid-2027. Thar holds about 175 billion tonnes of lignite.⁵²⁵⁵⁵⁴
• 1998. Reserves near $1 billion; foreign currency accounts of about $11 billion frozen; technical default.⁷ 2022 floods. 33 million affected; $14.9 billion of damage.¹¹

Three traps lurk in these numbers. Pakistan’s fiscal year runs July to June, so “2026” and “FY26” are different periods. Goods trade differs from goods and services, and customs figures differ from the State Bank’s, by billions. National and international measures differ too, as with the tax ratio. So name the year, the basis and the source every time.

24. Positions you can defend

Position A. The export base is the constraint, so the mix is the target. Debt and deficits are symptoms of exports that cannot pay for imports, so the measure of any policy is whether it changes what Pakistan sells, and the test of Uraan Pakistan is whether $60 billion arrives with a different mix. Support it with imports at 2.3 times exports, textiles at three-fifths of the total, and IT exports growing at 21 per cent a year.³²⁷²⁹

Position B. Revenue first, because the state cannot act without it. At about 10.5 per cent of GDP the state must borrow for ordinary functions, and borrowing abroad feeds the same trap. Broadening the base, including agriculture, retail and property, comes before everything else, and it is what the programme asks for. Support it with the tax ratio against the regional average and the budget’s interest bill.⁴¹²²

Position C. Stability has been bought and must now be used. The stabilisation worked, and the risk is treating it as the achievement. The years of stability are the window in which the export base has to change, because the next external shock closes it, and the oil shock of 2026 is a warning. Support it with the programme’s record, the widening trade gap and the return of inflation.⁵⁶²⁶²¹

Position A is the strongest for the 2016 and 2023 questions, because it answers the mechanism they name. Position C is the most sophisticated, because it credits what has been achieved and still argues for change, which is the register the paper rewards.

25. What loses marks on this subject

The commonest fault is recommending “increase exports” and “reduce imports”. Those restate the problem, so every proposal needs an instrument and an actor, meaning which sector, what changes for it, who does it and how it is paid for. Close behind is treating debt as the subject when the question is about the external account. Debt is what happens when exports do not cover imports, so state the mechanism and the debt paragraph writes itself.

Two opposite errors cost marks on the record. Writing that the economy is collapsing ignores published figures on growth and reserves, and a candidate who cannot acknowledge them will not be believed on anything else.⁵⁶ Writing that the crisis is over ignores the trade gap and the return of inflation in 2026. Steadiness restored is not the export base changed, so write both halves. Mixing fiscal years, or goods with goods and services, is the most common factual slip.

Blaming the IMF reverses cause and effect. The Fund’s conditions are about revenue and prices because that is what the Fund does, so criticise the design of a specific condition and say what you would do instead.⁵⁶ Finally, do not describe the load-shedding of 2012 as today’s energy crisis. Power cuts fell from about 18 hours a day to near self-sufficiency, and the energy problem now is the circular debt.¹⁵¹⁷

26. A worked answer plan

The question is from 2023, quoted as printed. “Critically discuss why Pakistan’s economic growth continues to be “Balance-of-Payments” constrained with a focus on the long-term challenge to Pakistan’s export performance under pressure.” (20 marks)¹

What the verb asks. Critically discuss why. The question already says growth is balance-of-payments constrained, so it is not asking you to prove it. It asks for the mechanism, and its second clause says where to put the weight, on export performance.

The position, in one line. Growth raises imports faster than exports because Pakistan’s export earnings rest on a few low-value products, so each recovery ends in an external deficit, and the constraint is the mix of exports rather than the level of debt.

The claims, each with its fact.

• The ratio states the problem. Goods exports were $30.13 billion against imports of $69.60 billion in 2025-26, and in the first two months of 2026-27 imports grew almost twice as fast as exports.³²⁶
• Concentration is why exports cannot respond. Textiles at $17.93 billion are about three-fifths of all goods exports, and the next group, food, earned $5.02 billion.²⁷²⁸ One sector cannot absorb a bad cotton year and a competitor at once.
• So adjustment falls on demand, and that stops the growth. The IMF programme’s conditions are revenue, compliance and energy priced at cost, all of which work on demand.⁵⁶
• The constraint is not electricity, which was the old answer. CPEC added 9,504 megawatts and ended the long power cuts, and the export structure did not change.⁹¹⁵
• Remittances hide the constraint without removing it. $41.6 billion of remittances kept the current account close to balance in 2025-26.³³⁴
• The state’s plan sets the right test. Uraan Pakistan targets about $60 billion of exports by 2029, and the question is whether that arrives as more textiles or as something else.³¹

The counter. Remittances and a near-balanced current account suggest the external position is manageable.⁴ Concede it, then answer it. A position held up by transfers is steady, not solved, and it leaves the constraint in place for the next expansion.

The measures. Name the radical reforms the question asks for, each with an actor. Energy at a competitive, predictable price through cutting distribution losses. Sales tax refunds paid on time by the Federal Board of Revenue. The special economic zones judged by new export products. Payment channels and internet access treated as export infrastructure for IT.

The conclusion. Name the test rather than restating the verdict. Pakistan will stop being balance-of-payments constrained on the day a recovery raises exports as fast as it raises imports. Nothing in the present structure does that yet, and Uraan Pakistan will be judged by its mix, not its headline number.

What to remember from this chapter

Pakistan does not have a growth problem. Growth itself brings on the crisis, because imports rise faster than exports can, so reserves fall and the economy is slowed on purpose. Imports were 2.3 times exports in 2025-26, and textiles were about three-fifths of everything sold abroad.

Remittances of $41.6 billion, not exports, kept the current account close to balance. IT exports of $4.6 billion, growing 21 per cent a year and needing almost no imports, are the first sign that the mix can change.

The energy crisis of 2013 was solved and replaced by the circular debt, which ended June 2026 at Rs 1.835 trillion and is paid for by a surcharge on consumers who pay. Debt is above the legal ceiling, interest takes a third of revenue, and tax is about a tenth of GDP.

The IMF’s list repeats because the same reforms keep being agreed and not finished, while FATF showed that specific, checked and penalised reform gets done. The stabilisation of 2025 and 2026 is real and should be credited before you argue that it is not the same as escape. Name the test, not the verdict.

Check yourself

Answer each from memory before you open it. The section that teaches each answer is named in brackets.

1. Explain in four steps why growth brings on a crisis in Pakistan. (Growth itself brings on the crisis.)

2. What happened in 1998, and why is it the best date to argue from? (How the economy reached this point.)

3. Why does a weak cotton crop damage both sides of the trade gap? (Agriculture and the land.)

4. What is the circular debt, where did it stand in June 2026, and who pays for clearing it? (Industry, energy and the circular debt.)

5. Give three poverty rates for Pakistan and say why they differ. (Work, wages and poverty.)

6. Why does a fall in inflation not mean prices have come down, and why did inflation rise again in 2026? (Money and prices.)

7. What are the two exchange rates, and why does the gap between them matter more than the rate? (The rupee, and the gap that matters more than the rate.)

8. What is inside Pakistan’s export total, and why does the mix matter more than the size? (What Pakistan buys and sells.)

9. Why are IT exports different in kind from textile exports, and how big are they? (The newer exports.)

10. How did Pakistan’s current account end 2025-26, and what kept it close to balance? (Remittances and the current account.)

11. What does the Fiscal Responsibility and Debt Limitation Act require, and how far above it was Pakistan in June 2025? (Debt, and what it costs to carry.)

12. Name the largest items in the 2026-27 federal budget, and say what they mean for a finance minister’s choices. (Taxes and the budget.)

13. Why was PIA sold, where did the money go, and what is the limit of privatisation as a way to raise money? (The state as an owner.)

14. What does Pakistan’s FATF record prove about reform, and how does it differ from the IMF record? (Pakistan and the IMF.)

15. Give the verdict on CPEC in three parts. (CPEC, and what it did.)

16. Name four ways natural resources can leave a country worse off. (What is under the ground.)

17. From the federalism chapter. Why does a province’s spending not create the same pressure for good use as its taxing would? (An earlier chapter.)

18. From the foundations chapter. What is the difference between a country’s current account and its reserves? (An earlier chapter.)