Annual Plan defeatism

Every year Pakistan publishes an Annual Plan with growth targets, sectoral targets, investment targets, export targets, and a Public Sector Development Programme. Last 4 annual plans targeted a growth rate of about 4 percent and exports of about 8 percent of GDP—a lackluster economy at best! Where do these numbers come from? Should they be taken seriously?

1. Where do the targets come from?

Not really — except for a day of newspaper reporting and commentary by old-fashioned economists. These are not bottom-up estimates built from firm-level investment plans, capacity utilization, productivity trends, export orders, crop conditions or serious sectoral modelling. They are numbers backed by little research or investigation. An aggregate growth target is chosen first — largely for political and presentational reasons — and then broken into sectoral numbers that add up neatly.

Agriculture is not assigned 3.6 percent because anyone has modelled water availability, seed quality, fertilizer use, climate risks or crop incentives. It is assigned 3.6 percent because the aggregate arithmetic needs it. Industry and services are treated the same way. The number comes first. The story is written later.

A number chosen to make the table work is not a forecast. It cannot be tested because it was never derived from a testable mechanism.

2. What policies or instruments are these targets based on?

This is where the Plan is thinnest. A credible growth forecast should specify the policy levers expected to produce the result: which reforms, which prices, which institutional changes, which market openings, which incentives, and which constraints will be removed.

The Annual Plan instead asserts outcomes without instruments. Investment will rise to 15 percent of GDP — through what mechanism? Tax policy when much of it is moving in the wrong direction. Deregulation of which sector? Lower policy uncertainty? Reform of which market? Serious public investment beyond the usual political projects? The document does not say.

Exports will increase sharply — based on what? Which exchange-rate assumption? Which trade-policy change? Which energy-price reform? Which resolution of the refund backlog? Which improvement in logistics? Given weak investment in recent years, where is the capacity to export? The Plan offers ambition, not mechanism.

Without instruments, the targets are wishes with decimal points. If growth misses, there is no policy lever whose failure can be identified. There is only the familiar appeal to external shocks, global uncertainty, weather, oil prices or political instability. But a serious plan would have built these risks into scenarios from the beginning. A risk paragraph is not a forecast.

3. Tested against existing research, do the targets hold up?

Pakistan’s academic and policy research seldom examines Annual Plan targets seriously. There is occasional nit-picking, often enough only to give the exercise undeserved credibility. The Plan therefore continues to publish point targets without forecast-error tables, fan charts or any decomposition of shocks, assumptions and implementation failures. Almost no one takes the numbers seriously beyond the day they appear in the press.

The strange thing is that the relevant knowledge exists. Forecasting research tells us to measure errors, publish uncertainty and compare forecasts with outcomes. Public-investment research tells us projects are delayed, overcommitted, underfunded and weakly linked to productivity. Yet the Annual Plan learns nothing from either tradition.

A serious Plan would ask: why were last year’s targets wrong? Which errors came from shocks, which from bad assumptions, and which from implementation failure? How did those errors affect revenue, imports, expenditure and borrowing? Which PSDP projects actually relieved which binding constraint? What did they add to growth?

The Annual Plan does none of this. The story resets every year. The problem is not merely that targets are missed. It is that misses are not studied by the institution that produces them. Without learning, planning becomes amnesia with tables.

4. Given PSDP, throw-forward and completion practices, is the investment plan reasonable?

No — and this is the sharpest contradiction in the document. Pakistan has its own literature on public investment failure. The Doing Development Better study showed how the planning system had deteriorated from its earlier role in public-sector-led growth. Other work documents delays, cost overruns and throw-forward. None of this alters the Annual Plan’s growth arithmetic. It still speaks as if PSDP projects are annual growth instruments, when the portfolio is really a decade-long backlog.

The IMF’s Public Investment Management Assessment found Pakistan’s PSDP had a cost-to-complete of roughly Rs10.7 trillion — more than fourteen times the FY2022–23 PSDP allocation of Rs727 billion. At that funding rate, the existing portfolio alone would take about fourteen years to finish with zero new projects. The Planning Commission’s PSDP 2025–26 document confirmed a throw-forward above Rs10 trillion against a PSDP of about Rs1 trillion. By 2026, the Planning Ministry again confirmed a throw-forward above Rs10,000 billion against a development ceiling of roughly Rs1.126 trillion.

So, the Annual Plan sets a one-year growth target while its main public-investment instrument is committed a decade or more into the future. Most of this year’s allocation goes to old, partially completed schemes — many with token allocations, many revised upward mid-cycle, many chosen for political rather than productivity reasons.

There is no published estimate of the marginal productivity of this spending. No counterfactual shows what the same money could achieve through regulatory reform, tax reform, city reform or energy-market reform. No demonstrated link connects this year’s PSDP allocation to the 4 percent growth target. The link is asserted, not shown.

This matters because a large fraction of the PSDP remains brick and mortar. Even on a narrow definition, nearly half of this year’s federal PSDP goes to highways, water, power, railways, housing and communications. On a broader definition, including special-area and Cabinet Division schemes, the share moves toward three-quarters. Development is still treated as construction, not productivity, markets, reform or firm growth.

5. Which growth model is being used — Haq/Harrod-Domar or endogenous growth?

This is the underlying issue. Pakistan’s planning apparatus still runs on the Harrod-Domar model inherited from the Mahbub-ul-Haq era: growth is constrained by a capital gap; public investment closes the gap; more capital stock produces more output; therefore, the PSDP is the growth engine almost by definition.

That framework had some validity in an early-development economy short of basic infrastructure. It does not describe an economy where the binding constraints are institutional: weak contract enforcement, regulatory sludge, distorted tax incentives, uncompetitive markets, policy uncertainty, low firm productivity, shallow capital markets and a public investment system that cannot complete what it starts.

Modern growth theory and cross-country evidence locate the real engine of growth in productivity: firm competition, technology adoption, human capital, trade openness, urbanization, market depth, knowledge and institutional quality. Capital spending matters only when it removes a specific binding constraint on private activity. A road connecting markets to ports may raise productivity. A road duplicating an existing route for political reasons does not, regardless of how it appears in the PSDP.

The Annual Plan does not engage with this distinction. It treats PSDP allocations as inherently growth-producing without asking which constraint each project removes, whether that constraint is actually binding, or whether reform would achieve the same growth at lower cost.

That is why the Plan is not merely weak. It is built on a static theory of growth in an economy whose real constraints are dynamic and institutional.

Copyright Business Recorder, 2026