Sanctity of statistics

GDP estimates are more than numbers; they shape fiscal policy, investor sentiment and IMF assessments. In Pakistan, the National Accounts Committee (NAC) is tasked with accurately reflecting economic activity. Yet, its FY2024–25 figures – 2.68 per cent real GDP growth and $411 billion nominal GDP – raise serious concerns amid sectoral contractions in crops, manufacturing and mining.

This critique outlines five problems: contradictory agricultural math, erratic utility output, unjustified industrial optimism, narrative-driven revisions and inflated per capita income based on outdated data. These flaws suggest a shift from evidence-based reporting to politically motivated data shaping, undermining the credibility of Pakistan’s economic narrative.

We deploy a triangulated method combining official NAC estimates and sectoral decompositions; base-year adjusted weights from PBS and SNA (System of National Accounts 2008); cross-checks using external data from SBP, Nepra and Planning Commission sources; and residual analysis and implied rate derivation to detect statistical anomalies. Pakistan's GDP is compiled using the production approach (Laspeyres index with 2015–16 base year), aggregating sectoral Gross Value Added (GVA) across agriculture, industry, and services. But when sector-level contractions (-13.5 per cent in crops or -1.53 per cent in LSM) coexist with headline growth, something is amiss – either in weights, deflators or unexplained residuals.

To probe this, we simulate counterfactuals: for example, whether livestock growth of 4.72 per cent (with outdated Census 2006 benchmarks) can realistically offset collapses in wheat, maize, cotton and sugarcane. We also derive implied exchange rates and inflation differentials to assess the plausibility of dollar-denominated GDP and per capita income figures.

We conduct a sensitivity analysis to assess whether small subsector gains (like forestry, fishing) could offset major GDP contractions. Sectoral weights are cross-checked with Supply and Use Tables, the 2022 PBS rebasing, and NAC reports for consistency. Quarterly data (such as EGWS) is extrapolated to test annual claims. Our standard is methodological coherence: the GDP story fails if parts don’t align.

Agriculture is reported at +0.56 per cent growth despite a -13.49 per cent collapse in major crops (wheat, cotton, maize, sugarcane), which form a third of agricultural GVA. Cotton ginning fell -19.03 per cent. The positive headline is driven by livestock’s fixed +4.72 per cent growth, based on 2006 census projections – not current data – ignoring fodder shortages, disease outbreaks and inflation.

The disconnection deepens when one considers that forward-linked industries like textiles, dairy, and leather did not mirror this supposed livestock or crop rebound. For instance, while cotton fell -30.7 per cent, textiles reportedly grew +2.15 per cent and apparel +7.62 per cent – defying raw material logic.

Quarterly growth in EGWS was negative: Q1 (-2.30 per cent), Q2 (-3.40 per cent), Q3 (-7.72 per cent). Yet, annual growth is somehow +28.88 per cent. This implies Q4 grew 129 per cent year-on-year – a mathematical and operational impossibility absent mega-projects or a policy shock.

Nepra data confirms no such leap in energy output or infrastructure commissioning. The sector remains crippled by circular debt (Rs2.393 trillion), underinvestment, and system losses. No reported structural reforms or capacity jumps justify the surge. This disjointed result likely reflects statistical smoothing, residual balancing, or political calibration.

Industrial growth of +4.77 per cent is claimed, yet: LSM (45 per cent of manufacturing) contracted by -1.53 per cent; mining and quarrying declined -3.38 per cent; construction is up +6.61 per cent, despite only 40 per cent PSDP utilisation; and EGWS’s outlandish +28.88 per cent is already discredited above.

High interest rates (>20 per cent), weak PSDP execution, and stagnant cement sales all contradict any meaningful industrial revival. With 75 per cent of the industrial base in decline, the headline growth is misleading.

PBS monthly data for LSM also confirms contraction in seven out of eight months up to February 2025. The declines in textiles, machinery, and food production undermine any narrative of rebound.

The NAC adjusted past GDP by just 0.01 percentage points: FY23 from -0.22 per cent to -0.21 per cent, and FY24 from 2.50 per cent to 2.51 per cent. Given significant sectoral re-estimates, such minor revisions are statistically implausible.

Rather than reflect new data, these revisions seem tailored to construct a neat narrative: minimise past contraction, amplify recovery and impress external observers. Sectoral revisions in agriculture (+0.22pp), industry (+0.28pp), and health (-2.65pp) should cause a bigger aggregate shift. That they don’t suggests backloaded residual smoothing.

This mirrors what development statisticians call ‘defensive revisions’ – edits designed not for precision but narrative continuity, often around IMF reviews or budget cycles.

Reported GDP of Rs114.7 trillion converts to $411 billion – at Rs279.15/USD. But the average interbank rate during FY24-25 was closer to Rs290. Using a realistic exchange rate would reduce dollar GDP by at least $15 billion. Similarly, per capita income of $1,824 assumes a population of 225 million. But the Census places it at around 246 million, implying a corrected figure of around $1,609 – a 12 per cent overstatement.

When inflation runs over 21–26 per cent, and real growth is just 2.68 per cent, nominal GDP growth is driven by price levels, not volume. This inflation-led nominal boost is passed off as a real gain through forex gaming and outdated demographics.

Research by local think tanks also suggests the real exchange rate is significantly misaligned, worsening inflation and debt dynamics. Overvaluation by 24 per cent (Rs278 vs real Rs211.5) suggests that inflated rupee values prop up GDP artificially while hurting macro stability.

The 113th NAC release is less a statistical portrait and more a crafted script. It sacrifices coherence and credibility for continuity and political convenience. Core sectors show contraction; headline aggregates report growth. Revisions are minimised to shape recovery stories; exchange rates and population figures are chosen selectively.

Without transparent methodology, sectoral reconciliation or independent audit, national income accounting risks becoming narrative management. Statistical independence is essential for credible policy, planning, and public trust.

Pakistan must invest in methodological reform, real-time data systems and depoliticised statistics. Otherwise, GDP will remain a moving target, shaped more by politics than production.

Nadeem Ul Haque is the former deputy chairman of the Planning Commission.

Yasir Zada Khan is a freelance contributor.