Pakistan accounts for nearly half of MENAAP’s extreme poor: World Bank

ISLAMABAD: Pakistan accounts for 48 percent of the extreme poor (USD3-a-day) in the Middle East, North Africa, Afghanistan and Pakistan (MENAAP) region, with the country’s poverty rate surging by 6.4 percentage points between 2018-19 and 2024-25, despite the country having emerged as one of the region’s economies with potential to benefit from artificial intelligence (AI), said the World Bank.

The Bank projected Pakistan’s GDP growth at 3.8 percent for 2027, below the government’s 4 percent target; inflation is projected to increase to 8.2 percent in 2027 from 7.1 percent in 2026. Pakistan’s per capita GDP growth is projected to rise marginally to 2.2 percent by 2027 from 2.1 percent in 2026, while the current account deficit is projected to widen to 0.8 percent and the fiscal deficit to 3.5 percent of GDP in 2027.

In its latest report, “Middle East, North Africa, Afghanistan and Pakistan Economic Update, from Divide to Opportunity: AI, Jobs, and Growth”, the Bank further stated that MENAAP was now the only region in the world where poverty remained above pre-pandemic levels and continued to rise, with Pakistan driving much of the deterioration, as it accounts for nearly half of the region’s extreme poor.

The increase was driven primarily by a rise in poverty in Pakistan, where the poverty rate rose by 6.4 percentage points at the USD 3.00 per day line and 3.2 percentage points at the USD 4.20 per day line between 2018-19 and 2024-25 following a succession of adverse shocks, including the COVID-19 pandemic, the devastating 2022 floods, a macroeconomic crisis marked by high inflation and currency depreciation, and a prolonged period of economic adjustment that weakened real household incomes and employment opportunities.

It said 14.3 percent of MENAAP’s population lived on less than USD 3 a day in 2024, compared with 10.4 percent globally, while 26.9 percent lived below the USD 4.20-a-day line against 18.9 percent worldwide.

Pakistan’s share of the region’s extreme poor was substantially larger than that of other countries. Afghanistan, the Syrian Arab Republic and Yemen together accounted for another 47 percent of people living below the USD 3-a-day threshold.

It projected that adverse poverty trends in MENAAP would persist through 2026, with poverty increasingly concentrated in conflict-affected and fragile economies.

The World Bank said Pakistan, along with Egypt, Jordan, Morocco and Tunisia, had technical talent and growing digital ecosystems but significant gaps in AI readiness.

Pakistan’s national AI strategy envisages training 200,000 individuals annually, including through 3,000 postgraduate scholarships, the largest absolute annual training target among the national AI strategies examined by the report.

Pakistan’s AI ambition also includes a USD 1 billion programme through 2030 covering shared GPU infrastructure, a sovereign multilingual model, 1,000 AI PhD scholarships and training for one million non-IT professionals.

The country is seeking to build on an estimated 75,000 annual IT graduates and record ICT services exports of USD 4.6 billion in FY2025/26, with AI increasingly viewed as a potential source of productivity, exports and employment.

However, the World Bank cautioned that Pakistan’s AI opportunity could remain constrained by weak innovation and economic integration despite relatively strong digital infrastructure compared with income peers.

Only 3 percent of firms in Pakistan reported product innovation and 1 percent process innovation, compared with lower-middle-income peer averages of 23 percent and 14 percent, respectively.

The report also identified a major localisation challenge. Urdu accounts for only 0.03 percent of global URLs collected by Common Crawl, compared with 0.7 percent each for Arabic and Persian, limiting the availability of local-language material for AI systems.

Frontier AI models perform broadly similarly when responding to English and Urdu prompts, the report said, but are less accurate on Urdu-centred humanities subjects, highlighting the shortage of culturally and linguistically relevant training data.

The World Bank said Pakistan’s AI adoption would therefore require simultaneous progress in digital connectivity, electricity, skills, local data, institutional capacity, regulation and private-sector investment.

It noted that AI could increase productivity in an estimated 13–20 percent of jobs across MENAAP, while near-term automation threatened less than 10 percent of jobs. However, the benefits were likely to be uneven because educated, urban and non-wage workers dominate AI-exposed occupations.

Pakistan also remains exposed to infrastructure constraints. Mobile broadband subscriptions are below the benchmark for economies with comparable income levels, while significant portions of the population still face inadequate access to electricity.

The report suggested that “small AI”—affordable, purpose-built applications capable of operating on basic mobile devices, low-bandwidth connections and intermittent power—could provide infrastructure-constrained economies such as Pakistan with a practical route to AI adoption.

Such applications could be deployed in agriculture, health and education without the computing capacity and continuous connectivity required by frontier AI systems.

The World Bank also highlighted Pakistan’s relatively strong performance in online public-service delivery, suggesting that the country could experiment with AI-enabled government services.

Meanwhile, Pakistan remains vulnerable to wider economic and climate shocks. Inflation rose to around 11 percent by midyear, while gasoline prices increased by 40 percent or more following the escalation of international energy prices. Diesel prices have risen by more than 40 percent since the start of the conflict.

Pakistan has introduced targeted fuel and farm assistance to cushion vulnerable households and sectors from the impact of higher prices.

The report said Pakistan’s growth was expected to increase from 3.2 percent in fiscal year 2024-25 to 3.7 percent in 2025-26 as services, manufacturing and livestock remained resilient despite rising import costs. Inflation is projected to increase to 8.2 percent in 2027 from 7.1 percent in 2026.

Pakistan’s real GDP per capita growth is projected to improve marginally from 1.6 percent in 2025 to 2.1 percent in 2026 and 2.2 percent in 2027, according to the Bank’s report.

Pakistan’s current account balance is projected to deteriorate from a surplus of 0.5 percent of GDP in 2025 to deficits of 0.1 percent in 2026 and 0.8 percent in 2027, while the fiscal deficit is projected to narrow sharply from 5.4 percent of GDP to 2.6 percent in 2026 before widening to 3.5 percent in 2027.

It warned that a prolonged slowdown in tourism, construction and related services could reduce labour demand and weaken income flows to labour-sending economies, particularly Pakistan and parts of the Levant.

Pakistan is also directly exposed to changing monsoon conditions, while climate-related heat stress and flood risks are adding to the country’s vulnerability.

The World Bank said targeted transfers were generally more efficient than broad subsidies and could help preserve fiscal space, particularly in countries facing significant debt and financing pressures.

Copyright Business Recorder, 2026