Swimming in Sand: Why Productivity Distortions Are Holding Pakistan Back

Key Takeaways

  • Pakistan's real GDP per capita grew by only 1.9% per year from 2000-2020, compared to 4.4% for South Asia as a whole.[1]
  • Policy-induced distortions impose a substantial productivity penalty: the World Bank estimates that eliminating allocative distortions could boost aggregate productivity by approximately 18%, while removing entry barriers could add another 12%.[4]
  • Nearly 40% of Pakistan's workforce remains in agriculture, yet manufacturing labor productivity is only about 30% higher than agriculture—far below the gaps seen in successful structural transformations.[2][4]
  • Among comparable developing economies, Pakistan's GDP per capita growth ranked second-lowest from 2000-2020, outperforming only Mexico and falling well behind Bangladesh (4.8%), Vietnam (5.2%), and Ethiopia (5.7%).[1]

The Productivity Puzzle

For over two decades, Pakistan's economy has been "swimming in sand." Despite periods of respectable GDP growth, the country has failed to generate sustained productivity gains that lift living standards over time.

From 2000 to 2020, Pakistan's real GDP per capita grew at an average rate of 1.9% per year.[1] This is less than half the South Asia regional average of 4.4% over the same period.[1] The gap is even starker when compared to high-performing peers: Vietnam averaged 5.2%, Ethiopia 5.7%, and Bangladesh 4.8%.[1]

This sluggish performance cannot be attributed to a single crisis or external shock. Pakistan's underperformance is structural and persistent. In the two decades examined, per capita GDP actually contracted in three separate years: 2008 (-0.6%), 2010 (-1.0%), and 2020 (-3.0%).[1]

The World Bank's 2022 Country Economic Memorandum, titled "Swimming in Sand," identifies a central culprit: policy-induced distortions that prevent resources from flowing to their most productive uses.

Pakistan's GDP Per Capita Growth (2000-2022)

Annual % change in real GDP per capita

Source: World Bank Open Data (NY.GDP.PCAP.KD.ZG). Accessed February 2026.

What Are Productivity Distortions?

In an efficiently functioning economy, capital and labor flow toward firms and sectors where they can generate the highest returns. Productive firms expand, unproductive firms shrink or exit, and new entrants constantly test the market with innovative approaches.

In Pakistan, this process is impaired by two categories of distortions:

Allocative distortions arise when policy interventions cause resources to be misallocated within an industry. Examples include size-dependent regulations that penalize firm growth, subsidies targeted at specific firms or sectors regardless of productivity, and tax enforcement that varies by firm size or political connections. These policies protect less productive firms while constraining the growth of more productive ones.

Entry barrier distortions prevent new firms from entering markets where they could compete effectively. High regulatory compliance costs, restricted access to infrastructure and utilities, and opaque licensing requirements create what economists describe as an implicit "entry tax."

The World Bank estimates that entry barriers in Pakistan are equivalent to an entry tax rate of approximately 400%—meaning a potential entrant faces costs four times higher than in an undistorted market.[4]

Quantifying the Costs of Misallocation

Using firm-level data from Pakistan's Census of Manufacturing Industries and a framework developed by economists Chang-Tai Hsieh and Peter Klenow, the World Bank estimates the productivity costs of these distortions.

According to their analysis, eliminating allocative distortions within sectors would increase aggregate productivity by approximately 18%.[4] Removing entry barrier distortions would add another 12%.[4]

Combined, these distortions impose a productivity penalty of roughly 30% on Pakistan's manufacturing sector. This is a substantial figure. Eliminating these distortions would be equivalent to more than a decade of normal productivity growth.

It is important to note that these estimates are derived from economic models applied to a single cross-section of firm data (the Census of Manufacturing Industries 2015/16). They represent potential gains under idealized conditions, not guaranteed outcomes of any specific policy reform.

The Missing Manufacturing Transformation

A hallmark of successful economic development is structural transformation: the movement of workers from low-productivity agriculture to higher-productivity manufacturing and services. This shift drives aggregate productivity growth even when within-sector productivity is stagnant.

Pakistan's structural transformation has been notably slow. As of 2020, approximately 38% of the workforce remained employed in agriculture.[2] This share declined only modestly from 42% in 2000.[2] By comparison, Bangladesh, starting from a similar base, has shifted workers into manufacturing at a much faster pace.

More troubling is the limited productivity gap between sectors. According to the World Bank analysis, labor productivity in manufacturing was only about 30% higher than in agriculture.[4] In services, the gap was larger at around 150%.[4]

This narrow gap matters because it reduces the productivity gains from structural transformation. If workers moving from agriculture to manufacturing see only modest productivity increases, the economy's overall efficiency gains are limited.

The manufacturing sector's share of GDP has also stagnated. In 2000, manufacturing value added represented 9.1% of GDP; by 2020, it had risen to only 11.4%.[3] This modest increase over two decades contrasts sharply with the rapid industrialization seen in Vietnam, Bangladesh, and other high-growth economies.

Within-Firm Productivity Decline

If structural transformation has been slow, one might hope that firms within sectors were at least becoming more efficient. The World Bank's analysis suggests the opposite.

Manufacturing sector within-firm productivity growth averaged only 0.9% per year from 2000 to 2018, compared to a 2.5% average among peer countries.[4] Services sector productivity growth was 1.3% per year, below the 1.9% peer average.[4]

The COVID-19 pandemic delivered an additional shock. Within-firm productivity contracted by 23% in 2020, according to the World Bank's estimates.[4]

Why are Pakistani firms becoming less efficient over time? The World Bank points to several factors:

Family ownership structures that prioritize control over growth and professionalism. The analysis found that family-owned firms showed sharper productivity declines than non-family firms.

Limited technology adoption due to high costs of capital and uncertain economic conditions that discourage investment.

Skill mismatches between what educational institutions produce and what firms need.

Weak competitive pressure resulting from the same entry barriers that protect incumbents from new entrants.

Barriers to Reallocation

The distortions described above prevent the natural process of resource reallocation that drives productivity growth in healthy economies. When entry barriers protect incumbents and allocative distortions reward connections over efficiency, the competitive pressure that typically forces resources toward more productive uses is weakened.

This manifests at the firm level as well. Unproductive firms persist longer than they should, consuming capital and labor that could be deployed more effectively elsewhere.

Cross-Country Context

Pakistan's productivity underperformance is not a matter of definition or measurement quirks. The gap is visible across multiple comparisons.

Among the set of structural and aspirational peers examined by the World Bank, Pakistan's 2000-2020 GDP per capita growth of 1.9% ranked second-lowest:[1]

Country GDP per capita growth (2000-2020)
Ethiopia5.7%
Vietnam5.2%
Bangladesh4.8%
India4.4%
Indonesia3.6%
Turkey3.6%
Egypt2.3%
Pakistan1.9%
Mexico0.0%

Ethiopia, starting from a much lower base, grew three times faster. Bangladesh, Pakistan's closest structural peer in terms of initial conditions and economic structure, grew more than two and a half times faster.

The gap with India is particularly notable given the countries' shared history and geographic proximity. India's 4.4% average growth rate was 2.5 percentage points higher than Pakistan's, compounding over two decades into a substantial divergence in living standards.

GDP Per Capita Growth: Pakistan vs. Comparators (2000-2020 Average)

Annual % average

Source: World Bank Open Data (NY.GDP.PCAP.KD.ZG). Calculated as arithmetic mean of annual growth rates. Accessed February 2026.

What the Data Cannot Tell Us

Causation versus correlation

The estimates of productivity losses from distortions are based on cross-sectional patterns in firm data. They identify correlations between policy environments and productivity outcomes but cannot definitively establish causation. Other factors—including macroeconomic instability, security conditions, and governance quality—may contribute to both the existence of distortions and low productivity.

The political economy of distortions

The data show that distortions exist and estimate their costs, but they cannot explain why such costly policies persist. The insider-outsider dynamics that maintain distortions, including the political influence of protected firms and sectors, are not captured in productivity statistics.

Implementation gaps

Even when pro-competitive reforms are enacted on paper, implementation may be uneven. The difference between de jure and de facto policy environments is substantial in Pakistan but difficult to measure systematically.

The informal economy

All firm-level analysis in this study is limited to the formal sector. The informal economy, which by some estimates accounts for more than 30% of GDP and employs the majority of non-agricultural workers, is largely unmeasured. Productivity dynamics in this sector may differ substantially from those observed in formal firms.

Model dependence

The specific figures for allocative and entry barrier distortions (18% and 12% respectively) are derived from economic models with specific assumptions. Different modeling choices would yield different estimates. These figures should be understood as indicative magnitudes rather than precise measurements.

Data vintage

The firm-level analysis relies primarily on the Census of Manufacturing Industries 2015/16 and SECP data through 2019. Conditions may have changed in subsequent years.

Data Notes

  1. World Bank Open Data. Indicator: NY.GDP.PCAP.KD.ZG (GDP per capita growth, annual %). Coverage: Pakistan, South Asia aggregate, Bangladesh, India, Indonesia, Turkey, Mexico, Egypt, Ethiopia, Vietnam. Years: 2000-2022. Accessed: 2026-02-22. data.worldbank.org
  2. World Bank Open Data. Indicator: SL.AGR.EMPL.ZS (Employment in agriculture, % of total employment, modeled ILO estimate). Coverage: Pakistan. Years: 2000-2022. Accessed: 2026-02-22. data.worldbank.org
  3. World Bank Open Data. Indicator: NV.IND.MANF.ZS (Manufacturing, value added, % of GDP). Coverage: Pakistan. Years: 2000-2022. Accessed: 2026-02-22. data.worldbank.org
  4. World Bank Country Economic Memorandum. "Swimming in Sand: Why Pakistan Needs to Break with its Past to Achieve a Better Future." World Bank, 2022. Chapters 1-2 and Overview used for firm-level productivity analysis, distortion quantification, and sectoral productivity comparisons. The 18% allocative distortion and 12% entry barrier estimates are model-derived using the Hsieh-Klenow (2009) and Fattal Jaef (2019) frameworks applied to Census of Manufacturing Industries 2015/16 data. These estimates carry MEDIUM confidence due to model dependence and data limitations.
  5. Methodology notes. Cross-country averages are arithmetic means of annual growth rates over the period 2000-2020 (21 observations). Pakistan's verified average is 1.89%, which we round to 1.9% for presentation. The World Bank CEM reports 1.7%, which may reflect a different calculation method (compound annual growth rate) or data vintage. The directional finding (Pakistan significantly underperforming peers) is robust to these differences.