Structural transformation of the labour productivity, 2000–2022

A shift-share analysis shows structural change has been productivity-neutral for two decades — and actively growth-reducing since 2017

By Dr Stephen LARTEY

For more than two decades, Ghana’s workers have been leaving the farm. Agriculture’s share of employment has fallen from 53 per cent in 2000 to 40 per cent in 2022 — a shift of 13 percentage points representing millions of people.

In any standard account of economic development, this should be cause for optimism. When workers move from low-productivity farming into higher-productivity manufacturing and modern services, the whole economy becomes more efficient. This is structural transformation at work, and it is the engine behind the East Asian economic miracle.

But in Ghana, something has gone wrong. A detailed statistical analysis of the country’s labour productivity record from 2000 to 2022 — using data from the GGDC/UNU-WIDER Economic Transformation Database and Ghana’s inaugural National Productivity Statistics Report published by the Ghana Statistical Service (GSS) in January 2025 — reveals a troubling picture.

The workers leaving agriculture are not going to manufacturing. They are going to construction sites and informal street trading. And since 2017, this pattern has stopped being a missed opportunity. It has become a drag on the entire economy.
“Workers leaving agriculture went to trade and urban services — not manufacturing. Since 2017, this has become a drag on the entire economy.”

The Mathematics of Structural Change
The analytical framework used here — the McMillan-Rodrik shift-share decomposition — breaks down economy-wide labour productivity growth into three components. The within-sector effect captures how much productivity grows because workers are becoming more productive within their existing sectors.

The between-sector effect captures how much productivity grows because workers are moving from lower to higher productivity sectors — this is the structural transformation term. When the between-sector effect is negative, the movement of workers between sectors is actively making the economy less productive.

Figure 1: Sectoral Employment Shares — Ghana, 2000–2022

Source: GSS Table 1.2 (Census/GLSS/AHIES).
Note: pre- and post-2013 shares are not strictly comparable due to the 19th ICLS definition change.

Employment Trends: Who Is Moving Where?
Figure 1 plots Ghana’s sectoral employment shares from 2000 to 2022. The structural shift is visible: agriculture’s dominance declining, services expanding. But the composition of that services expansion is the crucial detail the chart reveals.

The rise in services has not been driven by finance, ICT, or professional services — the sectors with the highest productivity. It has been driven by informal trade, construction, and other urban services — sectors whose productivity growth is among the weakest in the economy.

Between 2000 and 2011 — the ETD database period — the agriculture sector’s share fell from 53 to around 50 per cent. The workers who left did not go to manufacturing, which remained almost exactly flat at 10 per cent throughout the decade.

They went primarily to trade and repair services, characterised by low and stagnant productivity. The oil windfall of 2006 to 2011 raised the between-sector contribution modestly to 4.9 per cent as the extractives sector expanded — but this was a capital-intensive enclave effect, not genuine structural transformation.

Figure 2: Sectoral Employment Shares — Ghana, 2000–2011 (ETD, 7 sectors)

Source: GGDC/UNU-WIDER Economic Transformation Database (2014 release); author’s calculations.
Labour Productivity: The Sector Gap
Figure 3 maps the LP index for each sector in 2000 and 2011 — the economy average equals 100. The extractives sector sits at around 600, reflecting capital-intensive oil and mining operations. Modern services are above average at around 140.

But all the sectors absorbing the bulk of Ghana’s labour — agriculture, trade services, manufacturing — sit below 100. Workers leaving agriculture were moving between below-average sectors, which explains why the between-sector effect was near-zero rather than growth-enhancing.

Figure 3: Sectoral Labour Productivity Index — Ghana, 2000 vs 2011 (Economy avg = 100)

Source: GGDC/UNU-WIDER Economic Transformation Database; author’s calculations.
Right axis: Extractives (off-scale for main chart).
TABLE 1: GHANA SHIFT-SHARE RESULTS, 2000–2022
Period Source Between-Sector (% of total LP growth)
2000–2006 ETD (7-sector) 1.6% — Growth-enhancing
2006–2011 (oil build-up) ETD (7-sector) 4.9% — Growth-enhancing
2000–2011 (full) ETD (7-sector) 1.4% — Growth-enhancing
2013–2017 (pre-crisis) GSS (3-sector) 8.5% — Growth-enhancing
2017–2022 (COVID + debt) GSS (3-sector) −27.2% — GROWTH-REDUCING
2013–2022 (full GSS) GSS (3-sector) 5.6% — Growth-enhancing
Source: GGDC/UNU-WIDER ETD (2014); GSS Productivity Report (January 2025); author’s calculations.

The ETD Decomposition: Within-Sector Dominance
Figure 4 shows the shift-share decomposition for each ETD sub-period as a stacked bar chart. The dominance of within-sector gains — the blue bars — is unmistakable. Between-sector reallocation, the orange bars, is barely visible across all three periods.

Between 2000 and 2011, 92 per cent of Ghana’s labour productivity growth came from within-sector improvements — workers in each sector becoming more productive in place. The between-sector component contributed just 1.4 per cent. The economy grew, but it grew through effort and investment within existing sectors, not through the structural movement of labour to where it could be most productive.

Figure 4: McMillan-Rodrik Shift-Share Decomposition — Ghana, 2000–2011

Source: GGDC/UNU-WIDER ETD (2014); author’s calculations. All values annualised pp/yr.
Between-sector: 1.4–4.9% of total across all sub-periods.

A Brief Window — Then Collapse
The GSS data covering 2013 to 2022 reveals a more nuanced and ultimately more alarming story. For a brief period between 2013 and 2017, structural transformation was genuinely growth-enhancing, contributing 8.5 per cent of total productivity growth.

Aggregate labour productivity rose from GH₵7,927 per worker in 2013 to GH₵13,496 in 2017 — 70 per cent growth in four years. Workers, for once, were moving toward slightly more productive activities.Then came the collapse. By 2022, aggregate LP had fallen to GH₵11,961 — a decline of 11.4 per cent from the 2017 peak.

he between-sector component turned sharply negative at minus 27.2 per cent. The structural change that was occurring — in a period of COVID-19 disruption and the most severe debt distress in Ghana’s post-independence history — was actively destroying productivity.

Figure 5: McMillan-Rodrik Shift-Share Decomposition — Ghana GSS, 2013–2022

Source: GSS National Productivity Statistics Report (January 2025); calibrated 3-sector LP levels; author’s calculations. Note: GHC constant 2013 prices. Negative values for 2017–2022 reflect the LP decline from GH₵13,496 to GH₵11,961.
Where Are the Workers Going? The Sub-Sector Picture

Figure 6 plots each of Ghana’s 17 sub-sectors on the basis of employment growth (vertical axis) against labour productivity growth (horizontal axis), with circle size proportional to employment share. The picture is stark.

The upper-left quadrant — high employment growth, low LP growth — is dominated by Other Service Activities (+11.8 per cent employment per year, only 1.3 per cent LP growth) and Construction (+4.7 per cent employment, 2.1 per cent LP growth).

These are where Ghana’s workers are going. The lower-right quadrant — high LP growth, low or negative employment — contains Manufacturing (11.7 per cent LP growth, only +0.2 per cent employment) and Mining (21.7 per cent LP growth, minus 3.3 per cent employment). These are where the productive growth is happening, without the people.

Figure 6: Sub-Sector Employment vs Labour Productivity Growth — Ghana, 2013–2022

Source: GSS Table 2.1, Productivity, Employment and Growth National Report (January 2025). Circle size proportional to 2022 employment share. LP growth: real, average annual. Electricity (58.5% LP growth, 27.3% emp growth) is partially off-scale.

TABLE 2: WHERE GHANA’S WORKERS ARE GOING
— TOP SUB-SECTORS BY EMPLOYMENT GROWTH (2013–2022)
Sub-sector Emp. growth/yr LP growth/yr (real)
Other Service Activities +11.8% 1.3%
Construction +4.7% 2.1%
Finance & Insurance +17.6% 5.1%
Corporate Agriculture +28.3% 18.0%
Manufacturing +0.2% 11.7%
Trade & Repair −1.4% 15.0%
Mining & Quarrying −3.3% 21.7%
Source: GSS National Productivity Statistics Report, Table 2.1 (January 2025).

The New Structural Economics Diagnosis
Justin Yifu Lin’s New Structural Economics framework provides the theoretical lens for these findings. Lin argues that every country has a latent comparative advantage determined by its factor endowments. For Ghana — with abundant low-cost labour and proximity to West African markets — that advantage lies in labour-intensive manufacturing and agro-processing.

The country has the endowments to compete in garments, light assembly, processed food, and agricultural value chains. Lin’s central insight is that markets alone cannot make this happen. Firms in labour-intensive manufacturing face coordination problems that no individual business can solve: reliable electricity, trained workers who will not leave, co-located suppliers.

Without government coordination — industrial parks, targeted infrastructure, incentive alignment — comparative advantage remains latent. Ghana’s data is a textbook illustration. Manufacturing achieved extraordinary productivity growth of 11.7 per cent per year, but employment barely moved because the enabling environment for manufacturing to scale never materialised.

“Manufacturing achieved 11.7 per cent real productivity growth per year — the strongest of any large sector — but employment barely moved at just +0.2 per cent annually.”

What the 2022 Debt Crisis Did
The severity of the 2017-to-2022 reversal cannot be overstated. A negative between-sector component of minus 27.2 per cent means that the movement of Ghanaian workers between sectors — in aggregate — subtracted nearly a third of what workers could have produced had they remained in their existing activities. The debt crisis destroyed formal employment.

Firms in manufacturing, finance, and transport contracted. Workers entered the urban informal economy, swelling construction and street trading. This is not merely a cyclical story. Even the full 2013-to-2022 period, which includes the better pre-crisis years, produced a between-sector contribution of only 5.6 per cent. Ghana enters 2026 with a structurally weaker labour market than it had in 2017, at a moment when structural strength is most needed to service restructured debt and generate sustainable tax revenues.

What Must Change
The policy implications are not complicated, even if implementation is. Ghana needs structural transformation that is genuinely growth-enhancing — workers moving into sectors where they produce more, not less, than the economy average. This requires, first, resolving the energy crisis that makes manufacturing uncompetitive.

Second, scaling industrial policy: the 1D1F initiative pointed in the right direction but operated at insufficient scale. Third, redirecting commodity revenues from fiscal consumption toward productive investment. Fourth, the national dialogue on productivity-linked wages that the GSS itself called for at the launch of this report is now urgent — the data show earnings have consistently lagged productivity growth across all three broad sectors since 2013.

Conclusion
For twenty-two years, Ghana’s economy has grown primarily because workers in agriculture, manufacturing, and services have become more productive where they stand. That is good news, and it should not be discounted. But it is half of what development requires.

The other half — structural transformation, the movement of labour from lower to higher productivity activities — has been near-zero for most of this period and actively negative since 2017. Until the coordination failures that prevent manufacturing from scaling are systematically addressed, Ghana’s workers will keep going to the wrong places. And the country’s productivity record will keep telling the same story it has told for two decades.

About the Author
Dr Stephen Lartey is a development economist with a PhD in Economics specialising in institutions, fiscal policy, monetary and macroeconomic policy, and causal inference. The author can be reached at [email protected]

DATA SOURCES
1. GGDC/UNU-WIDER Economic Transformation Database (10-Sector, 2014 release). Available at: rug.nl/ggdc/structuralchange
2. Ghana Statistical Service (2025). Productivity, Employment and Growth — National Report 2024. Accra: GSS. January 2025. With ILO technical support and funding from Switzerland and Norway.
3. McMillan, M. and Rodrik, D. (2011). Globalization, Structural Change and Productivity Growth. NBER Working Paper 17143.
4. Lin, J.Y. (2021). New Structural Economics: A Framework for Rethinking Development. World Bank Publications.

Post Views: 11