Three Growth Paths: Reading India, Pakistan and China Side by Side

Deepest transformation - China. Agriculture is down to 7% of GDP while industry is 39% and per capita income is $12,600, roughly 5.3 times India's and 8.4 times Pakistan's. The mechanism: workers moved out of low-productivity agriculture into manufacturing, where capital per worker and hence output per worker is far higher. This is structural change in the classic sense - the shift of labour between sectors, not just growth within them, is what multiplies output. China's high investment rate, infrastructure build-out and export orientation financed that absorption.

Strongest record among the other two - India, but with a caveat. The sectoral shares look advanced (services 56%), yet agriculture still contributes 18% of GDP while employing roughly 45% of the workforce. That gap is the signature of disguised unemployment and low agricultural productivity: too many workers on too little land, producing little each. India has leapfrogged into services without absorbing labour in manufacturing (industry stuck at 26%), so growth has been capital- and skill-intensive, not employment-intensive.

Weakest - Pakistan. Agriculture 24% of GDP but industry only 19%: the country has not passed through the manufacturing phase at all. This is premature deindustrialisation - labour leaving farming for low-productivity urban services, not for factories. With per capita income at $1,500 and growth at 2.4% against population growth of 2.0%, per capita income is nearly stagnant.

The judgement to present: the key discriminator is not sectoral shares alone but whether the movement of labour raised output per worker. China's did; India's did partly, through services; Pakistan's largely did not.

Demography is a potential, not a dividend. A dividend occurs only when the growth of the working-age population is matched by growth of productive employment and by human capital. Otherwise the same cohort becomes a liability - unemployed youth, urban pressure, fiscal stress.

Pakistan. Median age 20 and population growth 2.0% mean the labour force grows about 2% a year. But growth is 2.4%, barely above population growth, so per capita income is flat; the investment rate is too low to create jobs at that rate, and an IMF programme constrains public spending. Large cohorts therefore enter the labour market without the capital or skills to be productively employed. Verdict: the demographic window is open and being missed - it will close within two to three decades and cannot be reopened.

India. Median age 28 with population growth 0.9% - the classic window, and a longer one. But the dividend has preconditions India has only partly met: female labour force participation remains low, skilled labour is scarce, and manufacturing has not absorbed low-skilled workers. India's advantage is that its window lasts longer and its growth rate is higher; its risk is that the window is squandered, as it partly has been.

China. Median age 39, population shrinking. China has already banked its dividend - the high savings and investment of the past three decades were largely produced by a large working-age population with few dependants. It now faces the reverse: a rising old-age dependency ratio, a shrinking labour force, and the need to raise productivity per worker rather than add workers.

Judgement: the sequence matters. China converted its window into capital and income; India still can; Pakistan risks passing through the window without creating the jobs or the skills. A young population with no jobs is not an asset on any balance sheet.

Start by rejecting the false binary. All three had heavy state involvement. China's post-1978 reform was gradual and dual-track: special economic zones, household responsibility systems in agriculture, township and village enterprises, and export processing - market entry alongside the state sector rather than instant privatisation. India ran a planned mixed economy with import substitution and licensing from 1950 to 1990, then liberalised, privatised and globalised from 1991, but retained substantial public provision. Pakistan oscillated between state-led five-year plans and repeated IMF stabilisation programmes.

Differences that actually matter:

1. Investment rate and its financing - sustained gross capital formation of 35–45% of GDP in China versus much lower in India and Pakistan.
2. Export orientation versus import substitution - exporting forced firms onto world prices and quality standards; import substitution sheltered them. India's 1991 reforms moved it towards the former, which is when growth accelerated.
3. Human capital - China's early gains in basic education and health; India's and Pakistan's weaker and uneven outcomes. Life expectancy 78 versus 70 and 66 is a summary statistic for decades of human-capital investment.
4. Macroeconomic stability and tax capacity - Pakistan's low tax base and repeated balance-of-payments crises shorten every policy horizon; India's 1991 crisis was the trigger for durable reform.
5. Political continuity and implementation capacity - reforms that survive a change of government compound; those reversed do not.

Ranked first: investment in productive capacity, financed domestically, aimed at tradeable goods. It is the variable through which all the others act: education without investment produces emigrants, and investment without education produces unemployable labour.

Caveat for the presentation: the data are approximate and a single year cannot establish a trend; growth in one year also reflects the base and the business cycle (Pakistan's 2.4% is constrained by a stabilisation programme). Conclusions should be phrased as patterns, not verdicts on a single observation.