Neoclassical and endogenous growth: Solow, TFP and ideas

Detail

1. The starting point: why Solow was needed

• Harrod-Domar model (1940s): growth rate g = s / v.
• s = saving rate. v = capital-output ratio, known in India as ICOR (Incremental Capital-Output Ratio). ICOR is the extra capital needed to produce one extra unit of output.
• Example: s = 30%, ICOR = 4 → g = 30/4 = 7.5%.

• Its weakness: capital never runs into diminishing returns, so in this model saving more always means growing faster. Indian Five-Year Plans used this logic.

• Solow (1956) added diminishing returns and a production function. With that change, saving alone could not give growth forever.

2. Solow-Swan model (1956)

• Output per worker: y = f(k). Here y = output per worker and k = capital per worker (machines, tools and buildings for each worker).
• Diminishing returns to capital: each extra unit of k adds less output than the one before.

• Giving a farmer a first tractor raises output a lot. A second tractor adds much less.

• Steady state: the long-run point where k stops changing.

• Condition: s·f(k) = (n + δ)k.
• s·f(k) = investment per worker (the part of output that is saved).
• (n + δ)k = break-even investment, meaning the investment needed just to keep k the same.
• n = labour-force growth. New workers need new machines.

• δ = depreciation rate, the share of machines that wear out each year.

• Worked example: let f(k) = √k, s = 0.20, n = 0.02, δ = 0.03.

• 0.20·√k = 0.05·k → √k = 4 → k* = 16, y* = 4.
• Now raise s to 0.30: 0.30·√k = 0.05·k → √k = 6 → k* = 36, y* = 6.
• Income per worker rises from 4 to 6. That is a level effect.
• Once the economy reaches the new steady state, growth per worker returns to zero, because there is no technical progress in this example.

3. Capital deepening vs capital widening

• Capital deepening: k rises, so each worker has more capital.

• Each worker has more machines → labour productivity (output per worker) rises → wages can rise.

• Capital widening: capital grows only as fast as the labour force, so k stays the same.

• Investment equal to (n + δ)k is pure widening. It only keeps the economy where it is.

• For India, with a large and growing workforce, a big part of investment goes into widening. Only investment above (n + δ)k deepens capital.

4. Key results of Solow (classic traps)

• A higher saving rate raises the level of income per worker, not the long-run growth rate.

• During the move to the new steady state, growth rises for a while. This is transitional growth. It does not last.

• Long-run per-capita growth comes only from exogenous technical progress.

• Exogenous = decided outside the model. The model assumes technology improves but does not explain why.

• This gap is what endogenous growth theory later tried to fill.

• Conditional convergence prediction: a country far below its own steady state grows faster, because each unit of capital there earns high returns.

5. Growth accounting and TFP

• Growth accounting splits GDP growth into three parts:
• g_Y = α·g_K + (1 − α)·g_L + g_A
• α = capital's share of income. g_K and g_L = growth of capital and labour.

• g_A = Total Factor Productivity (TFP) growth.

• TFP is the part of output growth not explained by more capital and more labour. It reflects technology, efficiency, better management and innovation.

• Worked example: GDP growth = 7%, α = 0.4, capital growth = 8%, labour growth = 2%.
• Contribution of capital = 0.4 × 8 = 3.2 points.
• Contribution of labour = 0.6 × 2 = 1.2 points.

• TFP = 7 − 3.2 − 1.2 = 2.6 points, so about 37% of growth comes from TFP.

• Solow residual (Solow 1957): about 7/8 of the growth in US output per hour of work (1909-49) was not explained by more capital. It came from technical change.

• TFP is called a "residual" because it is what is left over after inputs are counted. Critics call it "a measure of our ignorance".

• Krugman-Young critique (1994-95): East Asia's "miracle" was mostly input-driven.

• Growth came from very high investment, more workers and more schooling, with little TFP growth.
• Input-led growth → faces diminishing returns → must slow down. The Soviet Union showed the same pattern earlier.
• Lesson: growth from "perspiration" (inputs) has limits. Growth from "inspiration" (TFP) can last.

6. India KLEMS: measuring TFP in India

• The RBI hosts the India KLEMS database. KLEMS stands for Capital (K), Labour (L), Energy (E), Materials (M) and Services (S).
• It gives Labour Productivity (LP) and TFP at the industry level and for the whole economy, from 1980-81 onwards [2].
• It covers 27 industries. The Data Manual 2018 runs from 1980-81 to 2016-17, and later releases extend the series [2].
• It is built from National Accounts Statistics (NAS), Annual Survey of Industries (ASI), NSSO rounds and Input-Output tables [2].
• It uses the EU KLEMS method, so India's figures can be compared with other countries [2].

7. Convergence

• Convergence is the idea that poorer economies grow faster and close the gap with richer ones.
• Absolute (unconditional) convergence: all poor economies catch up, whatever their policies.
• Conditional convergence: poor economies catch up only when they have similar fundamentals, such as saving, population growth and institutions. Solow predicts this type.
• β-convergence: poorer regions grow faster. β is the slope when growth is plotted against starting income, and a negative slope means convergence.
• σ-convergence: the spread of incomes (their standard deviation, σ) narrows over time.

• β-convergence is needed for σ-convergence, but it is not enough on its own. Fresh shocks can keep the spread wide.

• Indian states show divergence. Richer states such as Maharashtra, Gujarat, Karnataka and Tamil Nadu pulled ahead of Bihar and UP.

• This happened inside one common market, with one currency and free movement of goods and labour.
• Reason: fundamentals differ across states (infrastructure, human capital, governance, industry). Any convergence is at best conditional.

8. Endogenous growth theory

• Endogenous growth theory: long-run growth comes from factors inside the economy, such as human capital, R&D, innovation and knowledge spillovers (ideas leaking from one firm to others).
• Because these factors respond to incentives, policy can change the long-run growth rate. In Solow, policy changes only the level.

Romer (1990): ideas

• Ideas are non-rival: one person's use does not reduce another's use. A formula or software code can be used by millions at once.
• R&D creates new ideas → ideas are non-rival → increasing returns at the level of the whole economy.
• Ideas are partly excludable through patents. This temporary monopoly profit is the reward for doing R&D.
• Romer shared the Nobel 2018 with William Nordhaus (climate and growth).

Lucas (1988): human capital

• Human capital = the skills, education and health of workers.
• Human capital accumulates → workers become more productive → others around them also gain (spillover) → no diminishing returns overall.

AK models

• Y = AK, where A is a constant and K is broad capital (physical plus human). There are no diminishing returns.
• Growth rate g = sA − δ (minus n in per-capita form). So a higher saving rate can raise growth permanently.
• Example: A = 0.4, δ = 0.05. At s = 0.25, g = 0.10 − 0.05 = 5%. At s = 0.30, g = 0.12 − 0.05 = 7%, permanently.
• Contrast with Solow: here saving affects the growth rate. In Solow it affects only the level.

Aghion-Howitt: Schumpeterian growth

• Creative destruction (Schumpeter's term): new firms and technologies replace old ones. Growth comes through this churn.
• Their model was published in 1992 in Econometrica: "A Model of Growth Through Creative Destruction" [5].
• Nobel 2025, announced 13 October 2025, for "explaining innovation-driven economic growth" [5]:
• Half to Joel Mokyr, for identifying the conditions needed for sustained growth through technological progress. He showed that innovations follow one another only when people understand why something works, not just that it works [5].

• Half shared by Philippe Aghion and Peter Howitt, for the theory of sustained growth through creative destruction [5].

• Examples: e-commerce disrupting retail shops, streaming replacing DVD rentals, and internet ads undermining newspaper ads [6].

• Policy message: competition and easy entry and exit of firms (a smooth Insolvency and Bankruptcy Code process) help growth. Firms that already dominate may block new entrants.

9. Policy levers for India

R&D spending (GERD)

• GERD (Gross Expenditure on R&D) was 0.66% of GDP (2019-20) and 0.64% (2020-21) [4]. (NCERT scaffold: about 0.6-0.7% of GDP.)
• In money terms, GERD rose from ₹60,196.75 crore (2010-11) to ₹1,27,380.96 crore (2020-21) [4].
• Other BRICS countries (2020-21): China 2.4%, Brazil 1.3%, Russia 1.1%, South Africa 0.6% [4].
• Structural weakness: government spends about 64% of GERD and the private sector only about 36% [3][4]. In advanced economies, business usually leads R&D.

Anusandhan National Research Foundation (ANRF, Act 2023)

• Its Governing Board is chaired by the Prime Minister and gives strategic direction to the RDI Scheme [3].

Research, Development and Innovation (RDI) Scheme / Fund: ₹1 lakh crore

• Cabinet approval: 1 July 2025 [3].
• Aim: raise private sector R&D in strategic and sunrise sectors, including deep-tech start-ups [3].
• Method: long-term financing or refinancing at low or nil interest. It is not a grant [3].
• A Special Purpose Fund (SPF) inside ANRF holds the ₹1 lakh crore corpus [3].
• Link to theory: this is a Romer-style policy, because public money crowds in private idea-creation.

Other levers

• Education and skilling (Lucas human capital).
• Competition policy (Aghion-Howitt creative destruction).
• Measuring and tracking TFP through KLEMS.

Prelims Hooks

• Solow steady state: s·f(k) = (n + δ)k. The right-hand side is break-even investment (capital widening).
• Trap: in Solow, a higher saving rate raises the level of per-capita income, not the long-run growth rate. In AK models, it raises the growth rate permanently.
• In Solow, long-run per-capita growth comes only from exogenous technical progress.
• TFP = Solow residual = output growth minus the weighted growth of capital and labour. Solow (1957): about 7/8 of US output-per-hour growth (1909-49) came from technical change.
• India KLEMS (K, L, E, M, S) is maintained by the RBI. It gives LP and TFP for 27 industries from 1980-81, using EU KLEMS methodology.
• β-convergence = poor regions grow faster. σ-convergence = the spread of incomes narrows. Solow predicts conditional convergence.
• Romer (1990): ideas are non-rival. Nobel 2018 with Nordhaus. Lucas (1988): human capital.
• Nobel 2025: half to Mokyr, half to Aghion and Howitt (creative destruction, 1992 model).
• India GERD: 0.64% of GDP (2020-21). China 2.4%. Government about 64%, private about 36%.
• RDI Scheme: ₹1 lakh crore, approved 1 July 2025, run through an SPF within ANRF, whose Governing Board is chaired by the PM. It provides low or nil-interest long-term loans, not grants.

Mains Points

• Input-led vs TFP-led growth: India's high-investment push (public capex, PLI) risks the Krugman-Young trap if TFP stays weak. The goal of Viksit Bharat 2047 needs innovation-led growth: higher GERD, private R&D through the RDI Fund and ANRF, and a better-trained workforce.
• Regional divergence despite one market: state gaps show that convergence is only conditional. Lagging states need spending on human capital, infrastructure and institutions. Finance Commission transfers alone cannot close the gap.
• Creative destruction and competition policy: the 2025 Nobel suggests that easy entry and exit, strong CCI enforcement, a working IBC and openness to new technology (AI, deep-tech) drive growth. But displaced workers need safety nets and reskilling.
• Saving vs growth debate: Harrod-Domar and AK models say saving and investment drive growth. Solow says saving raises only the level. So India's falling household financial savings matter more for medium-term growth than for long-run growth, and long-run growth depends on productivity.