Econ for EverybodyReferencesDownload PDF

1 / 20

A useful idea is only the beginning.

THE ECONOMICS OF INNOVATION

Why do useful ideas become ordinary practice in some societies, while remaining isolated achievements in others?

Ashish Kulkarni Takshashila Institution · October 2026

Carefully made mechanical hubs and workshop components in warm light.
AI-generated conceptual illustration

2 / 20 · 3 minutes

Useful ideas reshape economies
as they spread.

Create

Find a useful idea.

Use

Make it work in practice.

Spread

Make adoption worthwhile.

An innovation system needs all three.

“Until many users adopt a new technology, it may contribute little to our well-being.”

Hall & Khan · author draft, p. 3

Academic sources: Hall & Khan (2003)

3 / 20 · 3 minutes

New combinations.
Not just new machines.

ONE BAKERY. FIVE WAYS TO INNOVATE.

01

The good

A new bread, or better quality.

02

The method

A different way to make it.

03

The market

A new neighborhood to serve.

04

The supply

A new source of flour.

05

The organization

A different way to organize delivery.

A freshly baked golden loaf and slices on a quiet bakery counter.
Generated illustration

Schumpeter’s five combinations widen the meaning of innovation.

Academic sources: Schumpeter (1934)Broadberry & Zhai (2025 working paper)

4 / 20 · 3 minutes

A breakthrough needs
practical improvement.

01

Possibility

Macroinventions open new technical paths.

Conceptual sequence

Practicality

Better performance
Lower cost
Greater reliability

Microinventions make a possibility more useful.

A refinement may be what finally makes adoption worthwhile.

Academic sources: Mokyr (1990)Mokyr (1998 manuscript)

5 / 20 · 3 minutes

Knowing why.
Knowing how. Better together.

Where / whenKnowing whyKnowing how
Mediterranean–Atlantic · c. 1470Celestial altitude and latitudeNavigate with a mariner’s astrolabe
France · 19th centuryMicrobes drive fermentationImprove brewing; prevent spoilage
Britain · 1830sElectromagnetic inductionBuild an electric generator
Scotland · 1860sMicrobes cause putrefactionApply antiseptic wound treatment
Japan · 1950sUnderstand statistical variationMonitor processes; reduce defects

Practice also asks questions that science learns to answer.

Academic sources: Mokyr (2005)

6 / 20 · 3 minutes

Nobody knows enough alone.

Imagined connected landscape with a laboratory, workshop, farm, and shop, illustrating knowledge held in different places
AI-generated conceptual illustration
PRINCIPLE

The scientist

Knows why it could work.

PROCESS

The technician

Knows how to make it work.

PLACE

The user

Knows the local problem.

Innovation combines knowledge dispersed across people and places.

The institutional plumbing: legal and property rights · policing · capable taxation and public spending.

Our reading of this imagined community: dispersed knowledge works better when cooperation is dependable.

Academic sources: Hayek (1945) Greif & Laitin (2004)

7 / 20 · 4 minutes

What does winning reward?

CREATE VALUE

Improve
the offer.

A better product.
A real cost saved.
A problem solved.

SECURE A TRANSFER

Protect
the position.

Exclusive permission.
Protection from rivals.
Political favor.

What the rules reward shapes where effort goes.

Baumol’s point: the rules can direct entrepreneurial effort toward production, rent-seeking, or destruction.

Create value: win customers with faster deliveries.

Capture rents: secure an exclusive licence that shuts out rivals.

Illustrative strategies. A firm can pursue both; competition over favours can waste resources as well as transfer income.

Academic sources: Baumol (1990)

8 / 20 · 3 minutes

The new arrives.
The old loses value.

Film photography

Skills, equipment, and businesses built around an old method.

Digital photography

New capabilities.
New opportunities.
Old assets lose value.

The people who gain and the people who lose may differ.

Keep experimentation possible. Take transition costs seriously.

Academic sources: Aghion & Howitt (1992)

9 / 20 · 3 minutes

A rival can spur innovation.
Or discourage investment.

CLOSE RIVALS

Escape competition

An improvement can help either firm pull ahead.

A DISTANT LAGGARD

Discouraged investment

Fiercer rivalry can make catching up less rewarding.

The effect depends on who competes, and how.

The paper combines a model with evidence from UK firms’ patents. Rivals can change incentives differently depending on how close firms are technologically. It does not tell every country how much competition to choose.

Academic sources: Aghion et al. (2005)

10 / 20 · 3 minutes

Invention is one puzzle.
Sustained growth is another.

Technical achievement
≠
sustained economic growth

China had major technical achievements long before industrialization in Europe.

Later growth differed across regions and periods.

Cold War question: why did Soviet technical achievements coexist with weaker long-run growth than in the US?
One lens: incentives and diffusion. Also consider diminishing returns to capital; invention alone cannot explain the growth gap.

How do useful possibilities become cumulative economic change?

Academic sources: Broadberry & Zhai (2025 working paper) Easterly & Fischer (1995)

11 / 20 · 2 minutes

A rejected idea needs
somewhere else to go.

An idea
Authority ARejects it
Authority BMay support it
Authority CAnother possibility
↔

Can knowledge cross the boundaries?

Alternatives matter when people and ideas can reach them.

Law did not invent central banking. William Paterson, also Scottish, proposed the Bank of England.

John Law’s travelling proposal

The Scotsman proposed banking schemes in Scotland and Turin before France permitted Banque Générale in 1716. His system collapsed in 1720: another place to experiment is an opportunity, not a guarantee of sound design.

Academic sources: Mokyr & Tabellini (2024)Mokyr (2005) Buchan (2023)

12 / 20 · 4 minutes

Institutions are a process.

Institutions are the rules and arrangements that shape behavior.

01

Rely

Keep rules dependable.
People need to plan.

02

Observe

Watch the effects.
Conditions can change.

03

Revise

Use evidence.
Use a known procedure.

Dependable rules can include dependable ways to change.

This is my design principle; institutional change is not always improvement.

“Hence, institutions can be self-undermining and the behaviors that they entail can cultivate the seeds of their own demise.”

Greif & Laitin (2004), p. 634

Academic sources: Greif & Laitin (2004)

13 / 20 · 3 minutes

Protect the idea.
Keep knowledge moving.

INCENTIVE TO CREATE

Recover
the investment.

Protection can help
reward an inventor.

ROOM TO BUILD FURTHER

Adopt. Combine.
Improve.

Access helps others
use the pool of knowledge.

What does stronger protection change at the margin?

A trade-off to investigate. Not an estimated universal optimum.

“The key is to match problems to institutions.”

Alex Tabarrok · Launching the Innovation Renaissance (2011)

Compare patents, prizes, buyouts, and advance market commitments. Further viewing: How ideas trump crises (TED, 2009).

Academic sources: Stiglitz (2014 working paper)

14 / 20 · 3 minutes

Give the state a role.
Then scrutinize the role.

The appropriate role
remains open.

  1. What would otherwise happen?Examine horizons, uncertainty, and spillovers.
  2. Can the state do this well?Examine capacity, incentives, and accountability.
  3. Can unsuccessful support end?Make withdrawal part of the design.

A non-zero role establishes neither its size nor its form.

Ashish’s perspective: Reflections on The Entrepreneurial State ↗

Academic sources: Mazzucato (2011 report)Rodrik (2004 working paper)

15 / 20 · 4 minutes

AI capability is the start.
Work has to change.

MODEL

Capability

A task becomes possible.

ORGANIZATION

Adaptation

People change skills
and workflows.

PRACTICE

Useful diffusion

The change delivers value.

Access and usage are inputs. Productivity is an outcome.

The moving frontier matters. AI capabilities can improve while organisations are still adapting to the previous generation. Adoption choices need regular review.

Kwa et al. document rapid progress on selected software tasks. This is evidence of a moving frontier, not proof of equal progress in every domain.

Academic sources: Brynjolfsson, Rock & Syverson (2021) Kwa et al. (2025; updated 2026)

16 / 20 · 3 minutes

Delay has a cost.
So does error.

WAIT TOO LONG

Miss learning.

Useful adoption and capability building may be delayed.

Training and bounded trials

MOVE WITHOUT ENOUGH CHECKS

Risk harm.

Mistakes can be costly and damage trust.

High-stakes treatment decisions

Choose the pace for the activity, not just the technology.

My judgment: India should move faster at current margins.

Education
Personalised feedback ↔ confident errors

Health
Administrative help ↔ unsafe clinical advice

Justice
Find and organise records ↔ biased summaries affecting decisions

Discussion examples: separate bounded assistance from consequential decisions, with human accountability.

Academic sources: Hall & Khan (2003)Brynjolfsson, Rock & Syverson (2021)

17 / 20 · 1 minute

Try one change in one permit office.

Proposed trial · a district office in India

Can an AI assistant reduce incomplete permit applications?

This is a hypothetical test, not a reported programme.

  1. DiagnoseAre missing documents the binding constraint—or is delay caused by staffing and approvals?
  2. TestLet AI flag missing documents on copies. Keep the officer’s decision authoritative; compare with a human-only process.
  3. ReconsiderMeasure time, missed omissions, false flags, and appeals. Expand only if service improves without worse errors.

A faster model will not fix an unrelated bottleneck.

Academic sources: Hausmann, Rodrik & Velasco (2005)

18 / 20 · 3 minutes

Skills travel through people.

01

Learn beyond kin

A master outside the family opens access to useful skills.

02

Move between workshops

Compare methods. Carry know-how to another workplace.

03

Improve in practice

Supervised work makes a useful idea usable.

British watchmaking: skilled artisans and relatively few restrictions on hiring apprentices helped turn knowledge into practical improvements.

The evidence supports adaptable training and learning beyond kin, not a classless Britain.

For today: mass university education need not reproduce supervised learning through real work. How might we combine the two?

A possible route: degree programmes with sustained work under skilled mentors. A proposal to examine, not a result from the historical paper.

Academic source: de la Croix, Doepke & Mokyr (2018)

19 / 20 · 2 minutes

Make useful ideas worth trying.
Help them spread.

An innovative society makes it worthwhile to create, experiment, and spread useful ideas.

Its institutions must also evolve as the conditions for doing those things change.

Migration & open borders → people and skills

Patents → rewards and access

AI diffusion → organisational change

Education & apprenticeship → useful know-how

Academic sources: Hall & Khan (2003)Greif & Laitin (2004) de la Croix et al. (2018)

20 / 20 · 20 minutes

Which rule would you revisit?

Which rule was once useful
but now obstructs innovation?

What evidence would justify changing it?

What would you keep dependable?

Use a concrete case. Make the mechanism explicit.

Continue with the readings ↗

References and PDF readings

Full citations for the core slides. Earlier drafts and working-paper versions are labelled. Source links support the mechanisms described; they do not establish every policy judgment in this talk. Slide 10: Broadberry and Zhai, printed pp. 28–29. Slide 11: Mokyr (2005), printed pp. 297–298 and 341–342.

Hall & Khan (2003)

Hall, Bronwyn H., and Beethika Khan (2003). Adoption of New Technology. NBER Working Paper 9730.

Publication / repositoryRead PDF

PDF: author-hosted November 2002 draft. The reference is the 2003 working paper, also published in New Economy Handbook.

Schumpeter (1934)

Schumpeter, Joseph A. (1934). The Theory of Economic Development. Harvard University Press, Chapter II, p. 66.

Publication / repository

No open original PDF confirmed. Broadberry and Zhai (2025), pp. 3–4, provide an accessible secondary discussion of the five combinations.

Mokyr (1990)

Mokyr, Joel (1990). The Lever of Riches: Technological Creativity and Economic Progress. Oxford University Press.

Publication / repository

Original 1990 book attribution. No suitable open original PDF confirmed. See Mokyr (1998 manuscript) for direct discussion of macroinventions and microinventions.

Mokyr (2005)

Mokyr, Joel (2005). The Intellectual Origins of Modern Economic Growth. Journal of Economic History 65(2): 285–351.

Publication / repositoryRead PDF

University-hosted paper. Propositional and prescriptive knowledge; not the direct source for the macro/micro distinction.

Hayek (1945)

Hayek, F. A. (1945). The Use of Knowledge in Society. American Economic Review 35(4): 519–530.

Publication / repositoryRead PDF

University-hosted journal article.

Baumol (1990)

Baumol, William J. (1990). Entrepreneurship: Productive, Unproductive, and Destructive. Journal of Political Economy 98(5, Part 1): 893–921.

Publication / repository

Publisher access may require a subscription. No author/repository open PDF confirmed.

Aghion & Howitt (1992)

Aghion, Philippe, and Peter Howitt (1992). A Model of Growth Through Creative Destruction. Econometrica 60(2): 323–351.

Publication / repositoryRead PDF

University-hosted journal article.

Aghion et al. (2005)

Aghion, Philippe, Nicholas Bloom, Richard Blundell, Rachel Griffith, and Peter Howitt (2005). Competition and Innovation: An Inverted-U Relationship. Quarterly Journal of Economics 120(2): 701–728.

Publication / repositoryRead PDF

PDF: open 2005 journal article from Harvard DASH.

Broadberry & Zhai (2025 working paper)

Broadberry, Stephen, and Runzhuo Zhai (2025). Innovation and the Great Divergence. CAGE Working Paper 771/2025, September version.

Publication / repositoryRead PDF

Working-paper version. Compare regions and periods; the paper does not establish a universal Europe–Asia ranking.

Mokyr & Tabellini (2024)

Mokyr, Joel, and Guido Tabellini (2024). Social Organizations and Political Institutions: Why China and Europe Diverged. Economica.

Publication / repositoryRead PDF

PDF is the 2023 working-paper version, IGIER 697.

Greif & Laitin (2004)

Greif, Avner, and David D. Laitin (2004). A Theory of Endogenous Institutional Change. American Political Science Review 98(4): 633–652.

Publication / repositoryRead PDF

Author-hosted journal article. Explains change; does not establish that deliberate reform always improves institutions.

Stiglitz (2014 working paper)

Stiglitz, Joseph E. (2014). Intellectual Property Rights, the Pool of Knowledge, and Innovation. NBER Working Paper 20014.

Publication / repositoryRead PDF

Theoretical working paper. Not an empirically estimated universal optimal-IP curve.

Mazzucato (2011 report)

Mazzucato, Mariana (2011). The Entrepreneurial State. Demos.

Publication / repositoryRead PDF

Academic-authored policy report. This is the 2011 report, not the later 2013 book.

Rodrik (2004 working paper)

Rodrik, Dani (2004). Industrial Policy for the Twenty-First Century. Harvard Kennedy School Working Paper RWP04-047.

Publication / repositoryRead PDF

PDF is the September 2004 manuscript; the HKS working-paper record is November 2004. A framework for scrutiny, not a direct rebuttal of Mazzucato.

Brynjolfsson, Rock & Syverson (2021)

Brynjolfsson, Erik, Daniel Rock, and Chad Syverson (2021). The Productivity J-Curve: How Intangibles Complement General Purpose Technologies. American Economic Journal: Macroeconomics 13(1): 333–372.

Publication / repositoryRead PDF

PDF is NBER Working Paper 25148 (2018, revised 2020), not the 2021 journal version.

Hausmann, Rodrik & Velasco (2005)

Hausmann, Ricardo, Dani Rodrik, and Andrés Velasco (2005). Growth Diagnostics. Working paper, March.

Publication / repositoryRead PDF

Diagnostic framework. The proposed India application is ours, not a finding about India from this paper. Official Growth Lab PDF, March 2005.

Mokyr (1998 manuscript)

Mokyr, Joel (1998). The Second Industrial Revolution, 1870–1914. August chapter manuscript, printed pp. 1–2.

Publication / repositoryRead PDF

Author manuscript; companion to the original 1990 book attribution.

Personal reflections

These are Ashish’s notes, offered alongside the academic references:

Easterly & Fischer (1995)

Easterly, William, and Stanley Fischer (1995). The Soviet Economic Decline. World Bank Economic Review 9(3): 341–371.

Publication / repositoryRead PDF

Official journal PDF. Examines Soviet growth, capital accumulation and diminishing returns. The incentives-and-diffusion lens is a discussion hypothesis, not the paper’s complete explanation.

Buchan (2023)

Buchan, James (2023). John Law and Genoa. In Andrea Zanini (ed.), Per Giuseppe Felloni, ricercatore e maestro. Memorie e scritti di storia economica. Quaderni della Società Ligure di Storia Patria 13, pp. 191–204. DOI: 10.5281/zenodo.7530126.

Publication / repositoryRead PDF

Scholarly chapter, pp. 194–199 for proposals and outcomes. Possible remembered book: Buchan, John Law: A Scottish Adventurer of the Eighteenth Century (2018); identification is tentative. Book details ↗

de la Croix, Doepke & Mokyr (2018)

de la Croix, David, Matthias Doepke, and Joel Mokyr (2018). Clans, Guilds, and Markets: Apprenticeship Institutions and Growth in the Preindustrial Economy. Quarterly Journal of Economics 133(1): 1–70.

Publication / repositoryRead PDF

Author-hosted journal article. See pp. 6–7, 14–18 and 63 on learning beyond kin, apprenticeship flexibility and watchmaking. The modern university comparison is a discussion judgment.

Kwa et al. (2025; updated 2026)

Kwa, Thomas, Ben West, et al. (2025). Measuring AI Ability to Complete Long Software Tasks. NeurIPS 2025; arXiv:2503.14499v4, revised 10 July 2026.

Publication / repositoryRead PDF

Academic paper; linked version is the updated author preprint. Selected software tasks, not a universal measure of capability or organisational adoption.

Tabarrok (2011) · further reading

Tabarrok, Alex (2011). Launching the Innovation Renaissance: A New Path to Bring Smart Ideas to Market Fast. TED Books.

Publication / repositoryRead PDF

Author-hosted book. Complements the Stiglitz academic anchor on slide 13. The quoted sentence follows a comparison of innovation-support institutions.

Revision record

3 October 2026 — slides 1–20: removed all speaking notes and cover duration; reduced navigation; added rounded image corners and bread imagery; centred slides 4, 5, 7, 9 and the conclusion. Added verified quotations, historical why/how examples, institutional context, John Law’s banking episode, Cold War discussion, AI pace and domain examples, a concrete hypothetical permit-office trial, and a new apprenticeship slide. Added academic references and Ashish’s public state-innovation reflection. Rebalanced teaching time to retain the 75-minute session.

3 October 2026 — visual revision after Ashish’s review: rebuilt the 19 slides with point-making titles, varied compositions, locally hosted Fraunces and Source Serif typography, and two labelled generated conceptual illustrations. Kept all speaking notes, timing, academic references, and evidence boundaries. The design review tightened the diffusion headline to avoid treating widespread adoption as a condition for any economic effect.

3 October 2026 — prepared the October version from the February presentation and the subsequent discussion. Added the rewards-to-effort lens, institutional revision, and an activity-specific approach to AI adoption. Kept the role of the state open. Added academic sources, version labels, examples, and evidence boundaries to every substantive slide.

3 October 2026 — redesigned the October deck after content approval: message-led titles, varied mechanism diagrams, local presentation fonts, and labelled conceptual imagery. Preserved source boundaries, speaking notes, and timing.

Approved for publication by Ashish on 3 October 2026.