Resources
A working glossary of the economic ideas behind how technologies actually land, each tied to a case we have written about.
Transitions are usually discussed in the language of technology: what is newer, cleaner, cheaper, more advanced. But whether a technology actually reaches the people it was built for is decided far more often by economics than by engineering. Who bears a cost, who captures a value, how a cash flow is shaped, what a market will and will not price, which risk is real and which is merely perceived, these are the questions that decide a transition, and they are economic questions.
This page is a plain-language map of those ideas. Each concept is defined in a sentence or two, and then tied to a published piece where we applied it to a specific, real case, so the abstraction stays anchored to something that actually happened. It is meant to be read either straight through, as a primer on the economics of transitions, or dipped into, as a glossary.
Read this page alongside the human side of technology, which reads the same terrain from the behaviour-and-perception side; the technology & innovation dynamics companion, which takes it from the management-of-technology side; and the transitions primer, which sets out the socio-technical frame in full. A handful of foundational ideas, path dependence, network effects, increasing returns, appropriability, deliberately appear on more than one page, because they genuinely belong to more than one literature; each page treats them through its own lens and points to the others.
Once a system is established, the cost of switching away from it rises over time, so an inferior option can persist simply because it arrived first and everything else organised around it. The formal result is due to Arthur (1989), who showed that under increasing returns an economy can lock in to a technological path that is neither efficient, predictable, nor easily shifted by tax or subsidy, and to David (1985), whose study of the QWERTY keyboard made path dependence concrete. Lock-in is the central reason good technologies fail: the barrier is rarely the artefact, it is the system it must displace. We set out the full mechanism in the transitions primer, and its two-directional form, where the same barrier that keeps a country out of a market is also what makes entering worthwhile, in The Lock-In Runs Both Ways. Read this alongside the strategic reading of the same result on the technology & innovation dynamics page, where an early lead in a path-dependent technology is worth fighting for out of proportion to its immediate value.
Change arrives from protected niches, is resisted by an entrenched regime, and is pushed or slowed by a wider landscape of prices, politics, and crises. This multi-level view is the backbone of transition studies, and the frame beneath most of the concepts below.
A new technology can get past an incumbent system in one of three ways: by building a parallel route around it, by redirecting part of the existing system to a new use, or by eroding the foundations that hold the incumbent in place. Which mechanism fits depends on the barrier. This is the Lab's signature typology, set out in The BRW Framework.
Reading state capacity against niche strength sorts every real transition into one of four patterns: directed, coordinated, stalled, or bounded leapfrogging. The diagnostic, and how the Lab reads its own cases against it, is in Four Ways a Transition Lands.
A practice can be profitable over ten years and unaffordable in year two. The transition itself is the most vulnerable moment, because costs arrive before benefits, and the dip in between is where most attempts fail. We trace this through European farms in The Trough Before the Dividend. The behavioural cousin, why loss aversion makes the trough feel deeper than the numbers suggest, is on the human side of technology page.
Transitions rarely proceed by clean replacement. More often a new source is added alongside the old one, and the incumbent is retained as backup, so the picture is one of accumulation rather than substitution. We read this in African solar in Stacking, Not Switching.
Some technologies get more valuable the more they are used: each new adopter improves the infrastructure, deepens the supply of skills, and raises the credibility of the option for the next adopter. Formalised by Arthur (1989), this is the engine that lets a niche eventually overtake an incumbent, and the reason the early going is so slow, the returns only compound once a threshold is crossed. The idea has a long lineage, reaching back to Marshall (1890) on the external economies that accrue to a growing industry. The strategic reading of the same mechanism, as the engine that lets a niche cross the chasm, is on the technology & innovation dynamics page.
A niche does not break through on its own schedule; it breaks through when a shock, a price spike, a policy shift, a crisis, opens a window in the regime. A technology that is ready when the window opens succeeds, and one that is not may wait years for the next. The transition turns on readiness meeting timing, not merit alone.
A firm must decide which parts of its value chain to own and which to source from others. The choice turns on transaction costs: when depending on an outside party is too risky or too variable, a firm integrates. This is the founding question of the theory of the firm, posed by Coase (1937) and developed into transaction-cost economics by Williamson (1985), who tied the make-or-buy decision to uncertainty, frequency, and asset specificity. We apply it to battery-swap network architecture in Own the Battery, Rent the Shopfront, and to make-or-buy under uncertainty in The Anchor Tenant. The capability-strategy reading of the same decision, why outsourcing can hollow out the ability to build the next generation, is on the technology & innovation dynamics page.
Whoever owns the demand can finance the infrastructure. Platforms solve the chicken-and-egg problem of new infrastructure by controlling the customer relationship, which works, and quietly concentrates several dependencies on one counterparty. Set out in The Anchor Tenant: How Platforms Became the Utility Nobody Elected.
A product whose value to each user rises with the number of other users, a phone, a payment rail, a charging standard, tends toward a single winner, because adoption is self-reinforcing. Katz and Shapiro (1985) formalised network externalities and showed why compatibility and sponsorship decide these markets. Network effects are the flip side of lock-in: the same mechanism that helps a niche reach critical mass then makes the resulting incumbent very hard to dislodge. The competitive-strategy reading, of standards battles won on timing and compatibility, is on the technology & innovation dynamics page.
When a large customer can leave a shared system and supply itself, the economics of the system for everyone left behind change sharply. Whether self-supply strengthens or hollows out the grid is decided in tariff design, which we examine in The Customers Who Can Leave.
The friction that stops firms trading and depending on each other often sits inside a country, not at its border: unreliable power, warehousing, paperwork, and above all variance. It is variance, not average price, that prevents firms relying on one another. We show this for African trade in Behind the Border.
The right kind of money depends on the shape of the thing being financed. Equity suits uncertain, high-upside ventures; working capital suits predictable, short-cycle needs. Using the wrong instrument distorts what a founder optimises for. We work through this for a Nigerian agritech in Equity Is the Wrong Money for a Warehouse, and the mismatch between infrastructure economics and a ten-year venture fund in One Month Is Not a Trend.
An asset class is created when a stream of repayments can be made into collateral. Off-grid solar became bankable because a missed payment could be enforced by switching the light off remotely; assets without an equivalent enforcement mechanism, like water, cannot copy the model. Explained in What the Bond Is Actually Secured On.
When one side of a deal knows more than the other, the better-informed party can exploit the gap, and in the extreme the good options are driven out and only the bad remain. Akerlof's (1970) "market for lemons" is the founding statement, and it sits underneath much transition finance: a lender who cannot tell a good borrower from a bad one prices for the average, which pushes the good borrower away. Enforcement mechanisms, collateral, and independent verification all exist to close this information gap.
When quality cannot be observed directly, the informed party can spend on a costly, hard-to-fake signal, a certification, a guarantee, a track record, to prove it. Spence (1973) formalised signalling, and it explains why credible proof is worth paying for in exactly the markets where asymmetric information bites: the signal is not the quality, but it is the only way the quality becomes visible.
Blended-finance instruments like first-loss capital address credit risk. But if investors are behaving in a way that suggests credit risk is not the binding constraint, the instrument is aimed at the wrong hurdle. We read this in Nigerian capital allocation in The Hurdle Is Not the Risk.
When a development bank lends through a commodity trader to reach smallholders, the structure is itself a diagnosis of where the working relationships lie, and revives an old question about bundling credit with an offtake relationship. Examined in Resilience Is Downstream of the Buyer.
Some things are hard to sell because what the buyer is really being asked to purchase is a counterfactual, a saving relative to a future that will not happen, which is inherently hard to prove. This is why efficiency projects with fast paybacks still do not get done, as we argue in Nobody Buys a Chiller.
Infrastructure returns arrive over fifteen or twenty years; most funds must return capital in ten. When the tenor of the money is shorter than the life of the asset, good projects go unfunded not because they are bad but because they are the wrong shape for the available capital. The mismatch, not the project, is the problem.
For a capital-intensive transition, the interest rate matters more than the technology cost. A solar plant that is cheap to build can still be unaffordable where capital is expensive, so the same project pencils out in one country and fails in another purely on the cost of finance. This is why the price of money, not the price of panels, often decides where a transition happens.
Inventing something valuable and capturing the value from it are two different problems, and the second is often the harder one. The value in a commodity or technology chain is captured by whoever holds the scarce complementary asset, which is rarely the inventor. The full technology-strategy treatment, following Teece (1986), is on the technology & innovation dynamics page; the two entries that follow read the same fact as a value-capture and market-barrier question.
The value in a commodity chain usually sits not in extraction but in the next step: refining, processing, manufacturing. Capturing it is the aim of most resource-based industrial policy, and the constraints are rarely the ones announced. Set out in The Value Is in the Second Step.
To move up a value chain, output must be qualified, certified as good enough to enter a buyer's supply chain, and qualification, not technology or capital, is often the true gate. The same mechanism that shuts the door is why the door is worth opening. Drawn out for battery precursor in the DRC in The Lock-In Runs Both Ways, and for export bans in The Ban Is Not the Policy.
Networks of firms exchanging each other's by-products deliver real efficiency, but the successful examples accreted over decades and were never designed. Taking the ecosystem as a unit to be built from a site plan misreads how it formed. Set out in Symbiosis Does Not Arrive on a Site Plan.
When demand for a product exists only because a regulation requires it, the regulation is the asset, and its risk profile is political, not commercial. A plant built against an EU quota is a legitimate asset of a particular and unusual kind. Examined in The Mandate Is the Mine.
Small buyers of transition hardware, solar plants, data centres, factories, often order from the same constrained suppliers, and when lead times double it is the small and un-tracked buyers who wait longest. Set out in Standing in the Same Queue.
A headline price often hides that most of the cost is contractual rather than physical. Kenya can be 93 per cent renewable and still expensive because three of four cost bands are set in contracts, not at any power station. Broken down in Paying for Power You Curtail.
A very large buyer of power, a smelter historically, a data centre today, negotiates a bargain of a specific shape, and there is sixty years of evidence about how those bargains work out for the host system. The named instrument is the power purchase agreement, or PPA. Drawn out in The Smelter Contract.
Whether new self-generation strengthens or weakens a shared grid, whether a transition helps or harms the people left on the old system, is frequently decided in the fine detail of the tariff. A recurring theme, central to The Customers Who Can Leave and Paying for Power You Curtail.
Intermittent generation and firm, dispatchable power are different products with different values. Much of the difficulty in mineral processing and industrialisation comes down to the cost and availability of firm power, addressed in The Value Is in the Second Step.
Across a large sample of circularity projects, none of the binding bottlenecks was a recycling technology; they were all institutions for verifying a claim. A recycled material is a promise, and promises need machinery to be trusted. Set out in A Recycled Material Is a Promise.
As remote verification, satellite radar, sensors, gets cheaper, the last budgeted reason for anyone to physically go and look disappears, and something is lost with it. Argued for rice-paddy monitoring in The Cheaper It Gets to Verify, the Less Anyone Visits.
The size of a monitoring, reporting, and verification budget tells you something about the transaction it sits on: elaborate verification usually means the benefit and the burden have been separated, and separated benefits need continuous machinery to stay connected. Drawn out in The Distance Between the Work and the Reward.
An accuracy figure is an average; a clinic, a market, a grid experiences a distribution. When the errors fall along the same axis as the exclusion a system was meant to fix, the average is actively misleading. Set out for AI in Who Does It Fail For?, and as a general principle in the data analysis resource.
A results framework can only measure the benefits it was designed to measure, so real and valuable outcomes go uncounted simply because nobody specified them in advance. Shown for settlement upgrading in The Benefits Nobody Was Looking For.
When the party being evaluated is also the party supplying the evaluation, to a funder deciding on renewal, the evidence base tilts optimistic by construction, and more dashboards do not fix it. This is the reasoning behind the Lab's funder service, set out in The Reporting Loop. The behavioural foundation, courtesy bias in the interview itself, is on the human side of technology page.
A percentage without its base is not a finding, and one month is not a trend. Two of the commonest reading errors, drawn from a single month of venture data, in One Month Is Not a Trend.
The party a policy names is often not the party that ends up carrying the cost or receiving the gain. Asking where the burden and the benefit actually fall, rather than where they were assigned, is the core of distributional analysis, and it runs through most of the pieces in this section.
A subsidy aimed at vulnerable households can deliver comfort rather than savings, which may be worth having but is not what the fund promised. The distinction matters because it changes who is actually helped. Examined for Greece's Social Climate Fund in A Warm House Is Not a Cheaper One.
Some public goods were produced for free as a by-product of an economic activity, fire management as a by-product of farming, for instance. When the activity stops, buying the good directly costs far more and lasts only as long as the budget. Set out in The Fire Was Put Out by People Making a Living.
The binding constraint on a rural livelihood is often not who owns the land but who holds the specific right that matters, the permission to cut the tree, harvest the crop, use the resource. Drawn out for a restoration programme in The Right That Matters Is to the Tree.
A new trade corridor redistributes market access before it creates any, and the places that lose out often experience no single measurable event, just a slow relative decline. Examined in The Village Twenty Kilometres Off the Road.
Administrative and clearance burdens do not scale down with the size of the body carrying them, so a small municipality faces almost the same fixed hurdle as a large one with a fraction of the capacity. Set out in The Municipality Is the Instrument.
A transition that is good in aggregate still imposes concentrated losses on the people tied to the old system: the petrol mechanic, the coal town, the fuel vendor. Ignoring those losses is not only unjust, it is destabilising, because the people who lose are usually organised enough to slow or reverse the change. Whose livelihood is stranded, and whether they are compensated or abandoned, is an economic question before it is a moral one.
Many transition problems exist because a real cost or benefit is not priced: pollution that no one pays for, resilience that no one is rewarded for. Where the price is missing, the market sends the wrong signal, and correcting it, through a tax, a subsidy, a mandate, is often the whole substance of transition policy. Pigou (1920) proposed the corrective tax; Coase (1960) showed that where transaction costs are low the parties may bargain to the efficient outcome without one, which is why the transaction-cost structure, not just the externality, decides the right instrument. The absence of a price, not the presence of a bad one, is frequently the root problem.
Some benefits, once produced, are available to everyone and cannot be withheld from those who do not pay, clean air, fire management, a stable climate, so no private party has the incentive to supply them, and they are systematically under-produced. Samuelson (1954) gave the formal definition. Much transition infrastructure has a public-good character, which is precisely why it does not appear without public provision or a deliberately engineered price.
Some economic patterns are large enough to shape whole regions and whole technologies.
Invention versus commercialisation. Europe produces world-class innovation and struggles to scale it; the United States does the reverse. The asymmetry defines where value is captured on each side of the Atlantic, and where an independent evidence layer fits, as we argue in Europe Invents, America Scales.
Absorptive capacity. The climate question about AI is stuck on how much electricity it uses. The more decisive variable is which sectors are ready to convert AI into productivity, and the fossil economy has a forty-year head start, set out in The Asymmetry Nobody Is Metering.
The classical terms above draw on the following canonical sources. Where a concept is the Lab's own, the applied case is linked inline rather than cited here.
Akerlof, G. A. (1970). The market for 'lemons': Quality uncertainty and the market mechanism. Quarterly Journal of Economics, 84(3), 488 to 500.
Arthur, W. B. (1989). Competing technologies, increasing returns, and lock-in by historical events. The Economic Journal, 99(394), 116 to 131.
Coase, R. H. (1937). The nature of the firm. Economica, 4(16), 386 to 405.
Coase, R. H. (1960). The problem of social cost. Journal of Law and Economics, 3, 1 to 44.
David, P. A. (1985). Clio and the economics of QWERTY. American Economic Review, 75(2), 332 to 337.
Katz, M. L., and Shapiro, C. (1985). Network externalities, competition, and compatibility. American Economic Review, 75(3), 424 to 440.
Marshall, A. (1890). Principles of Economics. London: Macmillan.
Pigou, A. C. (1920). The Economics of Welfare. London: Macmillan.
Samuelson, P. A. (1954). The pure theory of public expenditure. Review of Economics and Statistics, 36(4), 387 to 389.
Spence, M. (1973). Job market signalling. Quarterly Journal of Economics, 87(3), 355 to 374.
Teece, D. J. (1986). Profiting from technological innovation: Implications for integration, collaboration, licensing and public policy. Research Policy, 15(6), 285 to 305.
Williamson, O. E. (1985). The Economic Institutions of Capitalism. New York: Free Press.
For the transitions-studies frameworks that sit alongside these (the multi-level perspective, strategic niche management, and technological transition pathways), see the transitions primer and the readiness levels explainer, which carry their own references.
These concepts are not academic ornaments. Each one is a lens the Lab actually uses when reading a real deployment: before a company enters a market, before a funder renews a grant, before a consortium claims an impact. The economics is where the decisive questions usually hide, and naming the concept is the first step to asking whether it applies to the case in front of you.
For the full arguments, follow the links to the individual pieces. For the frameworks behind them, see the BRW framework, the readiness levels, and the resources library. For how the Lab applies all of this to a specific case, see What We Do, and to discuss a study, Contact.
This page is a living map and grows as the Lab publishes. The economics of a transition is rarely the first thing discussed and is usually the thing that decides it.