The Cost of a Job: Nigeria's 8,538 Dollar Ledger vs Canada's Battery Bet
Nigeria ran a national cash-grant competition, published a peer-reviewed evaluation, and produced a verified job for $8,538. Canada committed at least $43.6 billion to three battery megaprojects meant to do the same thing, one of which delivered zero jobs before it collapsed. This is a line-by-line accounting of what each dollar actually bought, and a working model built from the mechanism that Nigeria proved.
In 2011, Nigeria stopped trying to lend its way out of unemployment
Nigeria had roughly 8 million young people who needed jobs, and a banking system that could not tell a good one-person business from a bad one, because almost none of them kept records a bank could underwrite. The finance ministry's answer, led by then finance minister Ngozi Okonjo-Iweala, was to stop trying to lend, and start giving.
The government opened a national competition called YouWiN!, the Youth Enterprise With Innovation in Nigeria. Anyone could apply with a short business plan. Nearly 24,000 people did, everything from a paint retailer to a catfish farmer to a dry-cleaning startup in a town that had none.
What made the design defensible, in a country where officials expected the public to assume the money would go to whoever had the right connections, was four specific rules, not the size of the cheques.
01 · Open competition
Public, national, low-friction entry. No connections required, no pre-selected shortlist. 23,916 applicants competed on paper alone.
02 · Blind judging
Applicant names were stripped before scoring. Assessors were brought in from outside government specifically to remove the appearance of patronage.
03 · Partial randomization
After weak plans were screened out, winners among the remaining qualified pool were chosen partly by score, partly by lottery, on purpose.
04 · Cash, not credit
Winners received an outright grant, averaging about $50,000 and reported as high as roughly $65,000, no collateral, nothing to repay.
Six years later, the World Bank could show its work
Because winners were partly chosen by lottery, economist David McKenzie could compare identical applicants who did and did not receive funding, then track both groups for six years, through a national recession. The result was published in the American Economic Review in 2017, one of the most cited studies in development economics.
McKenzie's own conclusion, later summarized by the American Economic Association, was that the effect came from the grant itself, not from added mentoring, confidence, or new business connections. Relieving the capital constraint was enough on its own. Economist Chris Blattman, reading the results cold, asked the question that became the title of his own write-up: is this the most effective development program in history.
Canada ran the opposite experiment, at a thousand times the price
Since 2020, Canada's federal and provincial governments have negotiated a handful of closed-door deals with individual multinationals to build electric vehicle battery plants, on the theory that a few very large employers would out-perform many small ones. Three case files below, all independently costed by Canada's Parliamentary Budget Officer.
Permanent operating headcount remains commercially confidential. The PBO could not verify the government's own job promise and said so directly in its report.
Commercially confidential · figures undisclosedThe temporary-worker detail drew public criticism precisely because the subsidy had been justified as a jobs program for Canadians. Stellantis has since sold its 49 percent stake to LG Energy Solution.
Partially disclosed · final headcount pendingNorthvolt's Swedish parent filed for bankruptcy in March 2025 before the Quebec plant reached production. Quebec's $270 million equity stake was declared a total loss. The site's roughly 50 remaining staff were laid off entirely.
Zero jobs, project terminatedWhat a billion dollars actually buys, side by side
Canada's own Parliamentary Budget Officer put the total government commitment across all three battery deals at $43.6 billion through 2032–33, 62 percent federal, 38 percent provincial, and warned the true figure could run several billion higher once tax treatment is fully accounted for. Converting every program to the same unit, jobs produced per billion dollars committed, removes the spin.
Canada figures use the PBO's $43.6 billion aggregate against (a) the three projects' own combined promise of 8,500 jobs, and (b) the roughly 2,700 jobs independently confirmed to date, a number still rising as Volkswagen and NextStar ramp toward full production. Northvolt is held out separately because it delivered zero and cannot rise further.
Per-project totals reconstruct each project's full pledged commitment, federal plus provincial, using the same up-to-ceiling methodology the governments' own announcements use: Volkswagen $16.3B, NextStar $15.5B (midpoint of a $15–16B range), Northvolt $4.9B ($510M disbursed by Quebec plus $4.4B pledged federally, never paid out). These are not an official PBO per-project split of its $43.6B aggregate, which was not published at that level of detail.
The broader pattern holds beyond these three. Independent tallies by labour researchers, corroborated in Globe and Mail reporting, put Canada's total EV and battery-adjacent subsidy envelope at close to $50 billion across roughly 16 projects since 2020. Umicore, in Loyalist Township, Ontario, took a subsidy covering 35 percent of its plant's cost and promised 700 permanent jobs; construction halted nine months after the groundbreaking, and the site has sat largely idle since.
The megaprojects are not the exception. Canada's own grant funds tell the same story
If the problem were only that foreign megaprojects are expensive, Canada could simply negotiate harder. But the pattern repeats inside the programs structurally closest to Nigeria's mechanism, the ones actually meant to fund small operators directly.
Auditor General, October 2025: the minister's office overrode non-partisan staff rankings on 54 percent of selections, sending $742M to medium, low, or poor-ranked applicants. $126M went to applicants who had hired lobbyists tied to the governing party.
Not fair, transparent, or accountableAuditor General, June 2024: 90 confirmed cases where conflict-of-interest rules were violated, tied to $76M in funding, plus $59M awarded to ineligible projects. The entire board resigned.
Agency abolishedPromised 50,000 jobs and $50B in added GDP. The PBO found spending running at 29 percent of pace by 2020 and modelled a realistic ceiling of about 27,000 jobs. By 2025 the GDP claim had been quietly rescoped down to $13–16B.
46% below its own targetCanada's largest youth entrepreneurship program is a loan, not a grant, capped at $75,000 and charged around prime plus 3 percent, repayable over 5 years. Its total scale is smaller than the capital-expense line of a single EV subsidy cheque.
A loan, sized like a rounding errorFour design choices explain almost the entire gap
None of this is a wealth or competence problem. Canada has audit institutions and financial sophistication Nigeria did not have in 2011. Line the two approaches up on the same four axes and the gap stops looking like a mystery.
Nigeria
Dispersed across 1,200 small, labour-intensive operators. A dollar given to a paint shop becomes wages roughly in proportion to headcount.
Canada
Concentrated in a handful of capital-intensive, highly automated plants, where a dollar buys machinery first and jobs second, if at all.
Nigeria
Names stripped, scored blind by assessors from outside government, specifically engineered to defeat patronage fears.
Canada
Ontario's Skills Development Fund let a minister's office override non-partisan rankings on the majority of its selections.
Nigeria
Randomization among the qualified pool removes discretion at the exact margin where favouritism usually happens.
Canada
Discretion is the mechanism. $126M went to applicants who had hired lobbyists connected to the governing party.
Nigeria
The randomization created a control group, which let a rigorous evaluation be published in a top economics journal.
Canada
The PBO has repeatedly flagged an absence of quantifiable objectives; many megaproject job figures are commercially confidential and may never be confirmed.
Concentrated subsidies to a handful of famous employers make for a better ribbon-cutting than a lottery that hands modest sums to thousands of small business owners nobody has heard of. That is a political-economy explanation, not a technical one, and it is worth naming honestly.
The mechanism has already travelled, twice, with mixed results
A fair investigation tests its own thesis. Nigeria's model has been copied elsewhere, and a closely related mechanism has been tested at even larger scale. The record is genuinely mixed, which is itself useful evidence.
Senegal also adopted a version of Nigeria's competition, per the same reporting that covered YouWiN!'s original rollout, though independently verified outcome data for that program was not available for this investigation and is not claimed here. The honest reading of the evidence is that the mechanism, public, blind, randomized, dispersed, evaluated, is a well-tested starting point, not a guarantee. Ghana's null result is the reason a Canadian version would need its own independent evaluation from day one, not an imported conclusion.
Run the reallocation yourself
This is not a forecast. It applies Nigeria's own revealed ratio, about 5.9 jobs for every funded business, to a slice of Canada's existing $43.6 billion EV subsidy commitment, so you can see the order of magnitude for yourself. Move the sliders.
Illustrative extrapolation from a single verified study, not a guarantee of Canadian outcomes. Exhibit I's Ghana result is the reason any real version of this needs its own randomized evaluation before anyone claims success.
A model, built only from what is already proven to work
This does not argue Canada should abandon large industrial bets. It argues Canada is missing the specific mechanism that Nigeria proved works for broad, verifiable, low-cost job creation, and that the fix is reallocation and design discipline, not new legislation or new money.
-
01
A single, arm's-length national competition
Explicitly insulated from any minister's office, unlike the Skills Development Fund model. Open to existing small businesses and prospective entrepreneurs, not pre-favoured sectors.
-
02
Blind scoring against published criteria
Run by assessors outside government, mirroring the exact fix Nigeria used to defeat patronage concerns before they started.
-
03
Partial randomization among the qualified pool
Removes discretion at the margin where most favouritism disputes happen, and builds in the control group that makes honest evaluation possible.
-
04
Grants, not loans, sized for the missing middle
Canada's own missing middle sits in the $10M–$100M revenue band, too big for venture capital, too small or early for private equity. A meaningful non-dilutive grant needs to sit well above Futurpreneur's $75,000 loan ceiling.
-
05
Independent, published evaluation from day one
Through Statistics Canada or a university partnership, so the program's own results become the argument for renewing or scaling it, the way McKenzie's published evaluation became the argument for Kenya and Senegal to copy Nigeria's model.
What does not transfer cleanly
Nigeria's result is not a universal law, and the Ghana null result in Exhibit I is the proof. Canada's wage levels, existing SME financing infrastructure, and cost base are different enough from Nigeria's in 2011 that grant size, sector focus, and eligibility rules would need real recalibration, not a copy-paste of a $50,000 figure.
What transfers is not the dollar amount. It is the mechanism: public, blind, randomized, dispersed across many small operators, and evaluated from the start. Canada already has every institutional capability required to run that evaluation well. It has simply not tried the mechanism.
Canada does not need to choose between large industrial policy and this kind of program. It needs both, at a moment when youth unemployment has been running near 13 to 14 percent, well above the pre-pandemic average, and Canadian manufacturing has shed more than 60,000 jobs to tariffs since early 2025.
How this was built, and where every number comes from
Methodology and honesty notes
This investigation deliberately compares Canada against Nigeria, and against Kenya and Senegal where Nigeria's specific mechanism was replicated, rather than against a broad basket of OECD peers such as Singapore, Estonia, or Ireland. That choice is intentional: the available research base is strongest and most directly comparable where the same mechanism, a randomized, blind business-plan competition, was actually tried. Extending the comparison to countries with structurally different entrepreneurship policy, without equivalent published, peer-reviewed evaluation, would trade rigor for false breadth, and this piece chose not to make that trade.
Where a figure is independently audited or peer reviewed, it is marked verified. Where a figure comes from a company or government announcement without independent confirmation, it is marked commercially confidential or flagged accordingly. Where sources disagreed on a number, for example press reports of the Sustainable Development Technology Canada conflict-of-interest count, the Auditor General's own clarified figures were used as primary, with the discrepancy noted.
The reallocation simulator in Exhibit J is an explicit extrapolation from one verified study, not a projection or a policy recommendation with guaranteed outcomes. It is presented as a magnitude check, not a forecast.
Full source list
- David McKenzie, "Identifying and Spurring High-Growth Entrepreneurship: Experimental Evidence from a Business Plan Competition," American Economic Review 107(8), 2017.
- American Economic Association, "The business competition that worked," research highlight summarizing McKenzie 2017.
- World Bank Microdata Catalog, "Nigeria: Youth Enterprise With Innovation in Nigeria (YouWiN!) Program Impact Evaluation, 2011–2016."
- NPR Planet Money, "Piles of cash and a town of solutions in Kenya, Nigeria," Summer School World Tour, July 2026.
- Office of the Parliamentary Budget Officer of Canada, "Costing Support for EV Battery Manufacturing," November 2023, and related PBO releases on the Volkswagen, Stellantis–LG, and Northvolt deals, and on the Innovation Superclusters Initiative, 2020–2023.
- CBC News reporting on the Volkswagen St. Thomas, NextStar Windsor, and Northvolt Quebec deals, 2023 through 2026, including PBO break-even and job-count coverage.
- Office of the Auditor General of Canada, "Sustainable Development Technology Canada," Report 6, 2024.
- Office of the Auditor General of Ontario, report on the Skills Development Fund, October 2025, as reported by CTV News and CP24.
- Innovation, Science and Economic Development Canada, evaluations of the Innovation Superclusters Initiative and of Futurpreneur Canada.
- CBGF, "The Missing Middle: Bridging Canada's Mid-Market Growth Capital Gap," 2025.
- Statistics Canada, Labour Force Survey releases, 2025–2026.
- Rank and File, "Canada's failed EV strategy is corporate welfare run amok," October 2025, cross-referenced against Globe and Mail reporting.
- Kenya Youth Employment and Opportunities Project (KYEOP) / MbeleNaBiz program reporting, Daily Nation, 2021.
- GiveDirectly, published randomized cash-transfer research in Kenya, including the 2022 general-equilibrium multiplier study and infant-mortality findings.
- World Bank working paper reference to Fafchamps and Woodruff (2016), business plan competition in Ghana, cited in "Using Experimental Evidence to Inform Firm Support Programs in Developing Countries," World Bank Policy Research Working Paper 9461.