The most expensive sentence in family finance is "my child must get a degree." Not because degrees are worthless. Because the sentence quietly converts the single most powerful wealth-transmission channel a family owns...
The most expensive sentence in family finance is "my child must get a degree." Not because degrees are worthless. Because the sentence quietly converts the single most powerful wealth-transmission channel a family owns into a single product, bought at the highest price on the shelf, on a payment plan that can bankrupt the buyer before the product ships.
Start with the data that makes education worth arguing about at all. In 2017, Fabian Pfeffer and Alexandra Killewald published a three-generation study of American family wealth in Social Forces. When they decomposed how parental wealth becomes child wealth, education carried roughly 25.5 percent of the transmission, the second-largest measured channel after homeownership at about 28.4 percent, and more than double direct bequests and gifts at about 12.3 percent (Pfeffer and Killewald, Social Forces 2017). Wealth mostly moves between the living, and a quarter of the movement travels through what parents pay to have their children taught.
Here is what the study does not say. It does not say the channel is called "university." The mechanism being measured is educational attainment converting into earning power. A degree is one vehicle on that road. It is not the road.
Walk through any trading town and count the assets that were created by training that never touched a lecture hall.
The welder with a recognized certificate charges double the welder without one and gets the contract for the school's window grilles because the contractor needs paper for the file. The certificate cost a fraction of one university semester. The salon chair rents out at a premium when the woman behind it holds a cosmetology certificate and can prove she trained, because brides pay for proof. The boda rider with a valid driving permit and rider training is insurable, hireable by the delivery apps, and legally defensible at a checkpoint; the one without is one accident away from losing the asset and the income together. A driving licence, in much of the world, is a income document. So is an electrician's wiring licence, a nursing certificate, a plumbing trade test, a food-handling permit, a Class B commercial permit.
Then the digital layer. A teenager who completes a serious bookkeeping or spreadsheet course can keep the books for three family shops. A cousin who learns to run advertising accounts or edit video sells that skill across borders from a phone. These credentials cost tens of dollars to low hundreds, take weeks to months, and start paying inside a year.
Set the two models side by side as an operator would:
| | Typical degree | Trade certificate or licence | |---|---|---| | Cost | High, often multi-year debt or land sold | Low, often under one semester's fees | | Time to first income | 3 to 5 years, plus job search | 3 to 18 months | | Payback period | Often 5 to 10 years, sometimes never | Often under 2 years | | Failure mode | Graduate unemployment with debt | Low sunk cost, retrain cheaply | | Stacks with family assets | Sometimes | Directly: the certificate runs the family shop, plot, or vehicle |
The last row is the one families miss. A welding certificate is not just an income for the holder. It is capacity the family can point at its own buildings, its own repairs, its own next venture. Credentials compound inside a family in a way a distant office job does not.
This is not an argument against university. It is an argument for buying university the way you would buy land: after checking the title.
A degree clears the bar in three situations. First, licensed professions where the degree is the legal gate: medicine, law, engineering sign-off, pharmacy, architecture. No certificate substitutes, because the state says so. Second, fields where the degree buys entry into a network and a labor market that genuinely pays the premium, and you have checked recent graduates from that specific course at that specific institution, not the national average. Third, the child who is authentically academic, the one who reads past the syllabus, for whom scholarship into research or specialized practice is the honest path.
Outside those cases, be suspicious, and here is the sharpest critique on the table. Economist Bryan Caplan, in The Case Against Education (2018), marshals evidence that a large share of the degree premium is signaling: the paper certifies pre-existing intelligence, conscientiousness, and conformity to employers rather than creating new skill. His own estimate attributes something like half or more of the premium to signaling rather than learning. Caplan's critics push back hard, and fairly: his estimates lean on contested assumptions, education has civic and personal returns his ledger ignores, and signaling itself pays the individual who buys it even if it wastes resources for society. But a family does not need to settle the academic fight to use its practical edge. If the premium is substantially signaling, then the degree's value collapses fastest exactly where degrees are most oversupplied, which is the situation in many countries where thousands of business and arts graduates chase a handful of formal jobs. A signal everyone sends stops signaling.
One more honesty note, since this piece leans on Pfeffer and Killewald: their decomposition is American data, built on American schools, mortgages, and labor markets. The 25.5 percent is not a law of nature and will not replicate exactly in Kampala or Lagos or Manila. What travels is the structure of the finding, that education is a first-rank channel and bequest a minor one, and if anything, in economies where formal jobs are scarce and self-employment is the norm, the case for income-producing credentials over general degrees gets stronger, not weaker.
Most families already run an education fund. It is called school fees, it is the largest line in the budget, and it has one approved product. The upgrade is to rename it and widen it.
Rule one: the pot funds credentials, not institutions. The question that unlocks money is not "which university" but "which credential, leading to which income, at what cost, with what payback period." A nursing diploma, a trade test, a driving permit plus defensive training, an accounting certification, an online qualification with real market recognition, and yes, a medical degree, all compete on the same worksheet. Write the worksheet: credential, total cost, months to completion, realistic first-year income, who in our network already earns from this.
Rule two: match the credential to the child, not to the family's ego. The girl who dismantles radios is telling you where her channel is. Paying for her to scrape through a business degree buys the family a framed photograph and four lost years. Paying for her electronics certification, then her electrical licence, buys an earner by 20. Graduation photos are decoration. Payback periods are inheritance.
Rule three: sequence cheap-and-fast before expensive-and-slow. A certificate that gets a 19-year-old earning does not close the door to a degree at 24, part-financed by the 19-year-old. The reverse sequence, degree first on family debt, closes doors every day it runs.
There is a second half of this channel that costs almost nothing and that wealthy families have never stopped using: placement. The credential says what you can do. The attachment under a real practitioner is where you learn to actually do it, and someone must open that door.
Treat placement as a standing family function. At the family meeting, keep a live list: which trades and professions do we have inside the extended family and close friendships, and which young person needs a bench. The uncle who runs a garage takes the nephew for a structured year, with a written expectation on both sides, not as cheap labor but as training with named skills to be signed off. The aunt in the hospital walks the niece's nursing application past the pile. The family friend with the hardware shop teaches stock-keeping to the cousin doing the bookkeeping course. Where the family has no practitioner in a target trade, the education pot pays a small honest fee to a master craftsman for a formal apprenticeship slot, and a family elder visits twice during the term so the master knows the child is watched over.
This is not nepotism in the corrosive sense. It is the family behaving like what the sociologists say it is, a transmission system, and doing deliberately what elite families do by reflex through internships and college networks. The difference between an old-money family and yours is not that they refuse to make calls for their children. It is that their calls are answered. Build the network that answers.
This week, sit down with your spouse or your council and write one line per child, including the grown ones: name, next credential, cost, time to completion, expected payback, placement contact. For some children the honest line will read "university degree in X, because the profession legally requires it and we have checked the graduate outcomes." For others it will read "welding trade test, four months, then attachment at Kasule's workshop." Both lines are education. Both lines are the 25.5 percent channel. Only one of them was visible before you widened the definition.
Fund the lines in payback order, cheapest and fastest first, and review the list every school-fees season like the investment portfolio it is.
This piece did its job if the next time someone in your family says "education," at least two people at the table ask "which credential," and nobody assumes the answer has a graduation gown in it.