Haute Lumière
Commerce · V.06 · MMXXVI · daylight
TEN CONCEPT BRIEFS · Chapter V.06 — The Cooperative Firm, Audited
One page each. A reader who reads only these ten pages has the chapter.
BRIEF 1 — The Routing of Volatility
The idea. A shock to a business has to go somewhere. The conventional firm sends it to employment; the cooperative sends it to pay. Neither abolishes it.
This is the single most robust empirical finding about worker cooperatives, and it has been found three times on three datasets that share no assumptions: Craig and Pencavel on the Pacific Northwest plywood cooperatives, Pencavel, Pistaferri and Schivardi on Italian firms, Burdín and Dean on Uruguay. Against the same demand shock, cooperative pay moves a great deal and cooperative employment moves very little. The conventional firm does the reverse.
Worked example. The John Lewis Partnership publishes its Partnership Bonus as a percentage of pay. Fourteen years read 18, 14, 17, 15, 11, 10, 6, 5, 3, 2, 0, 3, 0, 0. Mean 7.43 percent, standard deviation 6.595 points, coefficient of variation 0.888. Three years in fourteen paid nothing. No listed retailer's dividend behaves like that — and no listed retailer's pay behaves like that either, which is the point.
The honest qualification. John Lewis also reduced headcount over the same period. Partnership structure does not guarantee employment stability. It changes which variable moves first, and the trade has to be read on both axes or not at all.
You already know this because you have watched a business hold its people through a bad year by freezing everything else, and you understood without being told that somebody was paying for it.
BRIEF 2 — Insurance Has a Break-Even
The idea. The cooperative's income deal is an insurance contract. Insurance is never a better expected value; it is a better shape. So the question is never is it better but for whom.
Take a bad year. Conventional firm: 10.0 percent of people laid off, the rest held flat, a laid-off worker retaining 60.0 percent of income through the spell. Cooperative: 1.0 percent laid off, everyone takes an 8.0 percent cut.
expected income, conventional 96.00 % of normal pay
expected income, cooperative 91.68 % of normal pay
the cooperative member expects 4.32 points LESSWorked example. Value both under constant relative risk aversion and find where they cross.
| Risk aversion | Conventional | Cooperative | Preferred |
|---|---|---|---|
| 1.0 | 0.9502 | 0.9161 | conventional |
| 2.0 | 0.9375 | 0.9151 | conventional |
| 3.0 | 0.9214 | 0.9138 | conventional |
| 4.0 | 0.9019 | 0.9121 | cooperative |
| 5.0 | 0.8795 | 0.9099 | cooperative |
Break-even: 3.446.
Why it matters. Chetty's estimate of relative risk aversion, built from labour-supply behaviour rather than surveys, sits near one and below two. On that estimate the cooperative's income smoothing alone does not pay for itself — which means the cooperative's real advantage has to be firm survival, voice, and the job continuing to exist. That is a narrower claim and the data will carry it.
You already know this because you have bought insurance knowing perfectly well the expected value was against you, and bought it anyway, and been right to.
BRIEF 3 — The Unsellable Claim
The idea. A cooperative member holds a claim she cannot sell. There is no market, no buyer, and usually no capital gain — the account is withdrawn at book value on leaving. An unsellable claim is worth less to whoever must hold it, and that discount is a cost of capital.
Worked example. Price it the way an analyst prices any unmarketable minority holding.
risk-free rate 4.0 %
equity risk premium 5.0 %
equity beta 0.90
conventional cost of equity 8.50 %
discount for lack of marketability 25.0 % over 8 years
annualised illiquidity premium 3.66 points
cooperative cost of equity 12.16 %Check it by a second route. The member's holding is undiversified — her labour income and her savings are in the same firm. Scaling beta by the inverse of its correlation with the market gives a total beta of 1.80 and a cost of equity of 13.00 percent, a premium of 4.50 points. Two routes, two answers, and the honest thing is to quote the lower one and say the other exists.
Why it matters. This is the cooperative's structural cost and it does not go away with better management, a better year, or a better argument. Every serious proposal in this chapter is an attempt to pay it down.
You already know this because you have priced a private company against a listed one and knocked something off for the fact that you could not get out.
BRIEF 4 — The Refused Band
The idea. A higher cost of equity is not an abstraction. It is a specific range of real projects that one firm accepts and the other must decline.
Worked example. Carry the two costs of equity into a weighted average cost of capital at 40.0 percent equity, 60.0 percent debt, debt at 5.0 percent pre-tax and tax at 25.0 percent.
after-tax cost of debt 3.75 %
conventional WACC 5.65 %
cooperative WACC 7.11 %
the hurdle gap 1.46 pointsThe band. A project returning between 5.65 percent and 7.11 percent creates value inside the listed firm and destroys it inside the cooperative. Same machine, same customers, same country, two different answers — and the cooperative's answer is correct given its capital.
What to do with it. Go and look. Pull everything declined in the last three years and mark what fell inside your own band. That schedule is the business case for every instrument in this chapter, and it is the one document that makes the cost of capital feel like money rather than theory.
Why it matters. Most arguments about cooperative underinvestment are arguments about culture. This one is arithmetic, and arithmetic can be financed.
You already know this because you have watched a well-run division turn down a sensible project purely because of the rate it was being charged internally.
BRIEF 5 — Growth Without External Equity
The idea. Two firms can earn identically and grow differently, because one of them has a third source of capital and the other does not.
A conventional firm funds growth from retained profit, debt, and new equity. A cooperative funds it from retained surplus, debt, and new members' stakes — and new members' stakes are small.
Worked example. Both firms earn 12.0 percent on equity, pay out 40.0 percent, retain 60.0 percent. The listed firm issues net new equity worth 3.0 percent of opening equity a year. The cooperative's members contribute €15,000 on joining against €150,000 of invested capital per worker, with headcount growing 3.0 percent a year.
member-stake inflow = 15,000 x 0.030 / 150,000 = 0.30 % of capital
conventional sustainable growth 12.0 x 0.60 + 3.0 = 10.20 % a year
cooperative sustainable growth 12.0 x 0.60 + 0.30 = 7.50 % a year
the gap 2.70 pointsThen compound it. After ten years the conventional capital base is 28.2 percent larger; after twenty, 64.2 percent; after thirty, 110.5 percent.
Why it matters. The cooperative is not less profitable. It is less able to convert profit into scale, and scale is what decides who is still in the market in year twenty. This is the chapter's honest negative and every other part of the design exists to answer it.
You already know this because you have seen two equally good businesses diverge over a decade for no reason anybody could name except that one of them could raise money.
BRIEF 6 — The Horizon Problem, Priced
The idea. A member evaluates an investment over the years she has left in the firm, not over the asset's life, because she cannot sell the claim on the way out. Furubotn and Pejovich named this in 1970. It is usually asserted. It can be computed.
Worked example. An investment costs €100 and returns €12 a year in perpetuity. The discount rate is 10.0 percent. To a holder of a tradable claim it is worth €20.00.
| Years remaining | Annuity factor | NPV to the member | Vote |
|---|---|---|---|
| 8 | 5.3349 | −35.98 | no |
| 15 | 7.6061 | −8.73 | no |
| 20 | 8.5136 | 2.16 | yes |
The break-even is 18.80 years. A member needs that much remaining service before she will vote for a project the market values at plus twenty.
And the fix falls straight out of it. Credit half the retained value to her own capital account, payable on exit, and the break-even falls to 14.55 years — 4.25 years of horizon bought by an accounting policy rather than by asking anyone to be more patient.
Why it matters. A great deal of behaviour attributed to cooperative conservatism is a rational response to an instrument, and instruments can be redrafted.
You already know this because you have watched somebody two years from retirement decline to start something they would obviously have started at forty.
BRIEF 7 — Degeneration Is an Admission Rate
The idea. The degeneration thesis says cooperatives drift toward ordinary employment. Tested as arithmetic rather than asserted as a tendency, it turns out to be a policy variable.
The test is one ratio: worker-members divided by people employed.
Worked example. Start at 80.0 percent members. Hire at 5.0 percent a year. Admit new members at 2.0 percent a year.
after 5 years 69.2 %
after 10 years 59.9 %
after 20 years 44.8 %Nothing in that required a change of heart. No founder sold out, no principle was abandoned. The firm hired faster than it admitted, for a decade, for perfectly good operational reasons each time.
The recovery arithmetic. To climb from 59.9 percent back to 80.0 percent over ten years, admission must exceed hiring by 2.94 points a year — admitting at 7.94 percent while hiring at 5.0 percent. Admitting at the hiring rate holds the ratio exactly.
Why it matters. Cornforth's review, and Storey, Basterretxea and Salaman's comparison of Eroski and John Lewis, find degeneration real in some firms and reversed in others — exactly what you would expect of something governed by an admission policy. Publish the ratio quarterly and the thesis stops being a debate.
You already know this because you have seen a company's culture change without a single meeting about culture, purely because of who it hired and how fast.
BRIEF 8 — Reading a Survival Series Honestly
The idea. Survival percentages are the most quoted and least useful cooperative statistic, because they compare populations rather than firms. Convert them into an annual hazard and they become usable — and then say what they did not look at.
Worked example. The French SCOP confederation publishes survival against the national figures for firm creations: 82 percent at three years against 66 percent, and 66 percent at five years against 50 percent.
five-year annual exit hazard, SCOP 7.97 % a year
five-year annual exit hazard, all firms 12.94 % a year
hazard ratio 0.616
three-year hazard ratio 0.495What it did not look at. Sector mix, firm age at entry, and the fact that many SCOPs are conversions of going concerns rather than start-ups. This is a comparison of two published populations and must not be quoted as causal.
The second route. Burdín matched worker-managed firms against conventional firms on size, sector and age using Uruguay's social-security register, and still found the worker-managed firms less likely to dissolve. Two methods, two countries, same sign — which is how a claim earns the right to be repeated.
Why it matters. A lender can price off a hazard rate. A lender cannot price off a percentage in a brochure.
You already know this because you have already distrusted a statistic that was true and useless, and wanted to know what it had been compared against.
BRIEF 9 — The Marcora Conversion
The idea. A worker facing redundancy is owed unemployment benefit. Italy's Marcora law, passed in 1985, lets her capitalise it — take it as a lump sum and use it as founding equity in a cooperative buyout of the firm that is closing, matched by institutional money from a dedicated fund.
Worked example. CFI, the fund that writes the cheques, publishes its cumulative record: on the order of 560 enterprises and 25,000 jobs, an average of 44.6 jobs per enterprise, against reported cumulative investment near €350,000,000.
cost per job held EUR 14,000
two years of benefit at 1,000/month EUR 24,000
ratio 0.583Sensitivity, because the cumulative figure is reported rather than audited: at €250,000,000 the cost is €10,000 a job; at €400,000,000 it is €16,000. Every case in the band is below the benefit.
Why it matters. This is not a subsidy in a new envelope. The state is spending money it had already committed, in a form that produces a going concern rather than a claim. Vieta, Depedri and Carrano's Euricse study finds the resulting firms durable rather than temporary.
You already know this because you have known intuitively that it is cheaper to keep a working thing working than to pay for the consequences of it stopping.
BRIEF 10 — The Member Capital Bridge, and IFRIC 2
The idea. A cooperative cannot issue a residual claim without ceasing to be a cooperative. It can issue a non-residual one: non-voting, dividend-capped, redeemable, subordinated to debt and senior to member capital accounts.
The accounting crux, and it is the whole thing. IFRIC 2, Members' Shares in Co-operative Entities and Similar Instruments, decides whether members' shares sit in equity or in liabilities, and it turns on whether the entity holds an unconditional right to refuse redemption. Write that right into the rules and the instrument is equity. Leave it out and it is debt, and your gearing covenant moves against you on the day you issue.
Worked example. A cooperative of 1,000 people at €150,000 of invested capital per worker has a base of €150,000,000. A growth gap of 2.70 points implies a year-one tranche of €4,050,000, and an unfinanced decade leaves it at €309,154,734 against a competitor's €396,193,370 — a shortfall of €87,038,636.
The number that decides it.
return on invested capital > coupon + amortised issuance
7.05 % > 6.00 % + 0.50 %
margin 0.55 pointsThin, and saying so is the point: the instrument works above a 6.50 percent return on invested capital and not below it.
Three legs, not one. Italy's Law 59/1992 levy of 3 percent of profit gives this firm €216,000 a year against a surplus of €7,200,000 — 5.33 percent of the tranche, about one twentieth of the need. The rest comes from an institutional cooperative investor and from admitting members at the hiring rate, which lifts the member-stake inflow off 0.30 percent and fixes degeneration in the same motion.
You already know this because you have seen a perfectly good financing collapse over a definitional question that somebody should have settled with the auditors first.
All figures in these briefs are computed in lib/verify/V_06.py, printed by python3 lib/verify.py V.06, with every input labelled published, reported or model, and sourced in the chapter's Works Cited.