Legislatures decide many things at once, which lets members trade support across issues: I'll vote for your project if you vote for mine. This vote-trading — logrolling — is the engine of distributive politics, and it interacts with a commons problem to produce one of public choice's sharpest predictions: that representative budgeting is biased toward over-spending, and the bias grows with the number of districts.
Logrolling: vote-trading across issues
When votes can be traded, intensity gets expressed. A member who cares intensely about a local irrigation scheme but little about a rural road can trade their road vote for others' irrigation votes. Sometimes this is efficient — it lets minorities with intense preferences prevail over apathetic majorities, revealing information that a simple yes/no vote suppresses. But it has a dark side: it can assemble a majority for a bundle of projects, each of which fails a cost-benefit test, because each member values their own project and ignores the cost imposed on everyone else's taxpayers.
The fiscal commons
Concentrated benefit, shared cost
A district project delivers a benefit concentrated in one district but is paid for from a common national budget — so each district bears only a fraction of the cost of its own project. Every representative therefore faces a distorted calculus: they capture the full local benefit but internalise only 1/n of the cost (where n is the number of districts). Projects worth pursuing privately are pursued even when their national cost exceeds their benefit. The national budget is a commons, and like any commons it is over-grazed.
The law of 1/n
Weingast, Shepsle, and Johnsen (1981) formalised this in the result known as the law of 1/n. Under universalism — the norm that every district gets its project so a winning coalition is assured — each district demands its project up to the point where local benefit equals its 1/n share of the cost. The larger n is, the smaller each district's cost share, the more over-sized each project, and the greater the aggregate over-spending.
The 1/n prediction
Inefficiency and total spending rise with the number of spending jurisdictions sharing a common budget. A district internalises only 1/n of the cost of its own project, so it demands a project scaled to where local marginal benefit equals (1/n) × marginal cost — over-sized by a factor related to n. More legislators dividing a common purse means more over-spending, not less. The result is the analytic backbone of the pork-barrel critique and a caution about how many hands you let into one till.
Universalism and omnibus bills
Why does the over-spending coalition reach (almost) everyone rather than a bare majority? Because membership in the coalition is itself uncertain and valuable: a norm of universalism — everyone gets something — insures each member against being left out and makes the logroll stable. Omnibus bills (one giant package no one dares vote down because their own project is inside it) are the institutional expression. The same logic appears in 'Christmas-tree' budgets festooned with everyone's ornament.
The African application: constituency development funds
Kenya's Constituency Development Fund (and similar funds across Africa and South Asia) is a near-pure instance of the 1/n logic, institutionalised. Each MP directs a slice of the national budget to projects in their own constituency, with the costs borne nationally. The benefits — visible local projects that aid re-election — are concentrated; the cost is diffused across all taxpayers. The predictable results follow from the model: strong political durability (no MP will vote to abolish their own fund), a bias toward many small visible projects over fewer high-return national ones, weak cost discipline (since each constituency bears 1/290 of the total), and recurring questions about whether the projects would survive a cost-benefit test. The fund is not an aberration; it is the law of 1/n written into statute.
Exercise
A national legislature of 47 members funds local development from a shared national budget. Each member can propose projects for their own region. (1) Explain, using the fiscal-commons idea, why each member rationally proposes larger projects than they would if their own region paid the full cost. (2) State the law of 1/n and predict what happens to total spending if the legislature is enlarged to 94 members with the same total population. (3) 'Universalism' means every member's project is included rather than a bare majority's. Explain why this norm emerges and how omnibus bills enforce it. (4) Design two institutional reforms that would restore cost discipline, and name the mechanism each uses.