A charity can function as a 'middleman': a donor may give to a charity not only to constrain the use of aid but because the donor is ignorant about how to find real poor people or what they actually need, so the charity supplies the knowledge and matching the donor lacks.

Nonprofits as commonly conceived are better understood as an alternative to government rather than to for-profit firms, because they not only charge nothing but often have 'negative charges'—handing out money or aid—providing services governments might otherwise provide.

Costly charity stunts (walking 25 miles, shooting a thousand baskets, working a fundraiser booth) are economically irrational as production but persist because they function as costly signals of love and commitment to a cause—it is important that the activity be costly and meaningless, since that is precisely what makes it a credible signal.

The market for charitable giving suffers from a 'lemons' problem: because donors cannot easily observe whether a nonprofit wastes money on executive perks (limousines, Learjets), a few visible scandals can dry up contributions across the sector, which is why rating agencies disclosing administrative-cost percentages help solve the information asymmetry.

Nonprofit status may exist as a commitment device—a way for an organization to 'bind itself to the mast' and credibly signal it cannot extract profit—but in the modern legal setting this commitment is weakened because nonprofits can pay employees heavily and provide non-monetary perks (prestige, influence), making the commitment partly romantic rather than real.

Both for-profit and nonprofit organizations can credibly profit from signaling that they do not care about money: a nonprofit raises more donations by claiming all funds pass through to recipients and its staff are monastic, while a for-profit like Merck makes more by remembering 'medicine is for the patient, not for the profits—the profits follow.'

Hayek's later work extended beyond the 1945 information-and-prices argument to suggest that a wide variety of voluntary private organizations (including nonprofits) might solve social problems, and that the way to discover whether such forms work is to observe whether they survive and attract enough donations or revenue to operate—an evolutionary, Hayekian test that economists cannot specify in advance.

It's a pleasure to be on EconTalk again.

The original explanation for selling automobiles through franchises—that imperfect capital markets in the early 1900s and 1930s forced manufacturers to raise capital via franchise fees from locally-collateralized dealers—is a plausible but false story, because the practice continued even after auto companies were well capitalized.

It is necessary to close many unprofitable GM and Chrysler dealerships and there is no reason taxpayers should prop them up; bankruptcy is the important mechanism that finally allows this restructuring despite the human cost to dealership owners.

Haggling is a mechanism for price discrimination: the salesman's goal is to elicit the maximum price the buyer is willing to pay, which sellers can exploit when there is a gap between the buyer's maximum and the seller's minimum; this is why Priceline-style 'name your price' systems can actually disadvantage consumers who reveal their maximum.

GM cannot unilaterally terminate a franchise contract; by state law the franchisee is generally the only party who can end the contract unless the manufacturer can show cause (e.g., the dealer underperformed).

Ford is doing better than GM and Chrysler, appearing successful and widely sold in Germany/Europe.

The dealership franchise laws were originally enacted, at least partly in good faith, to reduce the perceived power imbalance between large corporations and small local businesses—the supposed backbone of America—but produced the opposite effect by entrenching dealer power.

GM has only two profitable lines, Chevrolet and Cadillac, with the others ranging from a small loss to an enormous loss, yet contract provisions and state laws prevented closing the unprofitable lines.

Franchise contracts included regionally exclusive licenses (no competing same-brand dealer within a certain distance), which made the franchise more valuable and kept franchise fees higher than they otherwise would be.

The political power in the manufacturer-dealer relationship is the reverse of the intuitive picture: rather than the large corporation dominating small dealers, local franchise dealer associations hold disproportionate political clout (especially via contributions to House members), and bankruptcy is the only way for manufacturers to escape these constraints.

In 46 of the main U.S. states it is illegal for automobile companies to vertically integrate by owning their retail dealerships directly; dealers must be locally owned.

State laws prevent manufacturers from closing a dealership without the dealer's permission unless they pay back the entire franchise fee, which forces GM to keep manufacturing unprofitable lines (like Pontiac) because stopping production would trigger massive franchise-fee repayments to dealers nationwide.

In any competitive market only the marginal buyer and marginal seller are indifferent at the market price; almost every buyer would have paid more and almost every seller would have accepted less, so money is 'left on the table'—but the gap is normally irrelevant because competition forces a single market price and buyers capture consumer surplus.