5. What the Research Actually Tells Us
The strongest lesson is not that one commission model is universally right or wrong. The evidence is more useful when it helps separate questions that are often bundled together.
Percentage and value are not the same question. A percentage tells you how the fee is calculated. It does not, by itself, tell you the level of service, the complexity of the transaction or whether the arrangement represents good value.
Brokerage economics and consumer price are related but not identical. Franchise royalties, splits, desk fees and corporate overhead affect the economics of a brokerage or agent, but they do not prove that every internal cost was passed directly to the client.
Historical U.S. commission patterns do not automatically describe the post-2024 market. The CFA research is useful for understanding the previous structure, while the NAR practice changes materially altered how buyer-broker compensation is communicated and contracted.
Economic models explain incentives, not individual outcomes. The Richmond Fed model helps explain how compensation design could influence search behaviour and welfare. It does not prove that every buyer is better off with hourly or à-la-carte representation.
A large brokerage is not automatically better or worse. What matters is what the consumer is agreeing to pay, what services are included, who will perform them and how the arrangement fits the transaction.