Static and dynamic efficiency measure different things and can trade off against each other

Static efficiency asks whether existing resources are well-allocated right now. Dynamic efficiency asks whether the system generates new knowledge, products, and possibilities over time. Improving one can damage the other.

Explanandum

Why can a market that scores well on productivity-based competition metrics (like Olley-Pakes) still be producing worse long-run outcomes than a monopolistic alternative?

Substance

The Olley-Pakes decomposition measures static allocative efficiency: are the most productive firms capturing the most market share right now? But this is silent on dynamic efficiency — whether the system is generating the innovations, research programmes, and new entrants that will matter in the future.

The Bell Labs case is the sharpest illustration. AT&T’s monopoly era produced the transistor, information theory, Unix, C, and the discovery of cosmic microwave background radiation. The post-breakup competitive era improved Olley-Pakes covariance in equipment manufacturing but gradually destroyed the institution that produced those breakthroughs. The market became better at selecting among existing firms while becoming worse at generating things worth selecting among.

Schumpeter’s argument was that monopoly profits fund dynamic efficiency — firms need above-normal returns to justify risky fundamental research. This has better explanatory fit than the standard competition framework for industries with high R&D intensity and long development cycles (pharmaceuticals, semiconductors, fundamental research).

Supports

• Bell Labs produced most of its landmark innovations during the monopoly period, not after breakup
• The 1956 consent decree required free licensing of AT&T patents — monopoly rents funded research whose fruits were widely distributed
• Pharmaceutical innovation (Albrecht’s Ozempic example) relies on patent monopolies to justify R&D investment
• Intel’s decline illustrates the reverse: competitive pressure eventually arrived, but by then dynamic efficiency had already eroded

Challenges

• Most monopolies don’t produce a Bell Labs; AT&T was a regulated monopoly with unusual structural incentives
• The counterfactual is unknowable — perhaps competitive telecoms would have produced different but equally valuable innovations
• Static efficiency gains benefit consumers immediately and measurably; dynamic efficiency claims are speculative and long-term

Open Questions

• Is there a way to measure dynamic efficiency with the same rigour Albrecht brings to static efficiency?
• Does the trade-off depend on the type of good (network goods vs consumer goods vs research-intensive industries)?
• Could institutional design (e.g. DARPA, prize systems) capture the benefits of monopoly-funded research without the costs?

Source Context

Emerged from discussion of Albrecht’s use of the AT&T breakup as his cleanest case study, and the observation that Bell Labs’ research heyday was during, not after, the monopoly period.