The 2016 Nobel Memorial Prize in Economics was awarded yesterday to Harvard’s Oliver Hart and MIT’s Bengt Holmstrom for their ground-breaking work on contract theory.
Their work has profound implications for contracts in many fields: financing contracts between investors and firms, business contracts between customers and suppliers, and – most topically – employment contracts between firms and workers. How can we use Nobel Prize-winning insights – rather than anecdote and shooting from the hip – to guide the current debate on executive pay?
Most academic research is empirical – using real-world data to identify correlations, and in some cases causations. Empirical research is illuminating, but typically only tells us what does happen, not what should happen. In particular, it can only study policies that are already in place.
But many of the proposals for compensation reform are hypothetical – they aren’t currently practised, so we have no data on them. This is where theoretical research – Hart and Holmstrom’s methodology – comes in. “Theoretical” doesn’t mean “abstract”. Instead, a good theory constructs a model of the firm-worker relationship, changes certain variables (such as the strength of incentives), and studies the outcome – similar to a flight or city simulator. Of course, the predictions of any model depend on its assumptions. Hart and Holmstrom’s theories have become particularly influential due to their realistic assumptions, and their predictions have subsequently been supported in the data.
Arguably Holmstrom’s most famous paper shows that employees should only be paid for performance under their control. This suggests executives should be rewarded for firm-specific but not industry performance. While seemingly obvious, this principle is often ignored in real life. For example, homebuilder Persimmon gave its executives a £600m reward for strong stock price performance when it was due to a buoyant housing market.
A paper with Stanford’s Paul Milgrom studies contracts when the worker has to balance many competing priorities. Paying according to one performance measure will cause the worker to ignore others. Again, while intuitive, this principle is often violated. Concerns to avoid “excessive” pay led to clauses cancelling executives’ equity if they miss certain performance targets – and so they take short-term actions to meet these targets. Theresa May advocates linking executive pay to worker pay, which may cause the executive to neglect other dimensions of social responsibility (such as customers, suppliers, and the environment) – and indeed other dimensions of worker reward, such as on-the-job training.
A separate paper with Milgrom shows that, under certain conditions (subsequently relaxed), the optimal contract is linear. Pay should rise gradually with good performance and fall gradually with bad performance – with no kinks or step-changes. This contrasts with MP Chris Philp’s argument that pay should have a maximum. Such pay caps incentivise executives to build good firms, but then to preserve the status quo rather than go from good to great. It also contradicts the use of options, where the executive’s downside is limited upon poor performance, and step-change forfeiture of equity upon missing short-term performance targets.
Some of Hart’s contributions are on incomplete contracts. Contracts aren’t perfect – there are many decisions that can’t be enforced by a contract. Who should decide what decisions are taken when the contract is silent? The answer is the party who bears the long-term consequences of these decisions. In a firm context, this is shareholders. Evidence shows that the long-term stock price reflects not just shareholder value, but stakeholder value. In contrast, giving decisions to employees (also proposed by May) is suboptimal. Employees are less affected by decisions that affect customers, suppliers, and the environment, and so may not take these decisions optimally.
Both Hart and Holmstrom emphasise how contracts don’t just provide incentives, but also achieve risk-sharing. Thus, even if employees are intrinsically motivated, incentive contracts are still valuable. Banker bonuses can be cut in a downturn, ensuring the bank’s solvency. The European Union’s bonus cap has replaced bonuses with salaries, which are harder to cut and thus increase bank risk. In addition, executives are less risk-averse than employees. This is why bonuses are higher for executives – and on the flipside, poor company performance leads to pay cuts for executives (Burberry’s Christopher Bailey lost 75 per cent in 2015) but not employees. This also suggests that executive pay should not be linked to worker pay.
Finally, both Laureates emphasised that the optimal contract is not one-size-fits-all – it differs from firm to firm, and from employee to employee. While regulations aim to impose common rules for everyone (for example, by shaming any firm whose ratio of executive to worker pay exceeds the average), the diversity of contracts should reflect the diversity of the economy.