A solution to the Problem of High Pay

I think I agree with Richard Murphy [1]. I feel faint. He suggests removing tax relief on salaries above a certain level. I think he means that when taxable profits are calculated, those salaries will not be counted as costs and subtracted from revenues, so will be taxed twice, once as profit and once via income tax. That will make it much more expensive for shareholders to employ high-paid staff. He suggested basing the salary ceiling on median earnings; I’d suggest tying it to the minimum wage.

 The old economic model of wage determination is that workers are paid their marginal product. If hiring the worker will add £X to your bottom line, it makes sense to pay them up to £X and if the labour market is competitive, £X is what they will be paid [2].

 Imagine that top salaries were capped at £250,000. Now imagine the sort of person you’d be able to hire to run your bank for that wage. How much more productive is the sort of person currently being hired for an average of $9.7m? My guess is: not enough to justify the salary. Under the proposed scheme, salaries would not be capped, but the cost to shareholders of paying salaries above a certain level would increase.

 A more recent economic model of wage determination is to suppose that the combination of workers and capital in a firm generates a surplus, and workers and shareholders bargain over that surplus. The surplus generated by a bank or a multinational is breathtakingly enormous. My contention is that the bargaining power of top executives is so strong, and the surplus up for grabs so large, that salaries have completely cut loose from any idea of marginal product and represent a market failure that anybody who thinks of themselves economically minded can not justify.

 I’m going to make ad hoc appeals to irrationality that will alienate any of the aforementioned economically minded. It’s no mystery how top executives pay themselves that much – they run the show – the mystery is why shareholders don’t discipline them. Shareholders worry about share prices falling. If the cost of disciplining executives who are looting the company is higher than the reward, they won’t do it.

For whatever reason, the idea exists that “leaders” of “global calibre” are a valuable asset to a firm [3]. Now, say you are a major shareholder who thinks this is bullshit and wants to contest pay deals. First, merely the disruption and uncertainty might knock the share price enough to make it better just to pay the swine their millions. Second, if your supposed global leaders do go off in a huff, which they very well may if you are moving unilaterally, your share price may take a really big hit. So it doesn’t happen [4].

I think the only thing restraining executive pay is the benchmark of other executives’ pay – what is the going rate? This is a multiple equilibrium situation [5]. I don’t know what governs the evolution of the going rate, but I don’t think it’s the marginal product of “global leaders”.

 To make concrete just how skewed the cost benefit picture is for shareholders wanting to restrain executive pay, if you own 10% of a £1bn market capitalization company, then cutting executive salaries by a whopping £10m could add £1m to your annual dividend. You lose £1m if your shares fall just 1% per cent. This is why shareholder activism on executive pay is rare. I suppose a corollary of this is that Richard’s proposal, which will merely increase the cost of high pay to shareholders, may have little effect. [paragraph somewhat edited, see comments]

 Of course, if shareholders of every corporation cooperated and agreed to pay no more than £X, global leaders would have to take £X or leave the game. This is why I am attracted to legislation that acts as a coordinating device to correct this market failure. Ideally high pay legislation is something that would happen globally, but I hope that the UK has a sufficient quantity of high ability individuals who would be prepared to work for a few hundred thousand pounds a year (peanuts!) that it is something we could do on our own.    

 Here’s a puzzle. According to most political economists, what motivates politicians is getting elected. My guess is this policy would produce howls of anguish from certain quarters, praise from the bien pensant lefties at those crucial Islington dinner parties, and be a smash hit at the ballot box. What’s stopping Labour?  

 Update! Tim Worstall informs me that something similar to this idea has been enacted in the USA since the days of Bill Clinton. If true, this suggests two things: 1. It doesn’t restrain high pay and 2. it is politically feasible and firms will not decamp overseas.

[1] With the proviso that I expect people to point out pragmatic reasons why what he proposes wouldn’t be feasible – most likely to do with creative lawyers and accountants cooking up other ways for company directors to make out like bandits. That, and every large UK company would instantly relocate its head office overseas. I use the world salary here: I mean to include any method of remuneration including payment in shares.  

[2] This is a useful start, but a few moments thought tells you there must be more to it than that. For example, with certain production functions, say when people have to work as teams, the concept of the marginal product of a single worker is not meaningful.  

[3] Possibly because it’s true, but I doubt it. This is my appeal to irrationality – I’m saying a false belief is prevalent amongst investors.

[4] Investors (shareholders) also rely on good relations with management. If rocking the boat results in being denied some insider tidbits guidance from management, it’s not worth it.

[5] The “bargaining over surplus” model supports multiple equilibria, within boundaries set by the size of the surplus and the participation constraints of the parties involved (so what matters is the “outside option” – if any party can do better elsewhere, they’ll leave). Nobody has a very good theory of where, within this feasible set, the outcome will lie.    


10 thoughts on “A solution to the Problem of High Pay

    1. I believe by 5 or 6 orders of magnitude. Slipped through my rigorous ctrl+c ctrl+v editing process as I put this up while on lunch.

      I’ve updated it, Luis, is that correct now?

  1. oh good Lord, £10m is 100th of £1bn isn’t it? but seeing as my hypothetical shareholder owns only 10% of the £1bn, to wipe £10m off that 10%, the shares would need to fall, um, 10%

    I don’t think that paragraph works at all now! how on earth did I get it that wrong? I did 10 not 10m didn’t I.

    Right, well I don’t see why shareholders don’t cause more of a ruckus.

    um, I still think shareholders don’t cause a fuss because they don’t see it worth their while. I suppose in the example is wrong to think £100m could be cut from the annual salary bill – that figure is too high. If it was more realistic to cut £10m, I think that means the shares would have to fall 1%, which isn’t too hard to believe happending if you lose your CEO.

    L.O I don’t know how you are going to salvage my dignity – either chop the para, or re-do with £10m cut and 1% fall. (is that right?)

    1. “I think that means the shares would have to fall 1%, which isn’t too hard to believe happending if you lose your CEO.”

      Certainly I can believe that. But, on the other hand, the dividend is an annual dividend. The value of the shares is a one-off gain if you decide to sell the shares at their current price.

      So: year 1 – gain 10m in increased dividend, lose 10m in lowered share price
      year 2+ – gain 10m in increased dividend

      (That’s assuming that expensive CEOs and cheap CEOs are no different in quality and therefore the long-term movement of the share price would be the same under either, excluding the cost to change between them)

  2. yeah, i realised I was comparing an annual thing with a one off, but thought it would do as a rough thing, but I really need the NPV of a £10m income stream. Oh dear.

  3. I suppose £10m is still quite a lot to trim from the execuative salary costs of a £1bn market capitalisation company. If you’re trading on a PE ratio of 10, that £10m saving, if added to earnings, is worth £100m, 1/10th of the value of the company!

    Ah, that’s the key isn’t it – you wouldn’t expect cost savings just to be added to earnings in perpetuity, which I think makes the one-off / recurring comparison more acceptable.

    Barclays is valued at £30bn. I think with realistic numbers, it isn’t worth a shareholder’s while to cause a fuss. Well, I’ve managed to convince myself the argument remains sound, as for the post itself, bit of a mess now. Sorry everybody.

  4. I’ve edited away again Luis, better now? I still think your point stands.

    The risks and rewards are definitely a bit skewif, so I think your point still stands.

    Plus the risks are more heavily weighted against action, losing a CEO is almost always bad in the short term, as it creates a lot of uncertainty. Just sucking it up and foregoing some income is easy, better to be conventionally wrong, than unconventionally right.

    1. thanks LO,

      unfortunately I think it’s wrong to say “just” a 10% fall in the shares – that’s a huge fall. Also £100m is a very silly amount of money to cut from executive salaries in a £1bn company – can you change it to a £10m cut (which is still too much) and a 1% fall in the shares?

      sorry about this

  5. Agreed, certainly not worth causing a fuss. If nothing else, the share price is going to be volatile by several percent over the course of a year just due to random fluctuation, which is likely to mess with both your dividend and the paper value of the shares themselves far more than the staff saving would.

    And, of course, if you’re getting 10m in dividends off the shares, themselves valued at 100m, going through huge amounts of hassle to potentially get an extra 1m is a waste of time even if it was completely risk-free.

  6. This is the annoying thing (cim totally not meant as a dig). A minor mistake can completely obscure the main thrust of a blog post. I’ve seen so many comment threads derailed by nit picking. Granted this was a rather large (100,000,000%) nit.
    Btw, all edited.

Comments are closed.