The IFS got Nuffield to look at earnings inequality. At which point they note that high earners tend to gain their incomes - more than the general population at least - from business activities. Also, that business activities tend to have lower tax rates than labour incomes. Well, yes, they would:

According to the report, business income – from either self-employment or owning and running a company – accounts for 21% of total incomes for the top 1% of adults and 29% for the top 0.1%, compared with just 9% for the rest of the population at large.

It said business owner-managers could effectively choose to take income out of their company through the form of a salary, dividends, or capital gains – allowing them to benefit from lower rates of tax.

The report pointed to a preferential 10% rate of capital gains tax, called business asset disposal relief, while saying that company owner-managers were able to access tax rates of just 27% on income taken in the form of capital gains.

The reason for this is that incentives are the expected return from an activity, not the actual return. This must be so - the incentive is the driver of the decision to try it, at the point of trying it there is an expectation of what the return will be, not a hard and fast actual return. What gets people to do things is what they think they’ll gain from doing it - again, obviously.

There’s risk involved in going into business. Just to trot out some not wholly accurate but well known statistics - only one in five of new businesses survives to its fifth birthday, VC investors expect 9 of 10 portfolio companies to tank. We’d not want to have to prove this but would at least suggest that the both modal and median experiences of trying to set up a business are money losing - certainly when opportunity costs are included.

Or, an actual number, in Jan 2022 there were 1,560 company insolvencies. A pretty average monthly number and one that only includes those business attempts large enough to bother with such a process.

Imagine we’re trying to design a tax system from scratch. We would want the expected return from innovation to be higher than that from labour. Even if the actual return were lower - we probably could assume a certain over-enthusiasm to carry us over that gap. Given the high risk of not making any money at all therefore the tax rate on money from successful innovation needs to be markedly lower than that on labour incomes.

In fact, given the high risks involved our taxation of innovation income is almost certainly too high. Even if that’s not the point that Nuffield and the IFS are making it is still near certainly true.