![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
A billion a year?
Jun. 23rd, 2010 12:13 pmThe increase of capital gains tax to 28% from 18% for higher-rate taxpayers "strikes the right balance", according to Chancellor George Osborne, and will raise about £1bn a year. Well, we shall see.
Presumably the advance-flagged 40% rate was a "softener-upperer", as the blindest fool would have known that this would have been such a steep hike that it would have drastically reduced the volatility of capital. This would not only have made tax receipts from CGT fall, but would also have reduced the supply of money/capital because, as we all know, the amount of money/capital in the economy is effectively dependent on how fast it moves around as well as how much there is of it. And reducing the supply of capital wasn't really a priority in this budget.
So, 18%/28% it is. Actually, these lower rates (Osborne conceded in his speech that the 40% level would have reduced actual governmental income) illustrate an interesting economic point, that being that elasticity and inelasticity of demand are not constants when it comes to the same good. Baby economics, the level understood by, say, George Osborne, dictates that some goods have elastic demand (i.e., very price sensitive), while others have inelastic demand (i.e., not very price-sensitive). Empirically, however, it would appear that goods can be inelastic at one point of price change and elastic at another.
This, of course, is little more than applying catastrophe theory to economics. It's also applicable to poker, where Sklansky observed in one of his many books that an opponent's propensity to call tended to be inelastic at a low level, then go through a sharp rise in elasticity (the "tipping point") before moving into inelasticity again. E.G., if the pot is $30 and you both have $200 in front of you, a bet of from $1 to $18 might be fairly inelastic (say, between 85% and 99% chance of being called). Then from $18 to $40 there is a period of elasticity (20% to 85%), while all bets above $40 become inelastic again (1% to 20%). So a 16% range of possible bet sizes covers a 65% range of "call or folds".
This seems to be the modern thinking when it comes to tax revenues. You can increase them for quite a while, generating increases in revenue all of the time, but if you go too far then a "tipping point" is reached, causing a dramatic fall-off in revenue.
Osborne reckons that much of the £1bn extra that he expects to raise will come from income tax rather than CGT, because the 28% rate would reduce the proportionate gain for higher-rate tax payers to convert taxable income into capital gains income. Well, that seems dubious to me. Under the elastic/inelastic spectrum, surely a 22% edge (28% vs 50%) is still sufficient incentive to shift from income to capital gain, even if it's less than the old 32% edge. Indeed, cynics might observe that the differential is back at the old level when the top rate of tax was 40%. Capital Gains Tax, BTW, only generates about £5bn a year in revenue, less than 3% of revenue from income tax (£150bn plus). That latter number indicates why one of the biggest worries for the Chancellor is, quite simply, higher unemployment. Even if you cut benefits to zero, the Treasury foregoes a fortune in lost income tax.
So, that's one thing that doesn't look to work. But, let's assume that players carry on "converting" at close to the same level as they did before -- that still gives you a 28% rate rather than an 18% rate.
But this is where it gets interesting. Although we know that CGT brings in £5bn a year, we don't know how "lop sided" this income is. What proportion of the money comes from, say, the largest 10% of deals? If the line is relatively flat (i.e., lots and lots of roughly £500k deals), then the income increase for the Treasury should be relatively inelastic.
I would think that a majority of CGT payers are higher-level income payers, even allowing for those who are retired. Let's guess at 60%, which would make up £3bn of total CGT revenue.
Osborne's £1bn in increased revenue therefore assumes that the 36%-odd hike in rate for 60% of CGT payers will result in a 33% hike in income -- which strikes me as an overly optimistic view, even if we assume that the distribution of CGT revenue is fairly flat (lots of roughly £500k deals).
But what if it isn't that flat?
Two obvious ways to avoid the 28% hike that could be worthwhile, if you had a potentially large liability, are either to (a) wait until you retire before booking a gain (the obvious strategy for anyone in their 50s, I would have thought), or (b), to take some kind of "Gap Year" that brings you back into the lower tax bracket for the year in question.
That in turn brings in the possibility of HMRC ruling on whether a person has deliberately manipulated his or her income for a particular tax year in order to pay a lower CGT rate. Would they take an "average" of the previous five years? If so, then anyone who happened to have only recently moved into the higher tax bracket could argue that such an average should also be applied to them. A veritable minefield of tax cases, I would have thought.
The major problem with CGT has always been that it is effectively several years' income compressed into a single tax year. Taper relief was an inefficient way of dealing with this, but it was at least an attempt to deal with it. A second problem, in times of inflation, is that you can end up paying tax on a profit that is not a real profit, because the value of the measuring stick (money) has changed.
I have no idea what the tax rules are on this, but I would be interested to see what would happen with HMRC if you chose to buy and sell stuff in gold rather than in sterling. Would HMRC "convert" the gold into the prevailing sterling price at the time? That could make an interesting tax case, if you paid 200 oz of gold for something and then sold it for 190 oz of gold, without ever converting it into sterling. Alternatively, what if a deal was denominated in dollars, or Chinese yuan? If you steer clear of sterling, I would have thought any attempted taxation for a "capital gain" would be hard to stick, because you could show that no gain had been made, unless you used HMRC's ridiculous yardstick (sterling).
But, I digress.
I can't really see this raising £1bn a year, because this seems to assume zero elasticity of demand. More likely will be a small fall-off in liquidity (fewer deals made that would generate a taxable gain) and a small fall in "tax diversion" from income tax to capital gains tax. Osborne seems to assume that the increased income tax revenue will counter the fall-off in liquidity. Can't see it myself.
__________
Presumably the advance-flagged 40% rate was a "softener-upperer", as the blindest fool would have known that this would have been such a steep hike that it would have drastically reduced the volatility of capital. This would not only have made tax receipts from CGT fall, but would also have reduced the supply of money/capital because, as we all know, the amount of money/capital in the economy is effectively dependent on how fast it moves around as well as how much there is of it. And reducing the supply of capital wasn't really a priority in this budget.
So, 18%/28% it is. Actually, these lower rates (Osborne conceded in his speech that the 40% level would have reduced actual governmental income) illustrate an interesting economic point, that being that elasticity and inelasticity of demand are not constants when it comes to the same good. Baby economics, the level understood by, say, George Osborne, dictates that some goods have elastic demand (i.e., very price sensitive), while others have inelastic demand (i.e., not very price-sensitive). Empirically, however, it would appear that goods can be inelastic at one point of price change and elastic at another.
This, of course, is little more than applying catastrophe theory to economics. It's also applicable to poker, where Sklansky observed in one of his many books that an opponent's propensity to call tended to be inelastic at a low level, then go through a sharp rise in elasticity (the "tipping point") before moving into inelasticity again. E.G., if the pot is $30 and you both have $200 in front of you, a bet of from $1 to $18 might be fairly inelastic (say, between 85% and 99% chance of being called). Then from $18 to $40 there is a period of elasticity (20% to 85%), while all bets above $40 become inelastic again (1% to 20%). So a 16% range of possible bet sizes covers a 65% range of "call or folds".
This seems to be the modern thinking when it comes to tax revenues. You can increase them for quite a while, generating increases in revenue all of the time, but if you go too far then a "tipping point" is reached, causing a dramatic fall-off in revenue.
Osborne reckons that much of the £1bn extra that he expects to raise will come from income tax rather than CGT, because the 28% rate would reduce the proportionate gain for higher-rate tax payers to convert taxable income into capital gains income. Well, that seems dubious to me. Under the elastic/inelastic spectrum, surely a 22% edge (28% vs 50%) is still sufficient incentive to shift from income to capital gain, even if it's less than the old 32% edge. Indeed, cynics might observe that the differential is back at the old level when the top rate of tax was 40%. Capital Gains Tax, BTW, only generates about £5bn a year in revenue, less than 3% of revenue from income tax (£150bn plus). That latter number indicates why one of the biggest worries for the Chancellor is, quite simply, higher unemployment. Even if you cut benefits to zero, the Treasury foregoes a fortune in lost income tax.
So, that's one thing that doesn't look to work. But, let's assume that players carry on "converting" at close to the same level as they did before -- that still gives you a 28% rate rather than an 18% rate.
But this is where it gets interesting. Although we know that CGT brings in £5bn a year, we don't know how "lop sided" this income is. What proportion of the money comes from, say, the largest 10% of deals? If the line is relatively flat (i.e., lots and lots of roughly £500k deals), then the income increase for the Treasury should be relatively inelastic.
I would think that a majority of CGT payers are higher-level income payers, even allowing for those who are retired. Let's guess at 60%, which would make up £3bn of total CGT revenue.
Osborne's £1bn in increased revenue therefore assumes that the 36%-odd hike in rate for 60% of CGT payers will result in a 33% hike in income -- which strikes me as an overly optimistic view, even if we assume that the distribution of CGT revenue is fairly flat (lots of roughly £500k deals).
But what if it isn't that flat?
Two obvious ways to avoid the 28% hike that could be worthwhile, if you had a potentially large liability, are either to (a) wait until you retire before booking a gain (the obvious strategy for anyone in their 50s, I would have thought), or (b), to take some kind of "Gap Year" that brings you back into the lower tax bracket for the year in question.
That in turn brings in the possibility of HMRC ruling on whether a person has deliberately manipulated his or her income for a particular tax year in order to pay a lower CGT rate. Would they take an "average" of the previous five years? If so, then anyone who happened to have only recently moved into the higher tax bracket could argue that such an average should also be applied to them. A veritable minefield of tax cases, I would have thought.
The major problem with CGT has always been that it is effectively several years' income compressed into a single tax year. Taper relief was an inefficient way of dealing with this, but it was at least an attempt to deal with it. A second problem, in times of inflation, is that you can end up paying tax on a profit that is not a real profit, because the value of the measuring stick (money) has changed.
I have no idea what the tax rules are on this, but I would be interested to see what would happen with HMRC if you chose to buy and sell stuff in gold rather than in sterling. Would HMRC "convert" the gold into the prevailing sterling price at the time? That could make an interesting tax case, if you paid 200 oz of gold for something and then sold it for 190 oz of gold, without ever converting it into sterling. Alternatively, what if a deal was denominated in dollars, or Chinese yuan? If you steer clear of sterling, I would have thought any attempted taxation for a "capital gain" would be hard to stick, because you could show that no gain had been made, unless you used HMRC's ridiculous yardstick (sterling).
But, I digress.
I can't really see this raising £1bn a year, because this seems to assume zero elasticity of demand. More likely will be a small fall-off in liquidity (fewer deals made that would generate a taxable gain) and a small fall in "tax diversion" from income tax to capital gains tax. Osborne seems to assume that the increased income tax revenue will counter the fall-off in liquidity. Can't see it myself.
__________
