"It is a great comfort to have you so rich" - Jane Austen, Pride and Prejudice.
Stripped down to the barest essentials, taxation works like this:
taxes owing = tax rate * tax base + lump sum taxes – income guarantees and tax credits.
The appropriate tax base, rate and income guarantee are the three fundamental issues in tax design.
Many economists believe that consumption is the best possible tax base (see this post, and the comments). Consumption is the best measure of what a person takes, consumes, and enjoys; consumption measures her welfare, how well off she is. It is thus the fairest basis for taxation. Moreover, these economists argue, taxing consumption has desirable incentive effects: it encourages saving; it discourages excessive, wasteful, environmentally destructive spending.
Consumption taxation could be achieved by allowing individuals to make unfettered contributions to registered retirement savings plans (RRSPs in Canada, IRAs in the US) and taxing all withdrawals from those accounts. Or relying more on sales taxes like the GST and HST.
There are well-known limitations to our ability to tax all consumption. What is produced and consumed in the home, for example, falls outside the tax system – something which is a problem for income taxes also. It can be hard to distinguish between business expenses, such as home offices or travel, and personal consumption.
Less often observed is that wealth itself generates consumption benefits, even if one never spends a dime of it.
I own a 12 year old Toyota Matrix. The front fender has collided with one too many snow banks, and is now held together with string. The exhaust system has seen better days. It breaks down occasionally. But overall it's very cheap to run.
If I was poor, it would be tough having an old, unreliable car. The unexpected, yet inevitable, major repairs would be a financial nightmare. $750 to repair the clutch. $200 to fix the axel seal. If the car broke broke down, and I couldn't get to work, I might lose my job.
But because I'm financially secure, I can afford a cheap car. I can self-insure against financial risks: unexpected repair costs, taxi fares, rental cars, and so on. I can afford to get my car towed. If it was beyond repair, I could get another car tomorrow.
The real value of having $10,000 in the bank isn't $200 in interest income, or the stuff $200 in interest income might buy. $10,000 in the bank creates a little bit of room to take risks. One could call it the "implicit value of self-insurance generated by own capital." It's the comfort of being rich (or having rich relatives). It's real. It's valuable. But it wouldn't be taxed if Canada had a consumption tax.
Admittedly, the insurance value of having wealth isn't taxed under an income tax either. But at least under an income tax some of the return on wealth is taxed, so there is, at least potentially, some shifting of the tax burden onto those with wealth.
The greatest freedom money offers is the freedom to walk away. Your bank doesn't offer you unlimited everything with no monthly fees? Walk away. There's always someone else who wants your money. Your phone plan is too expensive? Walk away (o.k., that may not be the best example).
People with money have alternatives, which makes their demand for goods and services elastic. Food may or may not cost more in poor areas. But a rich person can shop at Value Village if he chooses. A poor person may not be able to afford expensive purchases which save money in the long run, like bread machines or high efficiency appliances or pressure cookers. Consumption taxes aim to tax the amount of stuff people actually consume. But if poor people pay a higher price for their stuff than rich people, is a system that taxes only consumption spending, without taking into account the ability to command consumption wealth conveys, fair?
There is a more subtle benefit to having wealth: the power it gives in social interactions. Becker's Rotten Kid Theorem explains how those how have wealth can control the behaviour of others by giving or withholding transfers. Shakespeare's King Lear is a vivid example. As long as King Lear has wealth, his two eldest daughters flatter him and treat him with loyalty and devotion – because they hope their obsequious behaviour will, in time, be rewarded. Once King Lear gives over his kingdom to them, Regan and Goneril cast him out.
Some might argue that taxing consumption taxes capital – once that capital is spent. But wealth generates benefits for the holder even if the holder never spends a cent. Canada has relatively low taxes on capital – we do not have an inheritance tax, do not tax capital gains on principal residences, provide dividend tax credits to offset corporate tax paid, and provide room for tax-free savings within pension plans and tax free savings accounts.
The noted economist Tony Atkinson has recently made the case for introducing an annual tax on wealth. His argument is that taxing wealth would reduce inequality.
Even those who find Atkinson's argument for wealth taxation on purely distributive grounds unconvincing, and believe that consumption is the most equitable basis for taxation, should still be open to the idea of wealth taxes – because such a tax would recognize the comforts of being comfortably off.
Worthwhile Canadian Initiative 
Consider two people A and B with zero initial wealth and equal lifetime earnings. They are equally wealthy in present value terms. A consumes all his earnings immediately while B saves some and, hence, gains some insurance value from his savings. I believe that Frances’s argument boils down to saying that a consumption tax will tax all of A’s consumption but will not tax the self-insurance value that B consumes because such insurance is home produced. (If B purchased such insurance from an insurer, then the insurance premiums would be taxed.)
While I would concede that point, I don’t see how B’s home-produced insurance is any different from other home production that is also left untaxed. A handy person benefits not only from (untaxed) self-repairs of home appliances but also an insurance value that derives from the ability to fix appliances during times when no external handyman is available for hire. Similarly, a home cook or gardener derives (untaxed) insurance value from being able to satisfy hunger during hours and holidays when grocery stores and restaurants are closed. The insurance value described by Frances seems to be a special case of home production rather than a separate and different benefit associated with financial wealth per se.
The King Lear presumed social benefit of wealth also seems to result from cognitive illusion. True, King Lear can gain his daughters’ flattery with the promise of wealth transfer. However, he is not able to gain the promise of more wealth (or anything else) by flattering his daughters. (Presumably, his daughters do not value his flattery or else he would have been able to use it to avoid being cast out.) The exchange of wealth for flattery is a symmetric one. One could just as easily use the example to argue for a charisma tax as a wealth tax.
BC – “Consider two people A and B with zero initial wealth and equal lifetime earnings. They are equally wealthy in present value terms. A consumes all his earnings immediately while B saves some and, hence, gains some insurance value from his savings.”
Overwhelming majority of variation in wealth comes from variation in lifetime earnings and variation in inheritance. Family structure (which influences consumption patterns) and return on saving also explain some differences in wealth accumulation, but most of it is earnings and inheritance.
I’m not sure how much insight one can get into this issue by assuming away the stuff that matters empirically.
“I don’t see how B’s home-produced insurance is any different from other home production that is also left untaxed”
It isn’t any different from other home production: for all untaxed good and services there is generally some possibility of increasing the efficiency and/or the equity tax system if one can find some way to tax them either directly or indirectly. For example, the reason that a tax on earnings distorts labour supply is that leisure and household production is untaxed – if one could tax the value of leisure and household production, then a tax on the value of earnings+leisure+household production would be close to a non-distortionary lump-sum tax on ability. Which is an efficiency and equitable way of designing a tax system.
The collapse of marriage in our poorest communities — and its tragic impact — is a familiar story. But increasingly, marriage is becoming a marker of class privilege in America, something increasingly reserved for the affluent. If progressives want to tackle the scourge of inequality, then the retreat from marriage is an issue they can’t ignore.
FW: “Overwhelming majority of variation in wealth comes from variation in lifetime earnings and variation in inheritance.”
Yes, but much of the variation in earnings and inheritance is already captured by variation in consumption (plus consumption that one leaves to heirs). Thus, those variations are already captured by a consumption tax. Considering the merits of a wealth tax in addition to or as an alternative to a consumption tax requires examining what a wealth tax does that consumption taxes do not do. That’s why one needs to look at the differential impact of a wealth tax on two people with identical lifetime consumption. Controlling for lifetime consumption, differences in wealth are primarily due to timing of consumption, i.e., decisions to save vs. consume immediately. A wealth tax is (differentially, relative to consumption tax) a penalty on savings.
FW: “It isn’t any different from other home production…”
My point was that you were trying to give a good reason to tax wealth but ended up giving a good reason to tax home production. But, the wealth tax is not a tax on home production. It’s not clear that self-insurance is even a significant fraction of home production compared to cooking, home repair, etc. It’s also not clear that taxing wealth is a particularly good way to tax home-produced insurance. Two people with identical wealth could face very different actuarial probabilities of the insured events that you mention, and the wealth tax makes no attempt to account for that (which is required for determining what fraction of wealth goes to producing insurance and what fraction is merely delayed consumption, which will be captured by consumption tax). If one sought to design a tax on home production, it seems unlikely that one would end up with something that looks very much like a wealth tax.
BC,
Great points.
I might also add that Frances is arguing the use of wealth as a tag for other things that she would like to tax, but which are hard to tax directly. This is a strict utilitarian approach to redistribution and thus any information (in this case wealth) that makes the redistribution more efficient must necessarily be optimal. But we run into thorny problems with this line of thinking. Why stop at wealth as a tag and not tags of ability? The center of the shrubbery maze is that Frances wants to redistribute from high ability to low ability. Taxing any tag that correlates with the ability generate wealth must also increase the efficiency of the tax policy. Thus, if Frances wants to use wealth as a tag, why not use race, height, religious affiliation, or genetic markers? Strictly, Frances’ application here does not allow for horizontal equity as a consideration, and from a consistent utilitarian perspective, nor should it. The thinking that says “hey, let’s tag that [X,Y,Z] because it’s more efficient” is actually a very questionable approach to taxation. In my view it’s outright dangerous. Just think of how ugly the end point of that logic gets – you tax Jews and people of strong Protestant work ethic more, you tax taller and more attractive people more, you tax men more, you tax white people more, etc. and you do this regardless of actual market outcomes because all these groups had the potential or ability, on the average, to generate more wealth. Strict utilitarian approaches force you into this very ugly box.