The proper way to levy taxation

The full set of briefings to incoming ministers (BIMs) following the recent election are now helpfully available on a single page, and between them cover a host of quite interesting, practical, and in some cases timely economic questions. One quite meaty suggestion that I noticed in Treasury’s BIM related to taxation, but it was given very little space (perhaps they knew that it would be ignored?).

The full passage is quoted below, but the bit of interest is in the final bullet:

Over the medium term, there is a need to shift taxes from bases
that are internationally mobile and have the most detrimental
impact on growth to tax bases that are less mobile and less
damaging to productivity growth. Policy direction that will
contribute to productivity improvements and revenue
sustainability, include:
• reducing high marginal personal tax rates in order to improve
incentives for labour supply, entrepreneurship and the
retention of skilled labour within New Zealand
• equalising rates of tax on different forms of investment to
improve savings and investment, including reducing the rate of
tax on some existing forms of investment income and
introducing a tax on capital gains to reduce the diversion of
investment into tax-favoured or tax-exempt forms
• moving towards a tax system more heavily weighted towards
consumption taxes and, over a longer horizon, with a greater
contribution from property taxes.

Now I have always found our tax system (and most tax systems internationally, actually) bizarre. I’m actually not against progressive taxation, but taxing labour always seemed a bit daft to me. The only really good explanations that I have ever found for the government funding its activities in this way is that: (1) it is administratively efficient; and (2) it is easier to impose taxation progressively by imposing it on income. [A third point might be that ‘we have always done it that way’, and I think that this could be the biggest reason why we continue to levy funding for government activities through income tax].

Now I don’t necessarily disagree with these points, but I do wonder if we couldn’t manage to achieve these objectives through different tax systems if we really tried. Consumption taxes would certainly be a component of this (and already are to an extent), but I feel that they would need to be used in conjunction with other forms, since I don’t think that fully raising tax revenue on consumption is either equitable or constitutes good fiscal risk management.

Taxing externalities is generally supported in theory, but seldom introduced (barring cigarettes and alcohol and a few other product classes) and there is certainly room to do this more. I sometimes hear that taxing externalities isn’t a very good solution since you can’t fund much of the governments need for dough in this way, as the ‘economically efficient’ externality tax is only equal to the difference between social and private costs. However taxing labour is not in any way related to externalities – so why not over-tax the ‘bad’ things, since it is no worse than the status quo and at least some of it will be efficient? There could be a precautionary principal here – we don’t know exactly how large the externality is, but if we overshoot we will at least know we covered it. Again, I see how this is inefficient, just not how it is less efficient than the status quo.

  • Hi Dant03,

    This is a very important issue. I completely agree that, if we have a revenue target (which determines government spending) we should start with externality taxes.

    However, once we have these “efficient” taxes things become cloudier. Often people call on the “Ramesy principal” and state that we should tax things which will not respond, eg they have highly inelastic demand/supply curves. That way the deadweight loss is minimised.

    Now when thinking about the general “income tax” vs “consumption tax” principle there is one thing I would like to say. If both are levied in a flat fashion they are exactly the same thing.

    An “income tax” reduces peoples incomes, a consumption tax increases the price of all products – ultimately the budget constraint will move the same way in either case.

    “However taxing labour is not in any way related to externalities – so why not over-tax the ‘bad’ things, since it is no worse than the status quo and at least some of it will be efficient?”

    Going back to the Ramsey principle we have to ask – what has more inelastic demand. If labour demand/supply is highly inelastic, but demand/supply of goods subject to externalities is not – then “overtaxing” externalities in order to lower income taxes would be inefficient.

  • Note that we have written about income vs consumption before:

    http://www.tvhe.co.nz/2008/09/19/income-vs-consumption-taxes-whats-the-difference/

    This also has a link to a zillion other tax posts we have written – and we have done some more since

  • Kimble

    “However taxing labour is not in any way related to externalities – so why not over-tax the ‘bad’ things, since it is no worse than the status quo and at least some of it will be efficient?”

    Over tax the bad things and people will stop using them. Labour is, in general, a permanent source of income. Some people can avoid labour, but most cant. Labour also implies cashflow, so it is prefereable to a land-based tax that does not.

    In addition, having the primary tax on specific consumption triggers my libertarian gag reflex.

  • George Bolwing

    The proper base for taxation is a topic that has caused many trees to die.

    This is partly because it is such a wicked problem, especially once you get out of the class room and into world of politics and vested interest.

    A quick literature review:

    Smith (1776): general principles, which would today be described as balancing efficiency, equity and compliance and administrative costs. No real detailed policy prescription.

    Pigou (1920): use taxes to correct externalities. Essentially an argument about efficiency: externalities create welfare reductions and thus taxing them moves economy to a higher production possibility frontier. Good as far as it goes, but what if the state needs more revenue than it can get from correcting market failures?

    Ramsey (1927): answered a question posed by Pigou, which was “what is the best way to raise a given amount of revenue using a tax on consumption?”. His answer was also based on efficiency, and was that taxes should be impose the lowest deadweight loss, which in practice meant that taxes should be levied in proportion to the inverse of elasticities.

    Kaldor (1955): proposed a direct expenditure tax, which is a tax on labour income and economic rents, levied at progressive rates. Often called a “cash flow tax”, because on the business side, firms are taxed on their cash flows. Practical effect of this is that all capital expenditure is immediately deductible, while all capital gains are taxed on realisation.

    Diamond and Mirrlees (1971): equity and efficiency should be treated separately and taxes should not distort the production side of the economy. Lump sum taxes (i.e. a tax that is not levied by reference to criteria that the taxpayer can alter) are preferred, followed by taxes on labour.

    US Treasury (1977): “Blueprints for Basic Tax Reform”. Proposed two alternatives, a comprehensive income tax or an expenditure tax (under the name of a cash flow tax). Came down on the side of a cash flow tax. The reports principal author, the late David Bradford, would remain a strong proponent of cash flow taxes until his untimely death in 2005.

    Meade (1978): the Institute for Fiscal Studies in the UK appointed a committee, chaired by Sir James Meade, to review the UK tax system and make recommendations for review. Proposed a variant of Kaldor’s expenditure tax.

    In practice, the basic choices come down to:

    Income tax: taxes labour income and income from capital. Can be applied at progressive rates. Good for equity, as the rich can pay more. Wickedly complex on the business (capital) side, where valuation and timing problems abound.

    Indirect consumption tax (e.g. GST): tax on labour income (and accumulated wealth at the date of introduction of the tax). Simpler than an income tax, but can only apply at a single rate. Does not tax returns to capital.

    Direct consumption tax (i.e. a Kaldor expenditure tax). Taxes labour income and the rents from capital income (i.e. supernormal profits). Also a lump sum tax on accumulated wealth at date of introduction. Can be levied at progressive rates. Problem of transition from existing income tax probably impossible to solve.

    The NZ system of a broad-base, low rate income tax and a comprehensive GST is probably as good as it gets.

  • Hi George,

    Thanks for the clear run down.

    I am not sure I agree with the idea that an indirect consumption tax does not tax returns on capital. Implicitly it does, as it increases the price level of final goods – thereby reducing the real return on capital.

    “The NZ system of a broad-base, low rate income tax and a comprehensive GST is probably as good as it gets.”

    I agree. However, that does not remove the possibility that a set of externality taxes could help to reduce the burden of other, distortionary taxes. I think policy that aims to push externality taxes towards their optimal level is always going to be good policy – as long as the optimal level can be identified 😛

  • Kimble and Matt: Agree that labour has the advantage that it is pretty inelastic. However while it may appear that consumption is not (at least relatively), I suspect that its apparent elasticity is more in terms of relative prices – if it were imposed across the board there wouldn’t be much change in consumption levels (ie net income would be correspondingly higher, and relative prices would be the same with a uniform consumption tax). Ie if you were to cut income tax by an absolute value of x and impose consumption tax of x, I think that the net amount of goods and services brought would not change. Its just that slapping the tax on just some products would cause a shift away from those products.

    The risk with consumption taxes is not that consumption is particularly price elastic to the tax, but that consumption jumps around for a variety of reasons in ways that labour is not as prone to do (ie interest rates, GDP, business confidence). For this reason we can’t rely on consumption taxes too much to fund the government’s activities. Also, if you support progressive taxation (I do) you still need an element of income tax – but surely not the over-reliance we presently have? I think that there is room for more externality taxes and more consumption taxes, and less income tax.

    A distasteful feature of income tax is that it is prone to wasteful avoidance that externality and income taxes are not. Ever met the high income earning individual who pays negative taxes? A consumption tax disallows such distortions provided money is spent. Capital gains taxes as a complement help to ensure that distortions are avoided. Externality taxes when properly set cause efficient reductions in the activity that the tax is levied on, reducing the tax take, but thats okay, we want that, and the overall tax take is buffeted by other elements (consumption tax, some residual income tax). Both types have superior elements to income taxes, with the latter apparently only dominating because of ‘ease of imposition’. I think that there is a part for income taxes, just not that it should be so big a part. We are smart enough to make a more comprehensive system workable!

    I guess the idea in focusing on non-mobile sources of taxation than labout is to improve our relative competitiveness (obvious?)

    George: thanks – I feel illuminated!

  • “Agree that labour has the advantage that it is pretty inelastic. However while it may appear that consumption is not (at least relatively), I suspect that its apparent elasticity is more in terms of relative prices”

    I said that a consumption tax is probably just as inelastic – as it is “effectively” a real income tax. The distortion I was discussing stemmed from the fact that differing elasticities in different markets would see the relative price of goods change following the tax. However, we have the same inefficiency in the labour market when we compare different labour types.

    “The risk with consumption taxes is not that consumption is particularly price elastic to the tax, but that consumption jumps around for a variety of reasons in ways that labour is not as prone to do”

    However, remember that government has access to well functioning credit markets – even in this time of high credit dislocation. I am sure that a little more income volatility would not hurt government given its ability to borrow and thereby “smooth” its spending over time 😉

    “A distasteful feature of income tax is that it is prone to wasteful avoidance that externality and income taxes are not”

    True, but before selling them we need to try and figure out if there are ways that consumption taxes will be avoided – how will people avoid stating that something is a “final good”, what will happen to the “black market” for some products?

    “We are smart enough to make a more comprehensive system workable!”

    We might be smart enough – but if the compliance cost of “making it work” exceeds any efficiency gain doing so would not be a “smart” move.

    “I guess the idea in focusing on non-mobile sources of taxation than labout is to improve our relative competitiveness (obvious?)”

    Although we should be careful not to take this to an extreme – it is more of a message that supports integrated international taxation policy, rather than a message that we should slash taxes on mobile inputs.

  • George Bolwing

    In practice in New Zealand, GST is a very stable tax. It tracks nominal GDP pretty closely. This is not suprising, given that consumption is the largest component of GDP!

    PAYE taxes are also pretty stable, again because employment does not tend to swing radically. Most people who are employed in one year tend to be employed at about the same income the following year.

    The least stable tax is that on business profits, especially from unincorporated and small businesses. The business tax base is an interesting mix of economic, accounting and legal concepts. The economic cycle can have a large impact here, as can “tax mitigation” (the polite word for arranging your affairs to minimise the amount of tax you pay).

    To put this in prespective, in the year to June 2008, the revenue streams were:

    PAYE: $23bn
    GST: $11bn
    Corporate taxes: $8.5 bn
    “Other persons”: $5 bn

    Full data is here; http://www.treasury.govt.nz/government/financialstatements/yearend/jun08/fsgnz-year-jun08-pt5of6.pdf

  • George Bolwing

    On the point of consumption taxes not taxing income from capital, this was loose language on my part.

    the more correct statement is that a tax on income from capital taxes the returns to savings, while a consumption tax does not.

    Cash-flow, consumption and expenditure taxes have in common the feature that tax is levied only at the point in time when individuals spend their money and consume what they have previously earned and saved, or, indeed, spend what they have borrowed.

    The economic effect is that the return from savings is not taxed – removing the tax-driven incentive to consume now rather than later.

    Under an income tax (which by definition includes interest as income), however, the return on savings is taxed, which means that there is an economic disincentive to saving. Indeed, the difference between an income tax and a consumption tax lies entirely in the treatment of savings.

  • Hi George,

    But a consumption tax increases the price of future consumption – which is the goal of savings. As a result, it reduces the real return on savings in the same way an income tax would.

    Whether you are paying the tax on your “income” from savings, or you are paying the tax from the consumption you derive from that savings doesn’t really matter – in either case you are paying tax.

  • Although, the existence of a tax directly on savings implies that there is a “wedge” between the interest rate savers get and the rate borrowers pay – in the case of the consumption tax there is no such wedge, as this now occurs between the buyers and sellers of products instead.

    I’ll give you that part of the efficiency discussion 😉

  • I happily agree with the points you (Matt and George) make about consumption tax not being so unstable a source of government revenue after all. Both good points. I was repeating an argument I had heard elsewhere and wasn’t completely comfortable about it – now glad to be put right.

    I think that I am with George overall on the tax on savings issue, but it doesn’t seem like it would be a big effect.

    Matt: We might be smart enough – but if the compliance cost of “making it work” exceeds any efficiency gain doing so would not be a “smart” move”

    Yes of course, trite point. I am increasingly of the view though that if implemented with care and (importantly) enough impartiality, it would be at low enough cost to be an improvement. There have been generations of refinement with income taxes and we still can’t get it completely right. A phased introduction of a new system would allow institutions to adapt and risks to be minimised. Besides, as the stock arguments for efficiency as a reason for retaining such a large chunk of tax on income seem to keep falling over, I wonder whether the administrative cost argument is likewise a red herring. But of course it would need to be considered carefully.

    Matt: “True, but before selling them we need to try and figure out if there are ways that consumption taxes will be avoided – how will people avoid stating that something is a “final good”, what will happen to the “black market” for some products?”

    GST is done pretty well I think. You need to be hard-headed about these things, and if so it will innevitably be less abused than income tax, which is by its nature less transparent and open to abuse. I’m also a fan of jail time for people who knowingly skip GST payments of more than trifling sums… just putting that out there though.

    Overall, at the moment I am dead in favour of reducing income tax and countering that with increased consumption taxes and higher and broader externality taxes. Some income taxation should be retained so that progressivity in income gathering can be exercised.

  • George Bolwing

    My order of tax reforms would be:

    a) cut government spending by removing those things that clearly have no public policy justification. Big ticket items would be a lot of tertiary education funding — while there are some externalities if this area, they are nowhere near as large as the current level of spending would imply; a lot of science funding – ditto – and arts and culture – also ditto. Given some time, I think I could come up with a few billion in savings per year;
    b) put GST up to 25%, in a few stages. Adjust social welfare payments and New Zealand Superannuation to maintain real purchasing power. Given that we are an island nation, which makes smuggling hard, the evasion would not be that much of a problem. Internet is a problem for things that can be downloaded – things that have to be sent by post could be inspected at the entry point for a reasonably small cost;
    c) move as quickly as fiscal prudence would allow to a personal tax rate structure of 30/15, with the company rate at 30%. This would require a strong commitment to expenditure constraint, but could be done.