Government Spending Is Bad For Growth

‘Big government is bad for economic growth.’

No shit, Sherlock, i hear you cry.

But this isn’t my view, nor one of a neo-liberal free market think tank.

It’s from that respected inflation-fighting institution, the European Central Bank, which has come to this conclusion all by itself. No surprise to readers of this blog but papers like this will help spread the gospel of smaller government.

The paper concludes that each additional 1% of government spending reduces growth by 0.13%.

One interesting finding. The taxes that have the least harmful effects on growth are income taxes. Those that hinder growth the most are consumption taxes and government subsidies.

Something to chew on.

via UK left-libertarian and freakonomics-style blogger, Chris Dillow.

35 thoughts on “Government Spending Is Bad For Growth

  1. You say that the worse taxes are consumption taxes and government subsidies. However the paper seems to say consumption taxes and social contributions. I take the latter to mean social security taxes. Am I wrong?

  2. It seems that revenue reductions that occur mainly in terms of indirect taxes and social contributions, and cuts in
    government consumption and subsidies may contribute positively to fostering economic growth in the country samples analysed.

    That’s right, Terje. All three – consumption taxes, social security taxes and subsidies.

  3. You are right this is counterintuitive, I have skimmed through the paper but unfortunately it doesn’t really attempt to explain why this result for indirect taxes might have occured.

  4. So the question becomes how highly do you value economic growth? I certainly place it up there in the top 5 important goals of government policy, but I can’t see any justification for placing it at #1. Which means that as long as the government spending that slows down economic growth can be shown to be helping wrt other high-priority goals, it can still be justified, much as I would like to see it lowered in many areas.

  5. Income taxes don’t slow growth? This might be a methodological result. The impact of tax crediting might infer that corporate taxes have less of an impact on reinvestment than personal income taxes do.

    Let’s go back to Okun’s Law.

    The result of that relationship and this paper is we need to curb spending, cut taxes, end regulation and strucutural barriers and fight inflation through monetary and fiscal discipline.

  6. Definitely agree regarding evasion. The first priority of any tax system should be reducing the motivation to evade it. A critical component is ensuring that all citizens feel like they getting value for their taxes. I don’t assume that lowering taxes is always the easiest way to achieve this, however.

  7. actually my point was that the full negative effects of income tax in the analysis didn’t show up because a lot of productive people had already evaded/minimised it.

  8. Yes I gathered that, but surely the study is measuring actual collected income tax as opposed to “income tax that we expected to be able to collect”?

  9. The study extrapolates what growth rates would be by correlating previous actual growth rates with income tax collection. Therefore you would expect the growth correlation is with actual collected income tax.

  10. Don’t single purchasers in a market usually force prices down? When the government (a single purchaser, rather than lots of individuals) buy goods and services, prices SHOULD go down, therefore the GDP goes down, without necessarily decreasing the resources available to the population.

    As I see it, growth in GDP for a nation is like the weight of a person – more is not necessarily better – in fact it can be downright harmful.

  11. Income tax is highly invasive. It entails the government collecting loads of personal private data. In Australia and the USA it involves considerable paper work for nearly every taxpayer. Even if it really was better in terms of growth I’d be hard pressed to reverse my opposition to it. I would also suggest that if governments didn’t have income tax as an option then high taxes would be harder to sell in electoral terms because it would be quite difficult to target significant revenue raising taxes at minorities. So the analysis in terms of political economics would perhaps be quite different than in terms of pure economics.

    GST like Income tax involves considerable paperwork for businesses. However GST entails no paper work at all for the vast bulk of citizens.

    Mark Hill has suggested that income tax, the minimum wage and welfare should be replaced with a higher GST and a basic income. In spite of this paper I am still inclined to agree with Mark in this regard. Although Mark often attaches some other ideas regarding federal/state funding arrangements that I’m not too keen on.

  12. Dave,

    The approximation of a single buyer in the market for teachers and nurses would suggest to me that there is some evidence to support the idea that governments can drive down prices. However I suspect that it does so at the cost of quality. And obviously teachers and nurses and consumers of education and health may not be better off. I would only expect to see a sustained real saving from a single buyer if the market entailed a seller with a monopoly position charging monopoly prices.

    It is an interesting way to frame the discussion however and I think I’m going to have to ponder for a while before I can offer a more complete answer.


  13. David,

    You have confused many different concepts.

    Real GDP and nominal GDP.

    Real GDP is national income deflated by an inflation index. Ideally, you want to maximise GDP as you do real wages at the individual level.


    Monopsony: A single monopoly buyer is a monopsonist. However, public markets rarely are efficient. Furthermore, if Government buys everything, then the sovereign interest rate increases and the cost of capital rises too as private loans are corwded out. No Government has ever been as innovatiive as profit and non for porfit private organisations and GDP would ultimately suffer. As mentioned today on this blog, the Australian Government can only keep the monopsonistic PBS going as long as it tops ups big pharma with production subsidies. So in one industry alone, they pay double with funds that have a minimal cost of cpaital of 19% (the deadweight loss or lost production to society through incentive and price changes caused by taxes attached to each dollar of Government revenue).

    GDP and per capita GDP

    If real, per capita GDP went down, either nominal wages or prices could have fallen. Or both. The result here is falling per capita GDP is bad as everyone is less well off.

    1st world countries have high per capita real and nominal GDP. 3rd world countries have meager real and nominal GDP.

    Increasing GDP (i.e economic growth) is not dangerous. Low per capita GDP can be a very serious risk to your health (nutrition, resource wars, poor healthcare).

  14. Personally, I don’t buy the direct tax v. other revenue sources effect that the paper talks about. I went through the econometrics, and there are a few econometric reasons as to this effect may exist…I go through them a bit in my site (shameless plug). On a basic level, they boil down to either coefficient misinterpretation (viewing the coefficients as absolute instead of relative) or misspecification leading to the result being correlative, not causal.

  15. If you are going for the Henry George idea, it’s not actually a tax on land values, but rather a flat rate per unit of unimproved land (e.g. a per acre tax).

  16. If you are going to have a tax on land value I think it should be on the purchase value not the current market value. And if you are going to have such a tax on a recurring basis then best to abolish the one off stamp duty at the time of purchase.

  17. The Henry George idea is NOT a per-acre tax. It’s a tax collecting almost all of the RENTAL value of the site (the “almost” being necessary to allow some margin for valuation error).

    The modern “tax on site-only land values” is calculated on the capitalized (lump-sum) value of the land. So it’s not quite the Henry George idea.

  18. The Henry George idea is NOT a per-acre tax. It’s a tax collecting almost all of the RENTAL value of the site (the “almost” being necessary to allow some margin for valuation error).

    The modern “tax on site-only land values” is calculated on the capitalized (lump-sum) value of the land. So it’s not quite the Henry George idea.

  19. Terje, for sure, I think that Land Value Tax (in the UK) should replace ALL property taxes;
    Council Tax (local rates);
    Stamp Duty Land Tax (on purchases – up to 4%, ouch!);
    Inheritance Tax (up to 40% of deceased’s estate – ouch!);
    Capital Gains Tax (a tax on transactions – see below);
    The TV licence fee (a Poll Tax of £130 per household – ouch!).

    However, LVT is about encouraging efficient use of land; so the tax has to be levied on CURRENT market values, or else people would be discouraged from moving home after prices had gone up.

    For sure, people whine about ‘ability to pay’, heck, I also believe in redistribution via Citizen’s Income (about which I emailed you a while back, much cooler than 30/30) and of course pensioners would be allowed to roll up unpaid tax to be repaid on death (which is another reason for scrapping Inheritance Tax – else there’d be a double whammy on death).

    An LVT set as a % of current capital market values would act like an extra interest rate on land values, so would prevent these ludicrous land price bubbles that we have in the UK (and in Australia, AFAIAA).

  20. Mark

    Welcome 🙂

    But wouldn’t a land value tax also heavily penalise those cash-poor members of society who have spent a lifetime paying income and consumption taxes such as pensioners? would you grant just retirees an exemption or those over a certain age?

  21. Terje, such a tax will very much dampen price rises, it acts like a higher interest rate. But interest rate hikes (under current system) only hit people with an existing mortgage (for whom it is too late – they’ve already locked in the purchase price and the debt). It does not affect those about to buy because the price falls to compensate, and it does not really affect people who have paid off the mortgage and who have no intention of selling. So a LVT on everybody is much fairer between potential first time buyers, mortgage payers and outright owners. D’you remember that thing, fairness?

    And where does this ‘higher taxes’ come from? If LVT receipts increase, then use it to cut GST or Income Tax!

    LVT is also a bit like rent. So people who own vacant properties (up to one million in the UK) and/or ‘overoccupy’ (pensioners in three bed family homes) will be much more likely to sell the vacant property or trade down into a smaller one, rather than pay for space they don’t need. So existing housing is used more efficiently rather than building loads more which NIMBYs have more or less prevented over here.

    Pommy, yeah, yeah, the pensioners argument. Henry George and Winston Churchill referred to this as a ‘hoary old chestnut’ over a century ago. No exemption for pensioners – they can have a roll up. Or else the pensioner would bank the exemption and then let his family live there LVT-free, there’d be all sorts of fraud going on. Like with protected tenancies in the UK (long story) or with privately sub-let Council Housing.

  22. Mark,

    I think we got our wires a little crossed. When I said “making existing owners pay higher taxes” I was not refering to higher overall taxes but land taxes that are higher after property prices have increased. Owners have no price setting role, price setting being done by buyers and sellers. So how does including existing owners in tax hikes driven by a shift in the market value of property work to slow the formation of speculative price bubbles. In terms of taxing properties based on purchase price as opposed to current value I don’t see that the latter offers much advantage in this regard.

    Interest rates targeting is not a policy I advocate. I advocate floating interest rates not interest rates targeted by monetary policy. Interest rates should be determined by the supply of and demand for credit and we should not have a central bank in the business of supplying credit or using this power to manipulate interest rates. If your concern relates to controlling inflation then I don’t believe that taxation has any real role to play as a policy instrument. If anything the government should cut taxes on goods (including land) if there is a welfare issue associated with the price of that good rising. Promoting land taxes as an alternative means to raise revenue has a lot of merit, maybe even in terms of fairness, however promoting it as a tool of monetary policy is simply misguided.


  23. Terje, I could answer your questions one by one, but really you have to sit down and think hard and honestly about what would happen if we had LVT, the questions will answer themselves.

    I know all the hidebound objections to LVT. On a personal level, I have made a small fortune over the last ten years by buying at the bottom of the market and bailing out at the top (I now rent), so I (as an old and wealthy person) am against LVT on a purely personal level – I have every intention of making a large fortune in the next upturn from about 2010 to 2020, an upturn that will only happen provided politicians are too short-sighted to spot all the advantages of LVT.

    But from an economist’s point of view, and being brutally honest about it, rather than pandering to land owners and home owners who want to bank all their unearned profits tax free (and to hell with the first time buyers and screw the economy – these booms and busts are largely fuelled by land price/credit bubbles as in now being played out across the world), LVT is by far the least-worst tax, and unlike most taxes, has a lot of beneficial effects.

  24. Terje:
    The best example of what is possible with a single purchaser (but, I admit, not always realized, and useful only for need-to-haves, not nice-to-haves) is a nationalized pharmaceutical benefits scheme: regulated quality, and I’m pretty sure the statistics are available to look at the impact on the GDP and even adjust for different consumption patterns between countries.

    I’d imagine a libertarian perspective on this could be to consider the following question: What is the least burdensome way for a population to ensure the freedoms of speech and movement (as well as the ability to contribute and self-actualize) to the significant number of people with such freedoms constrained by [insert name of any disease – e.g. Parkinson’s – that has massive and long-term impact]?

    Mark –
    You said “Real GDP is national income deflated by an inflation index. Ideally, you want to maximise GDP as you do real wages at the individual level.”
    Actually, GDP is not necessarily a measure of human well-being. If people grow their own fruit’n’veg in their back yard, with any seasonal surplus given to friends and neighbors, the GDP is hit not only by the notional dollar value of that food, but by the dollar value of commercially-provided transport and storage – although the individuals concerned are probably better off. Another example might be the economic activity from legal actions to get freedom of speech in a litigious society compared to the lack of such economic activity in a society where people don’t try and constrain the speech of others – even if the laws are exactly the same.

    I’d also comment that Australia’s growth in GDP with explosion of the national debt is like having a pie that gets wider while decreasing in volume. IMHO, economic health, not just human well-being, is not highly correlated with “growth” as measured by GDP.

    On your comment about big-pharma being subsidized in Oz, it’s worth noting that US big-pharma gets a lot of for-free research (often with commercial-in-confidence results) done in government-sponsored programs. Administration of patents, and separating research from manufacturing when considering these issues is worthwhile. The other problem is that too often big pharma gets big money from PBS schemes with a new drug that has benefit calculated against NO treatment rather than treatment with a much cheaper analog: i.e. a 2% increase in efficacy with a 200% increase in cost.

  25. Dave – the flaws in GDP are obvious.

    None if these (proxy for well being, non counting of leusire or non traded goods) support your idea that “GDP is dangerous like being overweight”. As for the vegetables, then there are savings and income is generated elsewhere.

    National debt is not and will never be a problem. Why is it a problem? It is no different to Australians owing Australians. 20% of Australian capital stock is due to foreign direct investment – as are about 25% of wages. Reducing the wage bill (employees and wages) by 25% doesn’t make anyone better off.

    Growth is bad? How? Growth allows for technologies to emerge. Are you saying we sholdn’t have any technological progress?

    It’s strange you support PBS while acknowledging the amount of subsidies which are required to keep pharmaceuticals selling and producing in Australia after the single purchaser prices are paid.

  26. chal, I think you are misinformed or have failed to comprehend the literature.

    Most TFP measurements contradict your statement. However, TFP is not the entire picture. The burden of taxation and supply side effects on labour supply and capital investment are probably mopre important as the investment is needed to raise TFP in most cases.

    Have a read:

    M Feldstein – National Tax Journal, 1997 –

    Click to access v50n2197.pdf


    Journal of Policy Modeling
    Volume 24, Issues 7-8, November 2002, Pages 679-692

    Government size, factor accumulation, and economic growth: evidence from OECD countries

    Atul A. DarCorresponding Author Contact Information, E-mail The Corresponding Author and Sal AmirKhalkhali

    Department of Economics, Saint Mary’s University, Halifax, NS, Canada B3H 3C3

    Revised 11 December 2001;
    accepted 1 May 2002.
    Available online 19 September 2002.


    This study examines the role of government size in explaining the differences in economic growth rates of the 19 Organization for Economic Co-operation and Development (OECD) countries over the 1971–1999 period using a random coefficients model. Our results indicate that, on average, total factor productivity growth, as well as the productivity of capital, are weaker in countries where government size is larger. The advantage of a small government sector, in general, likely reflects the greater efficiencies resulting from fewer policy-induced distortions (such as the burden of taxation), the greater discipline of market forces which fosters efficiency of resource use, and the absence of crowding-out effects that weaken the incentives to create new capital which embodies new technologies. From a policy perspective, this does not mean that the optimal policy is one that minimizes the size of government. Rather, a small as opposed to a large government could potentially be as effective in providing the legal, administrative, and governance infrastructure critical for growth, as well as for offsetting market failures. At the same time, the country-specific results indicate that the nature of country-specific institutions as well as the mix of government activities are as important for growth performance as the aggregate size of government.”

    So yes it is not entirely because of the size of Government, but a mix of the size and efficiency. The objective is then to design efficient, small Government policies.

  27. Mark,

    As a former public servant I can assure you that public servants and politicians in the main will agree with you then do the exact opposite. It is against their interests to have small government so naturally, and perversely in light of Adam Smith’s ideas about the benefits of selfish behavior, they will strive for larger government.

Comments are closed.