The Chancellor wants to drive his tax hikes through the cabinet. But, explains David B Smith, people will pay a high price in the decades to come. Raising taxes can reduce government borrowing in the long run, but only at the cost of substantial collateral damage to private sector activity and employment.
Introduction
The Chancellor, Rishi Sunak, and the cabinet of Boris Johnson, now appear committed to introducing some of the largest tax increases in Britain’s economic history. The rate of Corporation Tax is scheduled to rise from 19 per cent to 25 per cent in April 2023 and both Employers’ and Employees’ National Insurance contributions (NICs) are to go up by 1.25 per cent this April. These tax hikes are occurring when the share of general government expenditure in the factor cost measure of UK Gross Domestic Product has been at the historically high average of 49.8 per cent in the four quarters to 2021 Q4 (Chart 1).
Chart 1: Ratio of UK General Government Expenditure to UK GDP at Factor Cost 1956 Q1 to 2021 Q4 (Smoothed Four Quarter Averages – Per Cent)
The corresponding picture for the UK non-oil tax burden (i.e., excluding oil taxes) appears in Chart 2. There are marked seasonal swings in the spending and tax ratios. The use of a four-quarter trailing average to smooth out this seasonality in Charts 1 and 2 means that the tax ratio of 37.3 per cent in 2021 Q4 is the average of the period 2021 Q1 to 2021 Q4. It is therefore still affected by some of the temporary tax ameliorations introduced to ease the consequences of the Covid-19 lockdown. The all-time quarterly peak on this measure (39.9 per cent) occurred in the four quarters to 1970 Q2 following the 1969 International Monetary Fund (IMF) bailout of the UK economy, while the low point was the 30.2 per cent recorded in 1960 Q4 and 1961 Q1 during the “never had it so good” era of Harold MacMillan.
Chart 2: Ratio of UK Non-Oil Taxes to Non-Oil GDP at Factor Cost 1956 Q1 to 2021 Q4 (Smoothed Four Quarter Averages – Per Cent)
The Office for Budget Responsibility (OBR) forecasts that accompanied last October’s Budget showed the ratio of total taxes to total factor-cost GDP rising from 33.4% in 2021/22 to 35% in 2022/23 and 35.7% in 2023/24 before drifting up from 36.0% in 2024/25 to 36.2% in 2026/27[1]. These numbers, if they can be achieved, are probably close to, but within, the sustainable upper limit of taxable capacity. However, this does not mean that hiking taxes on this scale does not imply serious adverse implications for the real economy.
The Laffer Curve and All That
It has been known in the public finance literature since the eighteenth and nineteenth centuries that increasing any given tax rate of tax leads to a weakening of the tax base of private sector transactions and that eventually higher tax rates lead to a collapse in the tax base, growing evasion, and ultimately falling revenues for the government. There was also discussion well over a century ago as whether this was also true of the aggregate tax burden. However, government spending was only around one tenth of national output in 1900, for example, so it was not really a live issue. The so-called ‘Laffer curve’ was certainly not invented by Professor Laffer, although he did a brilliant job of publicising the concept in the early 1980s.
It should also be emphasised that, in theory, there are three distinct stages of ‘Lafferness’. This important distinction is almost entirely ignored in the political debate.
First, there is the situation where the tax base is invariant to the rate of tax (or unchanged), so doubling the tax rate simply doubles revenues, for example. This is complete nonsense because it assumes economic agents are too stupid to change their behaviour – even in the slightest degree – in response to even gross alterations in the economic stimuli facing them[1]. It is equivalent to assuming that the tax authority is in the position of a monopolist facing a totally inelastic demand curve, which is a purely illustrative textbook concept. Unfortunately, this also seems to be the pervasive view among the political and bureaucratic classes, and alleged tax ‘experts’, whose entirely static calculations are often grossly misleading as a result.
Second, there is the quasi-Laffer situation in which hikes in the tax burden worsen the output/inflation trade off and lead to some destruction of home demand in the short term and aggregate supply in the longer term. This situation was widely found by people running traditional macroeconomic forecasting models in the past[2]. It was also not uncommon to find that the adverse effects of tax hikes build up over time because of the induced loss of international competitiveness and the disincentive effects on investment and the willingness to seek and offer employment.
Finally, there is the full Laffer curve situation in which cutting the rate of tax leads to increased revenues for the state, either because of reduced evasion or because of the stimulus provided to private activity. The concept of the micro-Laffer curve is so well established that it is almost a truism with the only uncertainty being the precise point at which it kicks in. Where there is more debate is the extent to which there is a powerful macro-Laffer curve applying at the level of the entire economy. It is the latter claim that gave rise to the sneer of ‘Voodoo Economics’ on the part of Big Government interventionists, including many self-proclaimed Conservatives.
The Bill – What Will it Cost?
The main purpose of this note is to report some counter-factual simulations for which I have used my Beacon Economic Forecasting (BEF) model of the international and UK economies. The ‘High Tax’ Scenario incorporated the increases in Corporation Tax and NICs that have already been announced and represents what modellers tend to call the ‘Base Run’. This might also be considered as the ‘most likely’ forecast, given the state of knowledge when the simulations were performed on 27th January. This contrasts with the counterfactual ‘No Tax Hike’ run, which held the rates of Corporation Tax and NICs at their previous levels – i.e., it simply assumes that Mr Sunak’s tax rises are never implemented. It was then possible to compare the two runs to estimate the ‘butcher’s bill’ for the Johnson/Sunak tax hikes.
It must be emphasised that all such exercises are heavily dependent on the properties of the model being employed, are highly uncertain and cannot cope with external shocks such as Covid-19 or a war breaking out in Eastern Europe.
The advantage of using a macroeconomic model is that it should allow, at least to some extent, both for the second-round feedbacks involved and the dynamic effects over time. However, such simulations also generate masses of quarterly data for several hundred variables. So, it is sensible only to deal with the key variables here. The BEF model is normally run for ten years ahead with the result that the final calendar year for this exercise was 2032.
The next paragraph reports the key differences between the Base Run and the No Tax Hike scenario in 2032 for a subset of the most important economic variables. This does not mean that the interim dynamics are not interesting; merely that space limitations preclude further detail and ten years seems a reasonable, if arbitrary, time horizon to allow the long run consequences of Mr Sunak’s policies to work through. It would also be possible to report results for each tax change individually. In the event, because the Corporation Tax hike was announced first, this change was included in our forecasts for some time before the increases in NICs was announced. However, going into such detail would probably only obscure the main picture.
Moving on to the results, there was indeed a reduction in projected Public Sector Net Borrowing (PSNB) of some 0.4 percentage points of national Gross Value Added (GVA) by 2032/33 because of Mr Sunak’s tax increases. However, this was at the cost of a 0.8% hike in the Labour Force Survey (LFS) measure of the unemployment rate by 2032, and a 299,000 increase in the claimant unemployment count over the same period. Furthermore, there was a 1.3% reduction in real GDP following on from these tax increases by 2032, and a 4.1% drop in real private domestic expenditure (PDE).
Two reasons why the projected loss in the volume of PDE – defined as non-welfare financed consumption plus private investment and stock building – was so much greater than that in GDP in the High Tax scenario are that: 1) government spending is a component of GDP and was assumed to be the same in volume terms in both simulations, and 2) because of the high marginal propensity to import out of private domestic expenditure. There is an important message here that, if one is primarily concerned with the wellbeing of the private sector activity that forms the tax base[3], GDP can be a very poor measure of the pressures being imposed upon it[4]. Three further adverse consequences of Mr Sunak’s tax hikes were that the sterling index was reduced by 1.8% in 2032, there was a 0.5% increase in the Consumer Price Index (CPI) over the same period, and a 0.7% of GDP deterioration in the balance of payments. Finally, both Bank Rate and the twenty-year gilt yield ended up 0.1 percentage points lower by 2032 in the No-Tax-Hike scenario.
Implications
I would be the first to admit that all such model simulations are massively imprecise and that it would be extremely useful to see the results of similar analyses carried out on other forecasting models. Unfortunately, neither the Bank of England nor OBR models seem to be technically capable of carrying out the sort of exercise carried out here, despite the long history of such analyses. This is because the Bank of England model does not appear to incorporate a detailed representation of the government accounts while the OBR sets long-run growth and CPI inflation by assumption, not endogenously from within its forecasting framework [5].
What seems to emerge from our research in terms of the Laffer-curve criteria above is that the aggregate UK economy is currently in the second ‘quasi-Laffer’ situation[6]. Raising taxes can still reduce government borrowing in the long run but at the cost of substantial collateral damage to private sector activity and employment. Thus, the reduction in real GDP in percentage points was 3.25 times the percentage point reduction in the PSNB as a share of GDP (i.e., 1.3/0.4), for example, while the corresponding ratio with respect to real PDE was over 10 (4.1/0.4) and the rise in the LFS unemployment rate was twice the reduction in public borrowing (0.8/0.4).
The questions that arise from such malign trade-offs are as follows. First, is the government morally justified in imposing so much pain on the private sector citizenry or has it become a purely predatory institution acting in its own selfish interests, notably those of the political and bureaucratic classes? Second, at what stage does the populace lose patience with being ‘ripped off’ by the provision of frequently poor public services at a high tax cost? Third, how much economic pain can the private sector endure before economic growth goes permanently negative, the state becomes unaffordable, and we end up with a Venezuela or Zimbabwe style economy? Finally, what is the point of the continued existence of a Conservative administration or the Conservative Party itself given its implementation of a hyper Big Government approach? It is not difficult to see why some Cabinet Ministers are arguing that the already announced hike in NICs should be cancelled or postponed. This is probably not because of any altruistic concern about the damage that their spend, borrow and tax policies are doing to the economy and the electorate but a naked panic about their personal careers once the tax-hiking chickens they let fly come home to roost. If the Prime Minister wants, as he says, to change course, there may be no better place to start than taking the shears to Treasury’s tax and spend plans.
References
King M (2021), Monetary Policy in a World of Radical Uncertainty, Institute of International Monetary Research Annual Public Lecture, London 23rd November 2021.
[1] A similar fault was committed by those epidemiologists who grossly over-predicted the death rate from Covid-19 because their predictive models did not allow for the spontaneous changes in behaviour that occurred as the risk of infection rose.
[2] Such simulations also indicated what were the most and least damaging tax increases to implement. There was a noticeable consensus that raising employers’ NICs was significantly the most damaging and the one tax increase that was unambiguously on the wrong side of the Laffer curve.
[3] In economic terms, it is logically impossible for government to generate real resources by taxing itself since any tax transactions within the government sector can all be netted out, even if this is not the way the official statistics are published.
[4] This distinction is also important for the monetary authorities who have zero control over the government sector because the fiscal authorities can always print money at will. Only the private sector has a demand for money and credit and the entire burden of monetary stabilisation policies falls on the roughly one half of the economy that is not socialised.
[5] Part of the problem is the fashion for Dynamic General Stochastic Equilibrium (DGSE) models in officialdom. The former Bank of England Governor, Lord King, has been rightly critical of such models in the monetary policy area. But the same criticism applies to their growth forecasts as well (see: King (2021)).
[6] However, this does not mean that the UK cannot topple over into Stage Three of the Laffer curve. The higher the tax and spending ratios are at the starting point, the more likely it is that there will be a transition into stage three. This should be a strong concern given where the UK is at present.
