Welsh taxes

Ple a s e no t e: Ch a n g e s m a d e a s a r e s ul t of p u blishing p roc e s s e s s uc h a s copy-e di ting, for m a t ting a n d p a g e n u m b e r s m ay no t b e r eflec t e d in t his ve r sion. For t h e d efini tive ve r sion of t his p u blica tion, ple a s e r ef e r to t h e p u blish e d sou rc e. You a r e a dvise d to cons ul t t h e p u blish e r’s ve r sion if you wish to ci t e t his p a p er.

Thi s v e r sio n is b ei n g m a d e a v ail a bl e in a c c o r d a n c e wit h p u blis h e r p olici e s. S e e h t t p://o r c a . cf. a c. u k/ p olici e s. h t ml fo r u s a g e p olici e s. Co py ri g h t a n d m o r al ri g h t s fo r p u blic a tio n s m a d e a v ail a bl e in ORCA a r e r e t ai n e d by t h e c o py ri g h t h ol d e r s .

Introduction
Fiscal devolution ensured that by April 2019 the Welsh Government and local authorities controlled nearly £5 billion of tax revenues, corresponding to 30% of their combined current spending. The taxes devolved are Council Tax and Non-Domestic (Business) Rates, Income Tax (partially), the Land Transaction Tax and Landfill Disposals Tax (replacing Stamp Duty Land Tax and Landfill Tax). This last pair of taxes are collected and managed by the newly formed Welsh Revenue Authority. In addition, a shortlist of four new tax ideas was announced alongside the outline draft Budget 2018-19a vacant land tax, a social care levy, a disposable plastics tax and a tourism tax. All the existing taxes are not only necessary to cover expenditure in the current fiscal year but also to cope with future trends. As warned by the Wales Centre for Public Policy, the current bases that are taxed are likely to constrain future spendingdue to low employment rates, wages and population growth (Ifan and Poole 2018). This paper examines some of the possibilities for Welsh tax changes and improvement.

Tax fairness and appropriateness
Taxes are rarely popular, and they mostly change people's behaviour counterproductively (Brewer et al, 2010). A good tax system should therefore aim to minimise adverse side effects while achieving revenue, distributional and other objectives. Some taxes might change taxpayers' behaviour helpfully, as would be the intention of a plastic bottle tax. Keeping tax administration and compliance costs as low as possible must be desirable. Taxes should be fair in the sense of avoiding unwarranted discrimination and satisfying taxpayers reasonable expectations. It is also preferable that people can and do understand how their taxes are assessed.
To prepare for the exercise of their tax powers the Welsh Government initiated a survey, between 2016 and 2018, to assess the extent to which Welsh people understood which taxes were controlled by the Welsh Government (Owens and Jones, 2019). For the purposes of considering tax reform it would be helpful also to understand attitudes to various forms of taxation. For example, the Lyons Review when considering the Council Tax reported a survey of English attitudes (Lyons, 2007, p.226). The survey found that 55% thought it unfair that they should pay more council tax if their house increased in value more than others, against 25% thinking it fair. The same proportion (55%) felt basing council tax on house values was unfair but about 30% thought it fair 1 . One explanation for the majority perception of unfairness may be simply that most taxes people pay are administered by firms; council tax is one of the few that is not and therefore hurts more. Or perhaps taxpayers believe they should not pay more when only the exchange value of their property has increased, because this confers no immediate advantage on them; the use value of their property has not been enhanced. There is also the matter of considering taxation over the lifecycle. A tax on property could be seen as a tax on cumulated life savings by asset rich and income poor pensioners.
For some economists, taxing housing, and more particularly the land on which it is built, is desirable, as is taxing the land on which business operates. Economists suggest that taxes on land, and on anything else in perfectly inelastic supply, does not trigger adverse behaviours. Ultimately an increase or decrease in business rates simply raises or lowers the rent landlords can charge. It does not affect the amount of land supplied. And it takes away from people who do not contribute. The ability to pay rent stems from productive efforts of businessmen or of residents, not of landlords. An implication is that dedicated student rental housing is more profitable for landlords than other forms of residence because students do not pay council taxthough they may also have less ability than the average occupant to pay rent. Economic theory indicates that what matters is that council tax and non-domestic rates are payable not only on land but on the structures and developments on it. These taxes are a disincentive to invest in improvement, though some maintain that housing is no different from other consumption goods and so housing services should be taxed at the VAT rate (Mirrlees 2011, et al. ch. 16) 2 . As the Lyons survey suggests, taxing housing services at 20% is not likely to be popular with tenants and householders.
Another, more easily remedied, objection to council tax is that it is regressive; the proportion of income spent on the tax typically reduces with the taxpayers' income. This could be addressed by introducing more valuation bands. Valuation of business properties is more problematic because the markets are more specialised and thinner. Valuation officers currently apply several rules of thumb that give rise to a large number of appeals. A review in England concluded nonetheless that this method was preferable to self-assessment (H M Government 2018).
The tax assessment problem points to an advantage of the capital gains tax; it taxes a definite undisputable market figure for the increase in the business property or second home value. The potential drawback for the tax revenue authority is that the revenue is only paid when the gains are realised, in contrast to the annual payments of the nondomestic rates and council tax.

Welsh income tax
As noted, a merit of the two taxes discussed above, insofar as they tax land, is that the tax base will not move or change if the tax rate goes up. In principle this is not the case for labour and capital, which matters for Wales especially if its income tax rates diverge from England's. The predicted effect on tax revenue of an alteration in the income tax rate may differ from the 'mechanical effect' (multiplying the change in tax rate by the incomes of the taxpayers). Those taxed may vary their effort or innovativeness, they may alter where they live and the consequences of all these decisions may have impacts on the economy that further affect tax revenue. Such migration and economy 'spillover' effects of possible income tax changes were investigated in a recent study (Foreman-Peck and Zhou, 2019).
Whereas the Scottish government reduced the lower Scottish tax rate and increased the higher rates in 2018, the Welsh government may be expected to be more restrained in making similar income tax changes. The economy in Wales is more closely integrated with England than the Scottish economy. Around 95,000 people commuted out of Wales and almost 48,000 commuted in during 2018 (StatsWales 2019a). In 2017-18 net migration to Wales from the rest of the UK was just under 9,000, with net outflows for persons aged 15-29 (Stats Wales 2019b). Over 1.4 million people in Wales (48% of the total) live within 25 miles of the border with England, and more than 4 million live within the same distance on the English side of the boundary (Figure 1). By contrast only about 200,000 people in Scotland and 260,000 in England are within 25 miles of the border between these countries. An income tax difference from England is therefore more likely to trigger substantial migration between England and Wales than between England and Scotland. Source: Holtham Report 2010, Chart 4.5 Whereas the Scottish government reduced the lower Scottish tax rate and increased the higher rates in 2018, the Welsh government may be expected to be more restrained in making similar income tax changes. The economy in Wales is more closely integrated with England than the Scottish economy. Around 95,000 people commuted out of Wales and almost 48,000 commuted in during 2018 (StatsWales 2019a). In 2017-18 net migration to Wales from the rest of the UK was just under 9,000, with net outflows for persons aged 15-29 (StatsWales 2019b). Over 1.4 million people in Wales (48% of the total) live within 25 miles of the border with England, and more than 4 million live within the same distance on the English side of the boundary (Figure 1).
By contrast only about 200,000 people in Scotland and 260,000 in England are within 25 miles of the border between these countries. An income tax difference from England is therefore more likely to trigger substantial migration between England and Wales than between England and Scotland.
An illustration of the possible migration effects of higher rate income tax changes is that a five percent increase in the higher rate of tax for an individual or household earning £80,000 would result in almost £2,000 a year more being paid in tax by the individual or household. Over ten years with a five percent discount rate this equates to about £15,000, which might more than counterbalance relocation costs. In which case, all the tax

Figure 1: The differences between the Welsh border and Scottish border
revenue paid by that individual or household would be lost to Wales. Tax rate increases may also provide a disincentive to relocations into the jurisdiction that might otherwise have taken place.
For higher earners the revenue impact of taxinduced mobility would be greater. If only 150 of Wales's highest earners moved out of Wales (10% of the total number earning over £250,000 per year) no additional revenue would be raised from Welsh residents as a result of a one pence higher rate tax increase (Holtham, 2010, para 6.22). Conversely cutting the higher rate could potentially increase Welsh tax revenue, although there could be problems of determining residential location and second homes.
The difficulty is that with no experience of differential tax rates there is no direct evidence of how much migration might be induced if income tax rates did diverge. A naïve solution is to use another country's experience to shed light on the UK scenario (such as Liebig and Souza-Poza, 2006, on Switzerland), but because of different institutional, geographical and economic conditions this empirical approach is not reliable. At the other extreme, purely theoretical models can be built and utilized, as in HMRC (2013). Nevertheless, such a model is not capable of capturing the tax effect on within-UK migration flows. Any judgement on this matter without a strict link with data is inevitably arbitrary. In view of the drawbacks of both empirical and theoretical modelling approaches, this paper uses a systematic framework consistently integrating both a structural economic model and an econometric model, utilising the only substantial source of tax differences between areas in England and Wales, the Council tax (for more details see Foreman-Peck and Zhou, 2019).
The effects of council tax differences (observable) and the effects of income tax differences (unobservable) can be equivalent if the magnitudes are carefully selected. Hence, the effect of an income tax difference can be simulated by translating it into a council tax difference. The two taxes affect taxpayers' budgets differentlycouncil tax raises the price of housing relative to other goods and services whereas income tax does not. In Figure 2, with housing plotted on the vertical axis and other consumption on the horizontal, the household budget constraint is indicated by the black solid line joining both axes. A cut in the housing (council) tax pivots the (now red) budget line because more housing can be consumed after the tax cut. The tangency of the budget line with the curved utility or wellbeing lines indicates the optimum consumption choicesfirst ℎ and , then ℎ′ and ′. The income tax equivalent of a council tax difference (the green line) -corrects for the different budget gradients while maintaining the same level of (subjective) well-being, here reducing housing consumption to ℎ′′ and increasing general consumption to ′′.

Figure 2: The Income tax equivalence of Council Tax
A simple Computable General Equilibrium (CGE) model of the Welsh economy is constructed and calibrated to simulate these two equilibria. Three types of individuals are specified to correspond with those paying the three rates of incomes (Basic, Higher and Additional). When the lowest rate of income tax changes (Basic Rate of income tax, 20%) all three types of individuals are affected, when the 40% Higher Rate changes those paying the Additional Rate of 45% are also affected. Council tax is more selective; a specific property tax only affects one type of taxpayer. Property (council) tax rates for each of the three types of individuals thus must be utilised to estimate the equivalent income tax. Supplementing the CGE model is an econometric model of migration because of the lack of prior information about tax-induced migration. The CGE model predicts the tax revenue consequences of hypothetical income tax changes. It does so by making appropriate adjustments as indicated above because current (here, property) taxes have different bases from the counter-factual income tax. When the econometrically estimated migration response is added in, the CGE model computes the full economic effect (Figure 3).

Figure 3: Workflow of the analysis
In the econometric model the variable to be explained is the bilateral flows of migration between 348 local authorities in England and Wales. The results suggest that between about half the local authority pairs there was no migration. Wage levels are assumed to be fixed by migration and trade, as Wales is a small open economy and in order to keep the exercise as simple as possible. This allows local policies, such as tax, to determine the long run equilibrium population and the migration necessary to achieve it. Migration determines output and, by changing tax bases, influences tax yields.
The 'mechanical effect' dominates the migration effect for the Welsh government when the Basic Rate (BR) of income tax is increased. If BR rises by 5p, income tax generated in Wales will rise by 8.02%, about 2 percentage points less than the full mechanical effect. Income tax kept by Wales will rise by 30% but total tax revenue generated in Wales drops by 1.9%. For tax revenue as a whole the migration effect dominates because the change in population also affects all the other tax revenues paid in Wales.
For Higher Rate (HR) changes, there is a similar asymmetric pattern between income tax generated in, and kept by, Wales for similar reasons. Behavioural responses exactly offset mechanical effects when the Additional Rate (AR) is altered, so there is very little effect on tax revenue.
Output per worker responds to the opening of tax differentials between England and Wales and the induced migration by changing the proportions of taxpayer types in the Welsh Notes: (1) answers the question "What is the relationship between the effects of different council taxes and different income taxes?" and (2) answers the question "What are the effects of different council taxes on the migration flow into Wales?" Combine (1) and (2) to answer the main research question "What is the overall effect of changes in income tax on the overall economy?" population. These types have different productivities and so average productivity of the economy responds to the type of tax changes. A Basic Rate increase affects all taxpayers and the proportional effect is greater on (higher productivity) HR payers than BR payers, reducing output per worker.
Total tax revenue responds in a similar fashion to output per worker. A BR or HR tax rise loses the UK government tax revenue while the Welsh government's revenue increases. In setting the HR and BR of devolved income tax the interests of the Welsh and UK governments are opposed. With such a small proportion of revenue devolved, tax-induced migration reduces Welsh government revenue by quite a small proportion. But the spillovers to the wider economy and the impact on other Welsh tax revenues collected by the UK government are greater.

Conclusion
If Welsh taxes were entirely devolved, the divergence of interest between UK and Welsh governments in income tax rate setting might be resolved. But the cautions about the prospects for even the limited Welsh tax base suggest this would be problematic for Wales.
Another difficulty with changing Welsh income tax rates is that beneficial effects may take some time to emerge and depend on whether the tax change is expected to be long lasting. So, a cut in income tax may eventually increase revenue when the behavioural responses offset the mechanical effect. But meanwhile the government's budget faces a shortfall that must be financed.
In the search for tax revenue, changing income tax rates seems perilous. If there are other policy objectives to be achieved by reducing them, it is advisable to seek a broader tax base. ap Gwilym (2017) has suggested a land value tax should replace income tax entirely and as noted, the Welsh Government is considering the practicalities (Bangor n.d.). Such a tax may appeal to economists on the theoretical grounds discussed earlier, though a shift of the proposed magnitude could present major administrative challenges. A less radical approach would be to make Council Tax (which includes an implicit tax on land) more progressive by introducing some higher bands, though the revenue yield would also be less exciting.