Cost of living crisis – Removal of 5% Goods and Services Tax (GST) from food (excluding alcoholic drinks) (P.100/2022): comments

STATES OF JERSEY

COST OF LIVING CRISIS –  REMOVAL OF 5% GOODS AND SERVICES TAX (GST) FROM FOOD  (EXCLUDING ALCOHOLIC DRINKS) (P.100/2022): COMMENTS

Presented to the States on 17th November 2022 by the Council of Ministers

STATES GREFFE

2022  P.100 Com.

COMMENTS

The Council of Ministers, from its very first days in office, has focused on helping people with cost-of-living increases. The Mini-Budget put almost £15 million back into Islanders' pockets this year and almost £42 million in 2023 - including changes to income  tax  allowances,  Income  Support,  Social  Security  contributions  and  the Community Cost Bonus. We have also included increases in the minimum wage and support with school meals and sanitary products.

This is part of an ongoing commitment to help people - targeting support to those who need it most, while ensuring we remain a responsible jurisdiction who manage the public's money carefully.

With this in mind, the Council does not support zero-rating food for the Goods and Services Tax (GST). While it appears attractive, it is not the best measure to help people, on the contrary, it is ineffective, complex, and burdensome.

The Council believes that GST should remain low (at 5%), broad (avoiding zero-rates and exemptions), so being fundamentally simple (and so low-cost) to operate both for businesses and Government.

This Proposition has been described as a response to the cost of living crisis. However, due to the necessary implementation lead times, Islanders would not feel the benefit of zero-rating, if passed on by retailers, until 2024. The recent Mini-Budget is currently delivering support to Islanders to weather the cost of living crisis, with further elements of the package of measures generating benefits in 2023.

It is also important to recognise that when GST was introduced, the States Assembly provided enduring compensation to Islanders for the impact on food prices through three mechanisms:

• Improved tax allowances, benefiting all taxpayers at the time of the introduction of GST;

• Improved  Income  Support  payments,  providing  uplifts  to  Income  Support components in line with the introduction of the 3%, and subsequent increase to 5%, standard rate of GST; and

• The creation of the GST Food Bonus, fully compensating households who did not pay tax and did not receive Income Support against the cost of GST on food.

These compensatory mechanisms remain embedded within their respective systems which have also seen additional degrees of uprating over time to recognise the impact of inflation. A move to remove GST on food would raise questions as to the future direction of these compensatory measures.

In fact, the recent Mini-Budget has provider additional support in all three areas in a targeted and timely manner:

• Income tax allowances will increase by 12% in January 2023, with taxpayers feeling the benefit immediately through updated ITIS rates;

• Income Support rates will see an increase of at least 10.4% year-on-year from January 2023; and

• The Community Costs Bonus (successor to GST Bonus) has been doubled in value and extended to a wider group of households for 2022.

In contrast, the removal of GST on food would create a disproportionate benefit to higher income households. With an estimated cost of £10m per annum to government revenue, analysis suggests that just £1.2m would benefit the lowest-earning 20% of households while around £2.9m would go to the top-earning 20% of households.

Given the current volatility in global markets - and the additional compliance costs associated with zero-rating - it is unlikely that the Proposition, if successful, would materially reduce the retail price of foodstuffs.

Existing support mechanisms provide direct benefits to Islanders, the value of which are not affected by prevailing market conditions as zero-rating food would be. In any case, it remains the case that there are wide variations in the prices of basic foodstuffs in Jersey's retail market.

Zero-rating  would  introduce  administrative  complications  for  businesses  and Government under any feasible option for delivering it. Depending on how the reduction was delivered, set-up and ongoing administration costs would fall on Customs; Revenue Jersey; and businesses, with the burden falling disproportionately on smaller firms.

Comments

Jersey's Tax Policy Principles

1. Since GST was introduced in 2008, the States Assembly has supported the principle that GST should be low (currently 5%) and broad (limited exemptions and zero-rates) to keep accounting for the tax simple, minimising compliance costs for businesses and keeping administration costs low for Government.
2. Removing GST from the retail supply of food would be a significant departure from this approach. While there is an undeniable case that Government policy should tackle the causes of hunger, the Council of Ministers considers it right that GST should remain broad-based, with complimentary compensation measures to support lower-income households.
3. It is estimated that charging GST on food collects more than £10 million revenue from all Islanders. Many households can afford to pay GST on their weekly food shop and those requiring support have benefited from improved tax allowances and social-welfare payments.

Disproportionate Benefit to Higher Income Households

4. It is well known that providing tax relief for universally consumed products, such as food, creates a benefit across the range of people's incomes. That means that a benefit will be generated for those who need it most, but also those who need it least.
5. In fact, as outlined in the table1 below, the benefit of zero-rating food in Jersey increases in absolute terms as income levels - and household expenditure on food

- increase. As a result, the proposed measure is poorly targeted, with almost 30% of the monetary value of relief benefiting the richest 20% of households by income. Whilst the lowest 20% of households will only pocket around 12% of the relief. With an estimated cost of £10m per annum to government revenue, this would equate to around £1.2m going to the 20% of households with the lowest household income while over twice as much benefit (£2.9M) will go to the 20% of households with the highest household income.

1 Table produced by Economics Unit from publicly available data apart from the average income by quintile, which was provided by Statistics Jersey and uprated by the change in average earnings over the period by the Economics Unit. Figure 4.3 of Preliminary Income Distribution report 2021/22 relates to the information provided by Statistics Jersey on income distribution and Table B2 of Jersey Household spending 2014/15 relates to the information provided by Statistics Jersey on household spending.

6. Please see Appendix A for more detail on the basis for these estimates.
7. Although the benefit for higher income households could be partially reduced by only zero-rating basic foodstuffs, this would not be achievable in the timeframe and would not eliminate the benefit entirely.
8. The extent to which these tax savings are passed on by retailers is also uncertain. New research on the UK's experience of Value Added Tax (VAT) zero-rating period products using Office for National Statistics data suggests that at least 80% of the tax saving was retained by retailers (please see Appendix B). It is, therefore, unclear as to whether the potential tax savings generated by this Proposition would materialise for Islanders. As the report states, "We should be cautious before lowering tax rates in the hope that benevolent retailers and suppliers will pass the savings on to those who need it. The evidence from the "tampon tax" is that they won't."[2]

How to Fund £10 million Estimated Revenue Loss?

9. The Proposition argues that zero-rating the supply of food in Jersey is affordable and would not materially affect public finances. The Council of Ministers does not share this view. The Government Plan, as drafted, protects the stability and sustainability of public finances, but the challenging economic situation has meant that the forecast financial position in 2025 is very tight. A clear counterbalancing income stream (or expenditure reduction) would need to be found to fund this Proposition.
10. The Government's rough estimate is that zero-rating the supply of food in Jersey would reduce annual revenues by around £10m.
11. The Government plan already includes the impact of inflation on GST based on the latest Fiscal Policy Panel assumptions. GST receipts are projected to grow over the period 2023-2026, driven by growth in the compensation of employees and the increased receipts from the lowering of the de minimis level from £135 to £60 from 1 July 2023, estimated at £1.1m from 2023 onwards.

12. As highlighted in the table above, GST receipts (not including International Services Entity fees) are projected to grow by around £3.7m from 2024 (the proposed year of introduction of this measure) to 2026. Whilst this increase is assumed in the Government Plan as part of the balance of public finances, it is also clear that the positive impact of inflation on GST forecasts is substantially lower than the cost of this proposed measure.
13. This  shortfall  of  GST  revenue  would  need  to  be  made  up  through  either expenditure savings or additional revenue. A leading option on the revenue side would be to increase the standard rate of GST from 5% to 6%, which should be sufficient to replace the foregone revenue, but not address broader inflationary pressures on public finances. This would be more efficient in economic terms than other tax levers, such as income tax, but would affect the price of other goods and services, such as road fuel and electricity.
14. If the supply of food was zero-rated, but the standard rate of GST was increased to 6% we see that the benefit of zero-rating is significantly reduced, and in some cases eliminated, across the income distribution.
15. As shown in the table below, it is estimated that a standard rate of 6% would essentially offset the benefit of zero-rating food and generate a net cost for the 60% of households at the middle- and higher-income levels.

Enduring Compensation Mechanisms Since 2008

16. As a tax on consumption, GST increases the prices of goods and services at each point in the supply chain and is considered mildly regressive as lower income households tend to spend a higher proportion of their income.
17. At  the  time  of  introduction,  the  States  Assembly  recognised  this  point  and developed a package of measures to address the impact of GST (initially charged at 3%) on Islanders. These included, amongst other changes:

• Uprating Income Support by 3%;

• Introduction  of  the  GST  Food  Bonus  which  is  now  called  the Community Cost Bonus (CCB);

• Uprating Income Tax exemption thresholds by 6.5%; and

• Uprating Income Tax child allowances by 20%.

18. To recognise the impact of increasing the standard rate of GST from 3% to 5% in 2011,  the  States  Assembly  approved  an  increase  to  Income  Support  and maintenance  of  the  GST  Bonus  (now  CCB)  to  compensate  lower  income households.
19. These measures are now embedded in the tax and benefit systems and the support they provide endures to this day as subsequent policy decisions have sought to either preserve or enhance the real value of support. Removing GST from food would raise questions as to the future direction of these support mechanisms.

Response to cost of living crisis

20. P.100 is identified as a response to the cost of living crisis. Implementation could not be achieved until 2024 at the earliest. COM have already taken timely and major actions to respond to the increases in food and other costs.
21. Immediate  extra  short-term  support  was  provided  through  the  Mini-Budget approved by the States Assembly earlier this year

• Doubling the value of COLTS payments to £40 per person per month from August to December 2022

• Reducing employee and Class 2 social security contributions by 2 percentage points from October to December 2022

• Doubling the value of the CCB for 2022 from £258.25 to £516.50 and expanding eligibility;

• Doubling and guaranteeing a fixed value at £70 per month for the Cold Weather Bonus and Cold Weather Payments, regardless of temperature, for the winter months (October 2022 to March 2023 inclusive).

22. Permanent extra support as set out in the Mini-Budget is provided from 1 January 2023:

• Personal income tax thresholds and allowances will be uprated by 12%;

• Income support components will increase, giving an overall rise of at least 10.4% over the last 12 months; and

• A new Community Costs Bonus scheme will be launched in 2023.

23. These  enduring  compensation  mechanisms  provide  more  timely  support  to Islanders than the measures proposed by this Proposition. The Proposition calls for GST to be removed from supplies of food by no later than 1 January 2024. As such, any benefit generated would not be felt by Islanders until 2024. The historic compensation  mechanisms  are  already  benefiting  Islanders  and  have  been augmented by additional permanent measures in the Mini-Budget which will benefit islanders from 1 January 2023.
24. GST exemptions and zero-rating are blunt instruments to deliver support to households  most  affected  by  the  cost-of-living  crisis.  As  highlighted  in  the previous section, there is a disproportionate benefit to higher income households from the proposal to zero rate food items.

25. Recent advice from the International Monetary Fund (IMF) on policy responses to the rising cost of living highlights that interventions on food prices should be targeted to the lower end of the income distribution.3 In keeping with this advice, the measures provided in the Mini-Budget are targeted at low- and middle-income households. They also provide more support than will be delivered through the proposed  zero-rating  of  food.  As  highlighted  in  the  distributional  analysis presented above, removing GST from food is likely to reduce household spending by less than £3 a week for the lowest income households with no benefit felt until 2024.

Speed of Implementation – Copy UK VAT Schedules?

26. In ordinary circumstances, contemplating such a substantive change to the tax system would be preceded by a lengthy consultation and design period and some aspects might need to be phased in over several years. However, implementing such a measure at pace only leaves two potentially viable options:

• Zero-rate all food; or

• Adopt the zero-rating approach developed by another jurisdiction (The Proposition has suggested adopting UK VAT legislation in its entirety).

27. While zero-rating all food has an appealing simplicity, it has two significant drawbacks. Firstly, a number of potentially arbitrary decision would still need to be made on the definition of food'. A widely used tax definition is anything that is for human consumption' but a number of widely consumed products sit at the margins of this definition, such as dietary supplements.
28. It will not be possible to answer these difficult questions while also developing the  necessary  administrative  changes  to  facilitate  introduction  in  time  for  1 January 2024. This then suggests that it will be necessary to adopt the approach of another jurisdiction.
29. The rate of GST or Value Added Tax (VAT) on the supply of food differs significantly across different jurisdictions. For example, Article 98 and Annex III to the EU VAT Directive 2006/112/EC provides that Member States may apply reduced (including zero) rates of VAT to the supply of foodstuffs (which is not defined).
30. This has led to a situation in the EU where some Member States subject the supply of food to the standard rate of VAT; some allow most foods to be sold at the zero rate; and the others specify which foods should be zero-rated and which should be standard-rated. Some, such as Ireland, provide a third – intermediate – rate.
31. Given the somewhat arbitrary differences in the approaches of other jurisdictions, the Council of Ministers would look to UK VAT legislation as proposed. This would allow the Government to benefit from decades of policy development on this area, would promote consistency for UK-Jersey traders and would align the Isle of Man's approach.

3 Amaglobeli, David, Emine Hanedar, Gee Hee Hong, and Céline Thévenot. 2022. "Fiscal Policy for Mitigating the Social Impact of High Energy and Food Price s." IMF Note 2022/001, International Monetary Fund, Washington, DC.

32. However, following UK VAT practice is not necessarily ideal.
33. The UK, including the Isle of Man, has not significantly altered its position on this matter since Brexit and it continues to zero-rate the supply of "basic foodstuffs", whilst subjecting other foods to the standard rate. The UK treatment of VAT on food is very complicated and VAT tribunals and the UK courts have a long history of dealing with challenges to the system by taxpayers arguing for the zero-rating of specific goods and by HMRC defending standard-rate assessments. This has given rise to a number of notable anomalies in the UK system. For example:

• Caviar and smoked salmon would be GST zero-rated (at home but not in a restaurant);

• A tub of ice cream would be charged at 5% but a cake would be zero- rated;

• A "Jaffa-type" cake would be zero-rated but a chocolate digestive biscuit would be taxed at 5%;

• A Christmas Hamper would constitute a "mixed supply" of standard and zero-rated items and be subjected to special accounting rules; and.

• A tin of plain biscuits could constitute a mixed supply of a standard- rated (5%) tin and zero-rated biscuits.

• A ginger-bread biscuit figure with 2 chocolate eyes would be zero-rated but the addition of a chocolate belt would make it standard-rated.

34. Adopting  the  UK's  VAT  schedule  would  expose  Jersey  businesses  and  the Government to such anomalies and would, in doing so, create practical challenges to the legislative process. It is also likely to create a good deal of confusion among local retailers, especially those which are too small to run sophisticated accounting and tilling systems.
35. It  would  be  necessary  regularly  to  update  Jersey  legislation  to  reflect  UK developments.  Alternatively,  making  references  to  UK  VAT  legislation ambulatory so that future changes to the UK Schedule automatically apply in Jersey would be more straightforward but would in effect delegate this area of law making to a foreign jurisdiction. Regardless, there could be further challenges stemming from the imposition on Jersey courts to follow UK case law on the VAT treatment of food, which remains an evolving body of law.

Options for Delivering Zero-Rating and their Respective Pitfalls

36. If food were zero-rated at importation, all importers of food (largely coming from the UK) would need to make full import declarations listing the commodity codes of each product, to establish each product's tax liability. This would significantly increase  the  volume  of  full  import  entries being  lodged  with Customs  who currently only require them for around 10% of importations coming from outside the British Isles Customs Zone.
37. As goods currently imported into Jersey from the UK enjoy free movement under the Customs Union, it is unknown how many consignments contain food stuffs which would therefore need to be declared using the commodity codes.
38. The growth in consignments requiring Customs declarations could lead to an increase  in  goods  being  detained  pending  declaration.  This  could  delay  the

clearing of other goods and could place consignments of food at risk due to their perishable nature.

39. The effect of this would be felt differently across food importers in Jersey, with the most significant impacts being felt by smaller importers which are either not GST-registered or do not have JCIS approved trader status. However, the headline effect would be an increase in costs for those affected importers/businesses, potentially removing scope for price reductions. This would also run counter to Common Strategic Policy of the Council of Ministers which aims to identify opportunities to cut red tape'.4
40. The impact would not be limited to just businesses. Consignments imported by individuals are currently only detained on arrival in the island from within the Customs Union if they are manifested with either no value declared or a value over  the  de  minimis  threshold  (currently  set  at  £135,  dropping to  £60  on 01/07/2023). If GST on food is zero-rated, then each of these consignments containing food products (over the de minimis), regardless of origin, will require a full declaration using the commodity codes to ascertain the GST liability.
41. If food is only zero-rated at the point of retail sale, this could require all GST- registered food retailers in Jersey to keep the more-detailed accounts needed to make the necessary declarations to Revenue Jersey. Although it would potentially avoid the Customs issues outlined above, it would in effect shift the compliance burden from importers to retailers. As a result, Revenue Jersey could see some businesses becoming regularly entitled to GST repayments which would increase administration costs.

New Administrative Costs

42. Aside from the time needed for officers to expand Schedule 6 of the GST Law to encompass the zero-rating of food, which would displace some policy activities programmed for 2023, Revenue Jersey would need to invest in internal training; new guidance and to engage with the business community to raise awareness. Taken together with ongoing compliance, the Comptroller currently estimates a new resource demand for at least three new full-time equivalent revenue officers.
43. As highlighted in the Council of Ministers' comments to P.103/2008, "there will also be an unknown but potentially significant amount of work to mirror changes in the [UK] as and when they occur and from local rulings provided by the Commissioners of Appeal in dispute cases which could be at variance with UK VAT legislation".5
44. If zero-rated at the point of importation, there would also be significant additional administrative costs for Jersey Customs & Immigration Service (JCIS).
45. The JCIS freight handling system (CAESAR) would require major software development. A library of goods codes will need to be built, accounting for every food  item,  so  it  can  be  ascertained  on  declaration  whether  GST  would  be applicable or not. It is currently unknown what the cost of this work would be or whether this could be implemented either on or prior to 01/01/2024, considering the ongoing impacts of Brexit and the changes to the de minimis being introduced on 01/07/2023.

4 Please see p.98-2022.pdf (gov.je)

5 Please see /assemblypropositions/2008/31858-10368-892008.pdf

46. JCIS would also need to consider the additional resource implications of ongoing stakeholder engagement (e.g., carrying companies), operating procedures and compliance and audit activities.

New Compliance Costs for Businesses

47. Without a full period of consultation on the viable models for administering a zero rate, it is not possible to estimate the potential compliance costs for businesses or any administrative problems which they may face.
48. The main items of cost are likely to relate to introducing and/or updating and maintaining till systems and pricing software; accounting for sales at both zero and standard rates; and the costs of making full import entries and/or repayment claims depending upon how a zero rate would ultimately be delivered. Additional compliance costs may be generated if changes to the GST return are required to support Revenue Jersey compliance activities and/or to provide better data for Government of Jersey revenue forecasting.
49. These costs are likely to fall disproportionately on smaller local food retailers who are less likely to have existing till systems and accounting software which might interface with UK-standard VAT schedules.

Technical Note

Relief from GST in Jersey is provided in two ways:

• Zero-Rating. If goods (and some services) are zero-rated, they are not charged to GST on importation and will remain zero-rated through the supply chain (importer to wholesaler to retailer to customer). The ultimate consumer pays no GST  and  any  GST  incurred  in  making  the  supply  (i.e.,  "input  tax")  is reclaimable by the registered businesses involved. Businesses need to isolate zero-rated goods in their accounting systems.

• Exemption. If certain services (and some goods) are exempt from GST, the supplier of the service will not charge GST but equally will not be able to reclaim GST input tax on the goods and services they use in supplying that service. Where a business is not wholly involved in making exempt supplies, rules exist restricting the amount of GST they can reclaim.

Appendix A: Methodological notes on distributional analysis

The analysis of the potential distributional impacts of zero-rating food was produced using the following assumptions:

• Expenditure  data  by  income  quintile  has  been  taken  from  the  2014/2015 Household Expenditure Survey and uprated by the relevant component of RPI. For example, expenditure on food (category 1) has been uprated by the growth of the food component of RPI over the period 2014/2015 to 2022. Uprating is required as the collection and publication of more recent household expenditure data was delayed by the Covid-19 pandemic.

• Income data has been taken from the 2020/2021 Household Income Survey and uprated by the change in the average earnings index from 2020/2021 to 2022. The average of each quintile has been used instead of the upper or lower bound as this provides a direct comparison to the expenditure data. Both the figures excluding and including housing costs have been considered to provide a comparison.

• RPI uprates have been calculated by taking the average of the component index over 2014-2015 and comparing this to the September 2022 index. Income uprates have been calculated by taking the average of the index over 2020-2021 and comparing this to 2022.

The impact of an increase in the standard rate of GST to 6% has been modelled using the following assumptions:

• All other expenditure has been taken from the total of categories 1-12 in the 2014/2015 Household Expenditure Survey and then categories 1 (food) and 4.1 (net rent paid) have been subtracted. Rent is not subject to GST and so has not been included.

• The same uprates have been applied as specified in the previous assumptions list.

Appendix B: Tax Policy Associates report on the impact of VAT zero-rating period products

How the abolition of the "tampon tax" benefited retailers, not women.

An analysis of pricing evidence in ONS data

8 November 2022

Executive Summary

5% VAT applied to tampons and other menstrual products until January 2021. Then, following the high-profile "tampon tax" campaign, it was abolished. Many expected that the benefit of the tax saving would go to women, in the form of reduced prices.

However, an analysis of ONS data by Tax Policy Associates demonstrates that the 5% VAT saving was not passed onto women. At least 80% of the saving was retained by retailers (and very possibly all of it).

The key piece of evidence is this chart showing price changes before and after the abolition  of  the  "tampon  tax"  on  1  January  2021.  Ignoring  the  large  spike  in December 2020, average prices after the change are only slightly lower than before

the change. For reasons explained further below, this likely reflects normal market movements rather than the passing on of the VAT saving.

Interactive  charts  that  illustrate  this  in  more  detail  are  available  at https://taxpolicy.org.uk/tampontax_chart

Background

Campaigners had been pushing for years for the 5% VAT on menstrual products to be scrapped[2]. EU law prevented this, but in 2016 the then-Government obtained in - principle agreement with the EU[3] that this would change. In the event, these discussions were overtaken by Brexit, and it was not until January 2021 that the tax was abolished. [4]

At the time, the Treasury said this would save the average woman an estimated £40 over a lifetime.[5]

There will be widespread interest in who in fact benefited from the abolition of the tampon tax – consumers or retailers. And there is also an important tax policy point. We have recently seen proposals that VAT or duties be reduced or removed from particular products or services (e.g. VAT on the tourism industry, VAT on petrol or fuel duty  on petrol/diesel). However, these campaigns often assume that the benefit of the tax cuts will be passed onto consumers. The "tampon tax" provides further evidence that this will often not be the case.

Analysis

We  used  Office  for  National  Statistics  data  to  analyse  tampon  price  changes around 1 January 2021, the date that the "tampon tax" was abolished. We were able to do this because the ONS includes tampons (but not other menstrual products) in the price quotes it samples every month to compile the consumer prices index. Since 2017, the ONS has published the full datasets for its price sampling. [6]

We set out our methodology on page 12 below. Qualitative analysis

Our methodology results in this chart of tampon prices before and after 1 January 2021. The data is normalised to December 2020, i.e. the price on December 2020 is set at 100% for ease of reference – this facilitates easy comparisons between different products.

Or, over a longer period:

The charts show  a 6% fall in tampon prices in January 2021 - that is largely a reversal of  a 4% increase the previous month; there is then a 2.5% increase in February 2021. Overall, the average price  for the period  after the VAT abolition  is about 1.5% less than it was beforehand.

The December 2020 price spike

It could  be  suggested  that  the 4% price rise in December 2020 was a deliberate strategy to make it look as though the 5% VAT cut was being passed to consumers the  next  month.  However,  we  regard  that  as  highly  unlikely.  It  would  require astonishing cynicism on the part of retailers. It would also suggest a degree of coordination between a large number of retailers that would be difficult to arrange in practice, as well as a flagrant breach of competition law.

As we will show below, other products also show price spikes at a variety of different times,  which we expect are  driven by a complex  and unpredictable  mixture of seasonal  and  situational  supply/demand  factors.  This  seems  the  more  likely explanation. However, the fact that the December 2020 price was a "spike" means

that it would be incorrect to conclude from the December and January data that tampon pricing fell by 5% when VAT was abolished.

Comparison with other products

It is insufficient to look at tampon pricing in isolation. For example, if many other consumer products were materially increasing in price in January 2021, then the absence of an increase in tampon pricing could be consistent with the benefit of the VAT abolition being passed to consumers. However, the evidence does not show this.

Our analysis includes price movements for thirteen other products that would likely be subject to similar supply and demand effects  to tampons  –  toiletries and products made of cotton. It is important to note that none of these projects were

subject to VAT changes over the period in question. Hence, if the benefit of the VAT abolition  was  passed  on  to  consumers,  we  would  expect  to  see  a  significant divergence  between price changes in tampons and price changes in the other products. We do not.

This chart  compares price changes in tampons (the red dotted line) with price changes in tissues (the blue line).

The two datasets seem reasonably correlated on either side of 1 January 2021 (with the exception of a large spike in tissue pricing in December 2019, and another spike at the start of the data in December 2017). If the benefit of the tampon VAT abolition

was passed onto consumers we would expect a divergence between the  two datasets after 1 January, as tampon prices fell but tissue prices did not. However, we see no such effect.

This chart compares tampon pricing (red dashed line) with t-shirts (cyan line). T - shirts are largely, but not entirely, made of cotton, and therefore are in principle subject to similar demand factors:

While t-shirt pricing seems much more volatile than tampon pricing, there is again no evidence of prices diverging after 1 January 2021.

There is an interactive version of the chart here that lets the user compare tampon price movements with the other toiletry and cotton products included in the CPI, as well as the CPI itself (which, towards the end of the period covered by the chart, increases steeply as energy costs etc start to rise). Clicking on the legend on the right-hand side will add/remove additional products.

Quantitative analysis

Comparing the average change in tampon prices  for the six months before the abolition to the six months after  confirms what we see in the  above  charts  - the change in the price of tampons is broadly in line with other price changes we see in products where the VAT treatment did not change:

A similar picture is apparent over a longer period:

This implies that none of the VAT abolition was passed to consumers in the form of lower prices.

What if, in the interests of prudence, we ignore the other products that showed a drop in price, and look at tampon pricing in isolation? How likely is it that the apparent 1.5% drop in price is a real effect, and not just a function of the high variability of the pricing of all of these products? Our statistical analysis (see the "methodology" section on page 12 below) suggests that, at most, 1% of the price reduction was a real effect.

Retailers or suppliers?

VAT is charged at every level of the supply chain. So, before 1 January 2021, manufacturers supplying tampons to retailers will have charged 5% VAT, in addition to the final retail sale being subject to VAT.

However, the VAT on the sale by manufacturers to retailers would have been recovered by the retailer – it is not a real cost[7]. Hence the contract between manufacturer and retailer will quote the before-VAT cost, and that is all the retailer normally cares about. For this reason, we do not think it plausible that manufacturers could have taken the benefit of the VAT abolition in the short-to - medium term by raising their prices – this would have been immediately visible to retailers, and the purchasing power of retailers is such that they would not have permitted it. The same reasoning applies to any additional intermediaries in the supply chain, for example wholesalers.

Manufacturers may take some of the benefit of the tampon tax abolition in the longer term, for example with prices rising slightly higher than they otherwise would have done.

Conclusions

Consumers did not in fact get the full benefit of the abolition of the 5% VAT "tampon tax". At most, tampon prices were cut by around 1%, with the remaining 80% of the benefit retained by retailers. More likely, the retailers took all the benefit – amounting to £15m each year.[8]

It is open to any retailer contesting the figures in this report to publish full data showing their pricing on either side of the 1 January 2021 abolition.[9] Our analysis is of ONS data across retailers as a whole. It is, therefore, possible that some retailers did pass on the benefit of the VAT cut, and provided lower prices than the average figure in the data. That would, however, imply that other retailers provided higher prices. And where, as happened in at least one case, a retailer[10] announced tampon price cuts ahead of the actual abolition of the tax, that retailer should be able to demonstrate that the price cut happened, and as a result their prices remained diverged from other retailers up to (and perhaps beyond) 1 January 2021.

It is our hope that the power of the "tampon tax" campaign means that public pressure will cause retailers, at this late stage, to pass on the full benefit of the tampon tax abolition to consumers.

Policy implications

This is an unusual case where a product is specifically included in ONS data. Price s changes are not normally so visible; nor are they normally subject to this degree of political pressure.

The public and policymakers should therefore be sceptical of those making proposals for cuts in VAT and duties, particularly if claims are made that this will benefit consumers, and/or those on low incomes (that was generally not the case

for the "tampon tax" campaign, which was largely argued on a point of principle). If we want to support those who can't afford to pay, then the answer is to put cash directly in their hands (through the tax and benefits system), or in some cases provide free or subsidised products. We should hesitate before lowering tax rates in the hope that benevolent retailers will pass the savings on to those who need them. The evidence from the "tampon tax" is that they won't.

Methodology

Source of data

The Office for National Statistics compiles detailed monthly price quotes for a large variety of products, and then calculates price indices for each of those products. Since 2017, this data has been published.

To assess the change in tampon prices, we extracted the ONS data from December 2017 (When the data starts) through to April 2022,[11] and consolidated the index data for tampons and another thirteen broadly comparable products. We ended in April 2022 because after that point inflation effects start to dominate[12].

We then wrote a short python script to analyse and chart the data. This is freely available on GitHub  here. For clarity, all the price indices are normalised to 31 December 2020.

T-test

The bar chart on page 88 above provides a reasonably clear indication that nothing exceptional happened to tampon pricing on the six months either side of January 2021. Whilst the average price after this date was higher than the average price before, there were greater differences in most other comparable products.

Apply statistical techniques to these datasets is not straightforward given the limited number of datapoints and very high degree of volatility. It was, however, thought appropriate to run an unequal variance one-sided t-test (using the python SciPy library) to compare the pricing datasets for the six months before 1 January 2021 with those for the subsequent six months. The null hypothesis would be that the price did not change; the alternative hypothesis was that the price was lower on and after 1 January 2021.

This resulted in the following set of p-values:

Product  p-value (six month t-test) Tissues-Large Size Box  0.00001

Women's Basic Plain T-Shirt  0.00627

Boys T-Shirt 3-13 Years  0.00709

Sheet Of Wrapping Paper 0.03236

Men's T-Shirt Short Sleeved 0.09442

Toilet Rolls  0.10948

Tampons 0.14045

Kitchen Roll Pk Of 2-4 Specify  0.53249

Disp Nappies Spec Type 20-60  0.89392

Baby Wipes 50-85  0.95865

Toothpaste (Specify Size) 0.99270

Plasters-20-40 Pack  0.99694

Razor Cartridge Blades 0.99831

Toothbrush 0.99847

As expected, the p-value for tampons is not significant (> 0.05). More significant p - values were achieved for six other products, four of which were actually significant at 5%. Those products did not receive any change in VAT treatment over the period in question so, absent other unknown factors, this should be regarded as a product of the volatility of  pricing decisions in a complex market environment, as well as potentially unknown supply and demand factors.

If we look at longer periods than six months then the p-value for tampons becomes more significant (because, although the average does not change, the number of datapoints  increases,  and  significance  becomes  "easier"  to  attain).  At  eight months, a significant result is achieved (p-value=0.03201), although  given that more significant results are (again) obtained for other products, this result should be treated with caution.

Nevertheless,  if  the  tampon  pricing  results  are  viewed  in  isolation,  they  are compatible with the price having decreased after 1 January 2021. We can estimate the maximum likely extent of that decrease by constructing a synthetic tampon pricing sequence, identical to the actual tampon pricing sequence but with an increase of x% from 1 January 2021. If  a t-test of that synthetic pricing sequence does  not  provide  a  significant  p-value  then  that  implies  that  the  statistically significant difference between the pre-January 2021 pricing and post-January 2021 pricing is limited to a tampon price reduction of x%.

Testing different values of x, the p-value ceases to be significant with x at 1% (p- value equal to or greater than 0.0935 for any given range of months) (the precise methodology utilised can be seen in the GitHub code).

Hence, we can conclude that there is only weak evidence of any change in tampon pricing reflecting the cut in VAT, with the greatest realistic extent of any price reduction being approximately 1%.

Limitations

The analysis in this paper is subject to a number of important limitations:

• The ONS adopts a methodology that is designed to produce statistically rigorous results across consumer prices as a whole. It is not necessarily intended to provide robust figures for changes in the price of one product. Nevertheless, around 250 tampon prices[13] are sampled each month[14].

• The prices of tampons and the other consumer goods considered in this paper will be affected by numerous factors - the supply of raw materials, energy costs, and sheer random happenstance. It is therefore unsurprising that we see a large amount of month-to-month variation in prices; this makes it difficult to identify separate real trends from noise. More sophisticated statistical methods than a t-test are therefore not helpful (difference-in-difference and synthetic control methods were attempted, but did not produce meaningful results).

• The January 2021 VAT change applied to all menstrual products,[15] however ONS data only covers tampons. It is therefore possible in principle that the

benefit of VAT abolition was passed to consumers of e.g. sanitary towels, although it is not obvious why that would be the case.

Acknowledgments

Thanks to Arun Advani of the University of Warwick for his advice on statistical techniques, to Gemma Abbott for her advice on the background to the "tampon tax" campaign, and to Laura Coryton for her help and advice. Any errors in our analysis are the sole responsibility of Tax Policy Associates, as are the conclusions we draw from it.

And many thanks to the Office for National Statistics for publishing the data at a level of detail that facilitates this kind of analysis, and also for responding swiftly and helpfully to our queries.

About Tax Policy Associates

Tax Policy Associates Ltd is not-for-profit company, with three key aims:

• Providing policymakers and politicians with expert and impartial tax policy advice, in the UK and worldwide.

• Partnering with academics and others researching tax and tax policy.

• Working to improve the public understanding of tax, by assisting journalists writing on and investigating tax policy, tax avoidance and tax evasion.

Tax Policy Associates Ltd is financially independent. It accepts no donations, charges no fees, and undertakes no commercial work.

To contact us please email info@taxpolicy.org.uk

About Dan Neidle

Dan Neidle spent almost 25 years as a tax lawyer, and was head of tax at the London office of one of the largest law firms in the world. During his career, Dan advised corporates, governments, regulators, central banks and NGOs on tax and tax policy.

Dan is now bringing the depth of his experience and specialist expertise to improving tax policy, and shaping and informing the debate around tax.

Legal

This report is licensed under the Creative Commons BY-SA 4.0 licence, and so its contents may be used in any form provided appropriate attribution is made to Tax Policy Associates Ltd.