Who pays sales taxes? Evidence from French VAT reforms, 1987–1999

https://doi.org/10.1016/j.jpubeco.2006.12.004Get rights and content

Abstract

The point of this paper is to provide visual evidence of tax shifting and to measure empirically the distribution of the sales tax burden between consumers and producers. For that purpose, two French reforms are studied. These reforms entail steep decreases of the VAT rate. September 1st 1987, the VAT rate on car sales went down from the luxury-rate of 33.33% to the full-rate of 18.6%. September 1st 1999, the VAT rate on housing repair services went down from the full-rate of 20.6% to the reduced-rate of 5.5%. The consumer share of the sales tax burden is 57% in the new car sales market and 77% in the housing repair services market.

Introduction

The present paper studies the economic impact of sales tax reforms, and attempts to provide visual evidence of tax shifting and to measure the distribution of the sales tax burden between consumers and producers. Sales taxation is heavily used all around the world, through different kinds of tax. In Europe, a VAT system is applied and those taxes make up a large part of fiscal revenue. For an example, the 2006 French budget forecasted VAT would represent 48.8% of the French total fiscal revenue.

Indeed, sales taxes are quite easy to collect. Moreover, it is generally assumed that they do not generate distortions, as they are paid only by consumers. However, the sales tax burden is not paid by consumers only but is shared between consumers and producers, and may be different for different goods.

The point of this paper is to point out this fact, through visual evidence first, then through econometrical estimates. Therefore, we use price data around the time of two French VAT reforms. The consumer share is found to be 77% in the housing repair service market, and 57% in the new car market.

Usual indirect taxation recommendations do not take tax shifting into account. The question of the consumer share of the tax burden may have political implications, as it may affect the optimal sales tax schedule. For an example, the values of VAT shifting on prices should have an impact on the European debate about reducing VAT rates for labor intensive services. However, little is known about the actual indirect tax share borne by the consumer.

A number of theoretical papers have studied the question of sales tax shifting on prices. Katz and Rosen (1985) studied a closed Cournot oligopoly. Stern (1987), Besley (1989) considered the free-entrance impact, based on the conjectural variation model developed by Seade (1980). Delipalla and Keen (1992) compared consumer shares of specific or ad valorem sales taxes. They found that the consumer share is higher in the case of specific sales taxes. These papers all pointed out the influence of imperfect competition on tax shifting on price. In essence, the consumer share may be higher than 100% under imperfect competition.

From an empirical point of view, Besley and Rosen (1999) tested tax shifting through a number of local sales tax variations in the United States. They found very few markets with a consumer share of the sales tax burden significantly different from 100%. The few goods whose tax shifting on prices is found significantly different from 100% are sold through the retail industry. Taxes on these goods are over-shifted, which means that the consumer share is higher than 100%.

Furthermore, there exists a literature on excise tax incidence on prices. Concerning tobacco, most studies estimate only demand elasticity. However, Barnet et al. (1995) estimated a general equilibrium model, with tax shifting on prices. They found that consumer share is equal to 102% for federal excise taxes and 90% for state excise taxes. Delipalla and O'Donnell (2001) tested the theoretical results of Delipalla and Keen (1992) in the case of the European tobacco market. They confirmed that consumer share of specific taxes is higher than consumer share of ad valorem taxes. Young and Bielinska-Kwapsisz (2002) studied the alcohol market with Washington D.C. price data. They found that excise taxes on alcoholic beverages are over-shifted on prices.

These papers studied small tax rate variations. Therefore, price shifting cannot be discerned in figures, and these studies cannot measure precisely the values of the consumer share. The present paper study large sales tax rate variations.

The rest of the paper is organized as follows.

Section 2 presents the data. The empirical work uses price data at the time of two reforms that induced ample VAT rate variations. September 1st 1987, the French VAT rate on car sales went down from the luxury-rate of 33.33% to the full-rate of 18.6%. September 1st 1999, the French VAT rate on housing repair services went down from the full-rate of 20.6% to the reduced-rate of 5.5%.

Section 3 presents the visual evidence and the results of estimates. The consumer share is 77% in the housing repair service market and 57% in the new car market.

Section 4 concludes on the distribution of the sales tax burden between consumers and producers, and its implication as a matter of optimal taxation. The point is that optimal tax rate should take into account tax shifts on prices.

Section snippets

Data

The point of this paper is mainly empirical; it is to estimate precisely sales tax shifting on prices. For that purpose, two tax reforms are studied. It is necessary to study tax rate variations large enough to get precise estimates of the consumer share. There have been few large tax rate changes, and the only reforms that can be studied are those changing the sales tax category of one kind of good.1

Tax shifting measures

The point of the present section is to estimate precisely consumer shares. Consumer share is the share of the tax burden paid by the consumers. This consumer share is noted s and also represents the tax shifting on prices.

The tax analyzed here is VAT. It is an ad valorem sales tax applied on the before tax price. The before tax price is p1+τ where p is the sale price. The amount of the tax is then τp1+τ. After a VAT rate variation, the consumer price variation is dpdτ, and the producer price

Conclusions

Analyzing huge VAT variations, this study provides sharp consumer shares estimates for two very different markets. The estimations find for both markets that taxes are under-shifted on prices, as in Delipalla and O'Donnell (2001). Moreover, the consumer share in the housing repair service market– a close to perfect competitive market– is very significantly higher than the consumer share in the car market– a quite closed oligopoly with very few firms competing.

These empirical results are

Acknowledgement

I thank the editor and referees for helpful comments. I am grateful to Thomas Piketty for his help and his comments. I am grateful to Michel Pierssens.

References (11)

There are more references available in the full text version of this article.

Cited by (92)

  • Temporary VAT rate cuts and food prices in e-commerce

    2024, Journal of Retailing and Consumer Services
  • I would if I could: Passing through VAT reductions in the german rail industry

    2022, Economics of Transportation
    Citation Excerpt :

    While tax reductions are often only passed through partially, the pass-through rates for tax increases are generally higher than for tax reductions (e.g. Benzarti et al., 2020; Doyle and Samphantharak, 2008; Kenkel, 2005). Additional analyses show that pass-through rates can differ significantly between sectors (Carbonnier, 2007), and that the degree of over- or under-shifting also depends on the type of taxation, with a greater effect of specific taxation than of ad valorem taxation (Delipalla and O’Donnell, 2001). In addition, the level of competition (Hindriks and Serse, 2019) or the type of store (Harju et al., 2018) can impact on pass-through rates, and constraints in the supply chain reduce tax pass-through rates (Marion and Muehlegger, 2011).

  • Are VAT reforms an effective tool for promoting culture? A quasi-experiment in Spain

    2022, Journal of Policy Modeling
    Citation Excerpt :

    Anti-competitive practices among suppliers to set (almost) uniform-pricing in movie theaters could also explain this undershifting. This result is also in line with previous evidence arguing that labor intensive services (present in many cultural activities analyzed here) tend to undershift VAT reductions, since increasing production in order to accommodate to a lower price is harder than to decrease production in response to a higher price (Carbonnier, 2007). A review of the literature highlights that the econometric analysis of the causal impact of VAT reforms on the consumption of cinema and performing acts have mainly used two different approaches: estimating demand equations (Prieto-Rodríguez et al., 2015; García-Enríquez & Echevarría, 2018) and applying quasi-experimental methods (Ateca-Amestoy et al., 2020).

  • The heterogeneous incidence of fuel carbon taxes: Evidence from station-level data

    2022, Journal of Environmental Economics and Management
View all citing articles on Scopus
View full text