Consequences of Consumer Sales Taxes in Light of Strategic Suppliers Anil Arya The Ohio State University Brian Mittendorf The Ohio State University March 2016 Consequences of Consumer Sales Taxes in Light of Strategic Suppliers Abstract Taxes levied on retail sales are a ubiquitous form of taxation, both in the US and abroad. While considerable study has examined the economic effects of such sales taxes vis-a-vis consumer demand, surprisingly little attention has been focused on the effects up the supply chain. In this paper, we consider a parsimonious model of retail products sold in a variety of consumer markets (each of which may face different tax rates) when retailers rely on strategic suppliers for inputs in the products they sell. We find that when suppliers have and use pricing power, the imposition of sales taxes at the retail level has reverberations on supply markets – sales taxes undercut consumer demand which also makes retailers more price-sensitive, and suppliers respond to this by cutting prevailing input prices. Not only does this "soften the blow" of sales taxes on retail profit in the market (tax jurisdiction) in question, it also boosts retail profit in other markets since the retailer is able to parlay the lower input prices into greater margins therein. Besides reversing several conventional views of economic consequences of sales taxes, the results also provide key implications for tax policy when firms operate in and care about multiple tax jurisdictions. 1. Introduction The economic consequence of imposing sales taxes on consumer products, a long- studied and oft-debated question, has taken on greater importance in recent years with the proliferation of online sales and the resulting taxation implications. Beyond the issue of effective implementation, questions of the efficacy of introducing widespread taxation on internet purchases (the latest legislative incarnation of which is the "Marketplace Fairness Act") have brought the broader issue of sales taxes to the forefront of many public policy discussions. While most are focused on how sales taxes affect consumer behavior and how that, in turn, affects retail sellers' decisions, little (if any) attention has been paid to the consequences of such end-user taxes on input providers at the wholesale level. Starting with the notion that end-user sales subject to taxation typically run through a nontrivial supply chain before reaching consumers, this paper seeks to examine if and how such supply chain relationships are altered by taxes imposed on end users. To elaborate, we consider a parsimonious model of a retailer selling to consumers, where consumer purchases are subject to taxation. In this model, we incorporate two distinct practical features: (i) the retailer relies on a wholesale supplier in providing goods to consumers; and (ii) the retailer sells to consumers in various markets, each of which may be subject to different tax rates. The former feature captures the notion that retail providers are rarely vertically integrated but instead rely on suppliers for their various products. The latter feature captures the idea that different states or distribution methods (e.g., online vs. in-store) are subject to different taxation despite the fact that the goods themselves are equivalent. Our model demonstrates that these two key features work in concert to alter traditional thinking about sales taxes and their consequences. First, we show that increases in sales taxes in one jurisdiction, while stunting consumer demand and thus restricting incentives for retail supply therein, have notable reverberations in input markets. By 2 restricting retail margins, high sales taxes limit retailers' willingness to pay for inputs which, in turn, incentivizes price cuts at the wholesale level. Second, we show that due to this input market effect, higher sales taxes in one jurisdiction boost demand in other jurisdictions even if consumers in the jurisdictions do not overlap. The reason for this is that high taxes in one market compel lower input prices and these lower input prices incentivize greater retail supply in other, low-tax markets. The cross-market interlinkage introduced by the nexus of differential end-user taxation yet reliance on a common input across markets has some notable implications for tax policies and firms' lobbying efforts (or lack thereof). First, a government seeking a tax increase may meet little resistance (or even tacit support) from a retailer if that retailer stands to benefit from input market price-cuts the proposed increase may engender. The paper provides precise conditions under which the multi-market retailer profits despite an increase in a market's sales tax rate. Second, one government may benefit from higher tax collections when another government imposes higher taxes even though the consumers in these markets are independent. Third, tax hikes in one jurisdiction can boost welfare if their effect is to boost supply chain efficiency and better level retail supply across markets. These results may explain, for example, tacit support by retailers of sales taxes at bricks- and-mortar locations yet adamant and organized opposition to sales taxes for online purchases. They may also explain the tendency for governments to boost sales taxes when their tax rates are lower than others, even when income and property taxes differ vastly (i.e., it is a differential in sales tax rate, not the overall tax burden, that introduces a benefit to altering sales tax rates). To extend the analysis and test its robustness, we examine two modeling variants. First, we consider consequences of unit (excise) taxes on the analysis, showing that it is taxation at the consumer level (not a percentage tax rate) that is the key feature, but also showing that input markets notably alter the traditional comparison of sales and unit tax methods. Second, we examine the consequences of competition, both at the retail and 3 wholesale level. This extension demonstrates that the key considerations identified herein persist under competition, but that increased competition in either the wholesale or retail arena mitigates the supply market effects we identify. This suggests that the input market effects should be most pronounced in practice for markets characterized by substantial supplier power and/or retailer market concentration. This paper's findings fit into two broad and heretofore distinct streams of literature: (a) economic consequences of tax policy and (b) supply chain pricing. In terms of extant research on taxes and their economic consequences, there is of course a voluminous collection of research examining how corporate taxes affect economic growth, income taxes affect labor incentives, and sales taxes affect retail firm behavior. Most closely related to the present study is the latter stream that hones in on consumer sales taxes. Much of the focus there is on if and how such taxes affect retailer behavior and whether imposing $1 of retail taxes actually leads to an increase of out-of-pocket consumer cost of $1. This tax incidence literature notes that the chilling effect of taxes on consumer demand may restrict retail quantities and boost consumer out-of-pocket price beyond just the tax imposed. Such "overshifting" can arise both when taxes are imposed on firms and when they are imposed on consumers. Studies have shown that such overshifting can be mitigated by, among other things, retail competition and excess capacity (Anderson et al. 2001a; Anderson et al. 2001b; Marion and Muehlegger 2011). What is notably absent from this extensive research on taxation policy, however, is an examination of the upstream consequences of sales taxes when the supply chain for retail goods is imperfectly coordinated, the focus herein. A consequence is that our paper demonstrates that conventional results that firm profits are always dampened and that welfare is necessarily reduced by higher taxes may well be reversed when supply chain efficiency gains are taken into account. Additionally, by considering an uncoordinated supply chain subject to a dominant supplier, we show that overshifting may be minimized due to the manner in which retail sales taxes convince suppliers to cut their prices. In fact, 4 when firms operate in multiple consumer markets, the supply market consequence can actually lead to lower retail prices in more than one market. These results may provide some support for the empirical evidence suggesting undershifting – tax increases of $1 leading to out-of-pocket consumer cost increases less than $1 – is prevalent in many industries (e.g., Besley and Rosen 1999; Poterba 1996). Given the critical role of multiple retail jurisdictions served by a retailer, our study is naturally tied to, and offers implications for, the nascent literature examining how differential tax rates between in-store and online purchases affect consumers and firms who each operate in both markets (e.g., Baugh et al. 2014; Goolsbee and Zittrain 1999; Hoopes at al. 2014). More broadly, the multi-market emphasis is in line the recent work of Hamilton (2008) who examines how taxes imposed in some retail markets affect other retail markets of multiproduct firms. In Hamilton (2008), the interlinkage comes about due to complementarities in goods and other intrinsically-tied consumer demand; here, in contrast, the interlinkage arises due to reliance on the same input for multiple retail markets with independent consumer demand. The second key stream of literature this paper builds upon concerns upstream markets and their behavior. At the core of the bulk of these studies (and this paper too) is the inherent conflict of interest in pricing. Starting with the seminal work of Spengler (1950), many have examined distortions introduced by above-cost pricing by suppliers and retail firm efforts to alleviate such distortions (for thorough and excellent reviews of the literature on supply chain pricing, see Katz 1989 and Lariviere 1998). Complicating matters are strategic efforts by suppliers to solidify excessive input prices, including sabotaging downstream investments (Pal et al. 2012) and even self-sabotage (Sappington and Weisman 2005). Existing literature has shown that concerns over supplier pricing can explain a variety of practices, including the introduction of direct sales by a supplier (Tsay and Agrawal 2004), cost-plus transfer pricing by a firm (Arya and Mittendorf 2007), product 5 returns policies (Pasternack 1985), quantity flexibility or revenue-sharing arrangements (Tsay 1999; Cachon and Lariviere 2005), and propping up loss-leader products (Arya and Mittendorf 2011). This paper adds taxation at the retail level as an additional consideration for supply chains, demonstrating that the imposition of sales taxes in one tax jurisdiction can have important ramifications for supplier pricing which, in turn, has ramifications for retail pricing and supply in other jurisdictions. The paper proceeds as follows. Section 2 presents the basic model. The results are presented in Section 3: the retail market equilibrium under sales taxes is identified in 3.1; the benchmark case of a vertically-integrated supply chain is presented in 3.2; the consequences of strategic supplier pricing are identified in 3.3; an extension to unit rather than ad valorem sales taxes is studied in 3.4; and the effects of competition, both at the retail and wholesale level, are examined in 3.5. Section 4 discusses implications and concludes the paper. 2. Model A firm relies on a supplier for a key input of products it provides in various consumer markets. In market i, consumer purchases are subject to an ad valorem sales tax, with the tax rate denoted t ≥0. The distinct markets can reflect different geographic areas i with varying tax rates (cities, counties, or states); different distribution methods that face different tax regimes (online vs. in-store); and/or different consumer uses taxed differently (food purchased for dine-in vs. to-go). To highlight the role of sales taxes, we presume that beyond any differential tax rates, the markets are identical and independent. In particular, (inverse) consumer demand in market i is pˆ = a−q , where q reflects the i i i quantity of units sold in market i and pˆ reflects the consumer's out-of-pocket price paid i for each unit. The consumer's out-of-pocket cost consists of stated retail price, p , plus i taxes, t p, i.e., pˆ = p[1+t ]. i i i i i 6 The monopolist supplier produces the inputs at unit cost v, v≥0, and sets a unit input (wholesale) price of w, w≥0. As is typically the case, consumer taxes are only levied at the retail level and not wholesale level. Thus, the retail firm's out-of-pocket cost for each unit of input is w; denote any subsequent costs to convert and sell each input by c, c≥0. Given this formulation, the supplier's and firm's profits in market i are ⎡a−q ⎤ Πs =[w−v]q and Π = i q −wq −cq , respectively. Similarly, the taxes i i i ⎢ ⎥ i i i 1+t ⎣ i ⎦ ⎡a−q ⎤ collected in market i are T =t i q , and total welfare (surplus) in market i is i i⎢ ⎥ i 1+t ⎣ i ⎦ q2 Ψ =[a−q ]q −vq −cq + i . i i i i i 2 The sequence of events is as follows: Tax rates, t , are Supplier sets Retail firm chooses Profits, taxes, and i established. wholesale price w. retail quantities q . total welfare are i realized. Figure 1: Timeline Denote the number of consumer markets the retailer serves in equilibrium by n, n≥2. Given the prevailing market tax rates, {t ,...t }, let t denote the mean tax rate and 1 n tmax denote the maximum tax rate. Given this, we presume consumer demand in the markets is sufficiently large to ensure interior solutions; in particular, ⎡(1+tmax)(1+t)⎤ a> ⎢ ⎥[c+v]. 1−tmax +2t ⎣ ⎦ Using this basic setup as a backdrop, we examine how (changes in) sales taxes affect firm profits, retail sales, tax collections, and welfare. The analysis is conducted with a focus on two key features novel to the setting: (i) retail sales occur in different consumer markets facing possibly different tax rates; and (ii) retail sales require wholesale purchases. 7 3. Results In deriving the equilibrium outcome, we work backwards in the game beginning with the retail market equilibrium. 3.1. Retail Market Outcome For a given prevailing wholesale price, w, the firm chooses retail quantities, q , i i =1,...,n, to maximize its total profit, Π = ∑n Π . The first-order condition of the firm's i=1 i problem yields the retail market quantities, consumer (out-of-pocket) prices, and retail prices: q (w)=[1/2] [a−(1+t )(c+w)] , pˆ (w)= a−q (w)=[1/2] [a+(1+t )(c+w)], and i i i i i p (w)= pˆ (w)/[1+t ]=[1/2] [a/(1+t )+c+w], i =1,...,n. (1) i i i i The outcome in (1) reflects the usual properties of retail output – it is increasing in consumer demand (a) and decreasing in cost (c + w). The new feature here is that retail quantities are also reduced by higher sales tax rates. Roughly stated, the tax-imposed cost on the consumer is borne by the retail seller in the form of an increase in the "effective" marginal cost (see, e.g., Anderson et al. 2001). Viewed in terms of prices, consumers' out-of-pocket payments are of course boosted by an increase in the tax rate. This reduces their willingness to pay and, as a consequence, leads to a cut in the retail price – this is captured by the term a/[1+t ] in p (w). i i This retail outcome results in tax collections in market i of: ⎡a−q (w)⎤ [a2 −(1+ti)2(c+w)2]ti T (w)=t i q (w)= . (2) i i⎢ ⎥ i 1+t 4[1+t ] ⎣ i ⎦ i The retail firm's profit across all markets, denoted Π(w) is: n [a−(1+t )(c+w)]2 Π(w)≡ Π = ∑ i . (3) qi≡qi(w) 4[1+t ] i=1 i 8 Finally, total welfare given the retail equilibrium, Ψ(w), is as follows with σ2, t σ2 = ∑n [t −t]2 / n, denoting the tax-rate variance: t i=1 i 2 n[c+v] [2a{3+t}−(c+w){3+2t −t2 −σ2}] 3na t Ψ(w)≡ ∑n Ψ = − . (4) i=1 i qi≡qi(w) 8 8 Given these outcomes, the paper's focus is on how supply market effects alter the traditional views of sales taxes. To do so most clearly, we first present a benchmark case where supply market effects are absent. 3.2 Insourcing Benchmark Say the retail firm makes its inputs in-house (at cost v), and does not need to rely on a strategic supplier for wholesale goods. Alternatively, one can view this as the outcome when there is an integrated supply chain or perfectly competitive input market. In each interpretation, the outcome corresponds to the equilibrium in Section 3.1 with w = v. Thus, using (1) through (4), with w = v, provides the solution in the insourcing case. This benchmark equilibrium yields the prevailing views of the consequence of higher sales taxes as summarized in the first proposition. (All proofs are provided in the appendix.) PROPOSITION 1. Under insourcing, an increase in sales tax in market i dΠ(v) (i) reduces the firm's profit, i.e., <0; dt i (ii) can increase or decrease tax revenues in market i while leaving tax revenues in other markets unchanged; and (iii) increases consumer "out-of-pocket" price in market i while leaving consumer dpˆ (v) dpˆ (v) prices in other markets unchanged, i.e., i >0 and j = 0 for j ≠i; and dt dt dΨ(v) i i (iv) decreases welfare, i.e., <0. dt i The benchmark proposition confirms the traditional thinking about sales taxes. First, due to the chilling effect on consumer demand, higher sales taxes undercut firm