The Effect of VAT Threshold on the Behavior of Small Firms Jarkko Harju, Tuomas Matikka and Timo Rauhanen∗ June 30, 2015

Preliminary version Abstract We study the effects of the value-added tax (VAT) threshold on the behavior of small firms using register data on the universe of Finnish firms. We find sizable bunching just below the sales-based VAT threshold (8,500 euros), which implies that small firms actively avoid VAT liability. We find that even considerable reductions in the VAT rate at the threshold do not affect the extent of the bunching response. This indicates that compliance costs of VAT are crucial for small firms. In addition, we find no explicit evidence of tax avoidance or evasion, which suggests that firms respond by reducing output. Also, we find that bunching behavior is very permanent, which implies that the VAT threshold decreases the growth of small firms. Keywords: Value-added tax, VAT threshold, bunching, small firms JEL codes: H21, H25, H32 ∗ VATT Institute for Economic Research (Helsinki, Finland). [email protected]

1

[email protected], [email protected],

1

Introduction

Value-added tax (VAT) is a commonly applied form of consumption taxation and a crucial component of tax revenue in many countries. Most VAT systems include varying thresholds below which firms are exempt from remitting and reporting VAT. For example, in the EU, VAT threshold varies between 0-100,000 euros. Despite the potentially large efficiency effects of these thresholds, there is very limited empirical evidence available on the effects of VAT thresholds on the behavior of small firms. Setting an appropriate VAT threshold is challenging (Keen and Mintz 2004, Kanbur and Keen 2014). One of the focal trade-offs lie between the large administrative burden of collecting VAT from small firms and the efficiency effects of excluding part of firms from VAT. Therefore, the key question in VAT system design is how small firms respond to a VAT threshold. In this study we present comprehensive evidence on the behavioral effects of the VAT threshold among small firms. We utilize detailed register data on the universe of Finnish businesses, including firms below the VAT threshold. We use the bunching method in order to provide clear and robust evidence of behavioral effects (Saez 2010). To our knowledge, this paper is the first to show how small firms respond to a VAT threshold. One further important aspect of VAT regulation is the role of small firms in economic development. Small and especially young businesses are often regarded as important determinants of economic growth (see e.g. Haltiwanger et al. 2013, Decker et al. 2014). Simultaneously, many tax rules and regulations, such as the VAT threshold, are size-dependent. These types of rules might decelerate economic activity among growing firms, in contrast to widespread objectives to enhance the growth of small businesses. In Finland, firms with annual sales below 8,500 euros are not liable to pay VAT and separately report sales to the Tax Administration. Relatively low VAT thresholds are very common. Half of the EU countries apply thresholds below 25,000 euros, including for example Germany, Belgium and Denmark. In general, small firms comprise a large share of all businesses. In Finland, one third of all registered firms have turnover below 25,000 euros. Among young and potentially growing firms, the share of businesses with small turnover is even larger. Over 40% of firms that are younger than three years have turnover below 25,000 euros. Firm responses to VAT threshold could imply several academically and politically relevant issues. Therefore, it is important to understand why and how firms respond to the threshold. First, firms could respond to the VAT threshold both because of increased tax liability and increased compliance costs above it. In addition to the direct cost induced by the remitted VAT, firms face other costs when the VAT threshold is exceeded. For example, firms need to frequently report their sales and purchases for VAT purposes, and the owners of the firms need to understand the complicated VAT rules and regulations. Also, exceeding the threshold could make tax avoidance and evasion more difficult. In order to understand the profound reason for avoiding VAT liability, we need evidence on whether the behavioral response is driven by tax incentives or compliance costs. 2

We utilize variation in tax incentives and compliance costs to analyze why firms react to the threshold. In Finland, before 2004 the VAT liability and the average VAT rate increased sharply at the threshold, as firms marginally above the threshold were liable to fully pay the VAT on all sales. In 2004, Finland introduced a VAT relief scheme in which remitted VAT increases gradually above the threshold, implying a significantly smaller increase in VAT liability. Thus the reform drastically changed tax incentives at the threshold, which allow us to disentangle the effects of tax incentives and compliance costs. In addition, targeted VAT rate reductions to certain specific types of services allow us to study whether the response is driven by tax incentives. In Finland, the VAT rate for hairdressers decreased by nearly 60% in 2007. However, the VAT rate for similar services such as beauty salons remained unchanged. This offers us additional variation to study whether tax incentives affect the extent of the behavioral response. Second, firms can respond to the VAT threshold both by reducing output, or by engaging in various tax avoidance activities or systematic underreporting of sales. This nature of the response entails important policy implications. For example, real output responses could be more detrimental in terms of welfare compared to avoidance activities with a smaller effect on the extent of business activity (see e.g. Slemrod 1992). Also, avoidance and evasion responses can be more easily affected by the policy maker, compared to influencing real economic activity of small firms. We analyze the nature of the response by studying how the production factors of the firms, such as equity and expenses, develop around the VAT threshold. Potential discontinuous changes in production factors exactly at the VAT threshold indicate changes in behavior caused by this regulation, and shed light on how firms respond to the threshold. For example, if firms would avoid exceeding the threshold by systematically underreporting their sales, we should observe larger firms bunching just below the VAT threshold. Third, in terms of longer-run efficiency, it is essential to know how the VAT threshold affects the growth of small firms. The VAT threshold could significantly hinder growth if firms avoid exceeding the threshold for a prolonged period of time. The panel structure of the data allow us to follow firms over time, which enable us to characterize the effect of the threshold on growth and the scale of business activity. We find that the VAT threshold has notable effects among small firms in Finland. Firms bunch very actively just below the threshold, which implies that small firms actively avoid VAT liability. We find that even considerable reductions in the VAT rate at the threshold do not affect the extent of the bunching response. This indicates that compliance costs are the key factor in explaining observed behavior. In addition, we find suggestive evidence that a decrease in compliance costs reduces bunching below the threshold, which further highlights the key role of compliance costs. We find no clear evidence of tax avoidance or evasion. This suggests that firms respond by changes in output and real economic activity, implying that the efficiency effects of the VAT threshold can be

3

notable. In addition, we find that bunching behavior is relatively permanent, as a significant portion of firms avoid exceeding the threshold for many consecutive years. This implies that the VAT threshold decreases the growth of small firms. Despite the fiscal importance of VAT and the generally applied sales thresholds, only a few previous papers study the effects of these thresholds. The theoretical literature has focused on determining the rules for optimal VAT threshold. Keen and Mintz (2004) and Kanbur and Keen (2014) show that the optimal VAT threshold depends on administrative and compliance costs, and the extent of the effect of the threshold on firm behavior. Empirically, Onji (2009) was the first to detect clear effects of the VAT threshold on the distribution of firms in Japan. He shows that relatively large Japanese firms reacted to the introduction of a VAT threshold by splitting into smaller entities, reflecting clear tax avoidance behavior. Li and Lockwood (2015) show that firms in the UK bunch actively at the relatively large VAT threshold (approx. £90,000). In addition, Li and Lockwood (2015) show that the behavioral response to the VAT threshold measures the excess burden of VAT. Also, Waseem (2015) observes a strong clustering of firms at the VAT threshold in Pakistan. This paper proceed as follows: Section 2 describes the VAT system and the VAT threshold in Finland. Section 3 presents the methodology and Section 4 describes the data. Section 5 offers the results and Section 6 concludes the study.

2

Value-added taxation (VAT) and VAT thresholds

2.1

Value-added taxation in general

Most developed countries use the value-added tax (VAT) as their primary consumption tax system. VAT is usually a broadly based tax assessed on the value added to goods and services. The amount of value added is calculated by subtracting the amount of externally purchased goods and services from the value of goods and services produced and sold. In short, the VAT assessment process is the following: each trader in the chain of supply from manufacturers to retailers charges VAT on their sales. Firms are entitled to deduct from this amount the VAT paid on purchases. VAT is remitted to the tax authorities by the seller of the goods and services. VAT is an important source of tax revenue in many developed countries. For example, among all OECD countries almost one-fifth of all tax revenue is collected by VAT. However, the variation in VAT revenue is large across countries. In many VAT systems firms with annual sales under a certain threshold are not required to register and remit VAT. Figure 1 shows these annual sales thresholds among OECD countries in 2014. The Figure highlights that thresholds vary notably across countries. While some countries levy VAT on all sales without a specified VAT threshold (e.g. Sweden and Turkey), some countries apply relatively high 4

thresholds around 100,000 euros (e.g. Switzerland and the UK).

VAT thresholds in OECD countries in 2014 (in euros) Chile Mexico Spain Sweden Turkey Netherlands Greece Belgium Norway Iceland Denmark Finland Portugal Estonia Israel Germany Hungary Korea Canada Luxembourg Average Austria Italy Poland Czech Republic New Zealand Slovak Republic Slovenia Australia Japan Ireland France Switzerland United Kingdom

0

20,000

40,000 60,000 Sales (in euros)

80,000

100000

Source: OECD Statistics

Figure 1: Annual sales thresholds of VAT registration among OECD countries in 2014 (in euros)

2.2

VAT in Finland

Finland, as a member of the European Union (EU), applies the general EU VAT legislation. All members of the EU apply a standard VAT rate of at least 15%. The EU allows member countries to use a maximum of two reduced VAT rates for specific products and services, such as food and pharmaceuticals. The standard VAT rate in Finland is 24% in 2014. The standard rate applies to most goods and services. Finland uses two reduced VAT rates: a 14% rate is applied to e.g. food and restaurant services, and 10% is applied to e.g. books and pharmaceuticals.1 Some goods and services are exempt from VAT. These include financial and insurance activities, letting and operation of dwellings, education, health services and social work activities. A firm that sells solely these goods or services are not liable to pay VAT. In addition, Finland has applied a special reduced VAT rate for certain selected labor-intensive services. For example, in 2007-2010, the VAT rate for hairdressers and barbers was decreased from 22% to 8%. However, other similar services such as beauty salons were not subject to the decreased VAT rate. In addition to hairdressers, the VAT rate for restaurants was decreased from 22% to 13% in July 1 Until 2010, the standard VAT rate was 22% in Finland. The standard VAT rate was increased to 23% in 2010, and to 24% in 2013. The first reduced rate was 17% until 2009. It was decreased to 12% in 2009, and increased to 13% in 2010 and to 14% in 2013. The second reduced rate was 8% until 2010, and was increased to 9% in 2010 and to 10% in 2013.

5

2010. VAT registered firms are obliged to regularly file periodic tax returns to the Finnish Tax Administration. The reporting obligation covers sales at different VAT rates, input purchases, zero-rated sales, imports and exports. From 2010 onwards, the frequency of the required VAT report depends on the annual sales of the firm: firms with annual sales below 25,000 euros are allowed to report annually, firms with sales between 25,000-50,000 euros must report quarterly, and firms with sales above 50,000 euros have to report monthly. Before 2010, all VAT registered firms filed a monthly VAT report. VAT threshold before 2004. In Finland, the VAT liability threshold for firms is 8,500 euros of annual sales. Below this threshold firms are exempt from reporting and remitting VAT. The threshold has remained constant from 1995 in nominal terms. Even though small firms below the threshold are exempted from VAT, they need to pay other taxes and report their overall sales to the Tax Administration for income tax purposes. Before 2004, firms that exceeded the threshold paid VAT for all value-added. This included valueadded on sales from below the threshold. Therefore, exceeding the VAT threshold created a notable jump in VAT liability and the average VAT rate. VAT relief scheme from 2004 onwards.

The VAT rate at the threshold changed in 2004 in Finland,

although the threshold itself remained at 8,500 euros. The VAT reform of 2004 introduced a VAT relief scheme. The VAT relief scheme is applicable for firms with annual sales below 20,000 euros in 2004. The threshold for the VAT relief was increased to 22,500 euros in 2005. The VAT relief decreases remitted VAT such that the average VAT rate increases gradually above the threshold, compared to a discontinuous jump in the average VAT rate before 2004. In practice, this implies that firms only pay VAT for the value added above the VAT threshold. Figure 2 shows the VAT remittance in euros (above) and average VAT rates (below) for different levels of sales (in bins of 100 euros). The Figure illustrates the introduction of the VAT relief region in 2004 and post-2005 in comparison to pre-2003 period for a representative firm that is subject to the standard VAT rate. For illustrative purposes, the representative firm is assumed to have no deductible VAT from purchases, implying that the value added equals the sales of the firm. The Figure shows that the pre-reform system created a salient VAT notch, implying a jump in remitted VAT and the average VAT rate from 0 to 22% at the threshold (standard VAT rate was 22% until 2010 in Finland). After the reform the notch was replaced by a VAT kink, implying gradually increasing remitted VAT and average VAT rate above the threshold. Within the VAT relief scheme, gradually increasing average VAT rate implies an increasing marginal VAT rate above the threshold up to the point in which the average VAT rate equals 22%. This leads to marginal VAT rates between 13-57% within the VAT relief region, which was 8,500-20,000 euros in 2004 and 8,500-22,500 euros from 6

2005 onwards. Figure 2 highlights the striking difference in tax incentives between the two VAT regimes. For example, consider a firm with annual sales equal to 10,000 euros, which therefore exceeds the VAT threshold by 1,500 euros. Before 2004, the average VAT rate on all value added of the firm was 22%. However, after 2004, the average VAT rate on total value added is around 2.5%, which is over eight times smaller than before the reform. Thus the difference of pure tax incentives is very notable, especially for firms barely exceeding the threshold. However, as can be seen from the Figure, the difference between the regimes decreases at larger sales levels.

Remitted VAT and average tax rates before and after the reform

0

Remitted VAT 3000 6000

Remitted VAT

0

5000

10000

15000

20000

25000

30000

20000

25000

30000

Average tax rate (%) 0 5 1015202530

Average tax rates

0

5000

10000

15000 Annual turnover

VAT pre−2003 VAT post−2005

VAT 2004

Notes. The Figure shows the remitted VAT and average VAT rates for a representative firm which is subject to the standard VAT rate in the year in question. For simplicity, the firm is assumed to have no tax-deductible VAT from purchases. This implies that the value added equals sales.

Figure 2: VAT remittance and average VAT rates for different levels of sales before and after the introduction of VAT relief region An additional important detail of the VAT relief is that it is not automatically granted by the Tax Administration. Before 2010, firms needed to apply for the VAT relief using a separate tax form. From 2010 onwards, firms can apply for the VAT relief with the same periodic tax form they use to declare remitted VAT. This can have important implications for the salience of the VAT relief scheme, which we study in more detail in Section 5. Finally, firms that do not exceed the threshold can voluntarily register and pay VAT. There are logical reasons for registering even when it is not necessary. A firm can only deduct the VAT from its purchases if it is registered. For example, voluntary registration could be important for businesses that 7

have large start-up costs. Also, firms below the threshold that have a large share of business-to-business sales have an incentive to register, as the VAT rebate is only possible from purchases of VAT registered firms. Furthermore, VAT registration can enhance the status of the firm, and give the appearance of the firm as a large and trustworthy partner. This can be appealing towards both customers and suppliers, and therefore increase business activity.

3

Methodology

3.1

Bunching at the VAT threshold

We use the bunching methodology presented in Saez (2010) and Kleven and Waseem (2013) to analyze firm responses to the VAT threshold. The baseline intuition behind the bunching approach is that if a discontinuous change in VAT liability at the threshold affects the behavior of firms, we should find an excess mass of firms located at the threshold. We illustrate the incentives related to the VAT threshold using a simplified model of firm behavior. First, we abstract from other taxes than the VAT. In addition, we assume that the firm is owned and managed by a single entrepreneur. As the VAT threshold in Finland is low, most firms around it are single owned. Therefore, it is reasonable to assume that individual owners make the relevant firm-level decisions. However, for conceptual simplicity, throughout the paper we denote that firms respond to the VAT threshold, not individual owners. We assume that a firm has the following profit function

π(s) = (s − zs) − c(s) − Tvat (zs) − [(Tvat (s) − Tvat (zs)) + δ(s∗ )] · 1(s > s∗ )

(1)

where s denotes the sales of the firm. We assume that the output of the firm is sold with an exogenous price equal to unity, which implies that sales directly reflect the output of the firm. zs denotes the linear function of tax-deductible purchases needed to generate s. We assume that 0 ≤ z < 1, which implies that the marginal unit of sales produces positive net income for the firm. We briefly discuss the implications of relaxing this assumption in the end of this subsection. In equation (1), c(s) is a convex function of the cost of effort of the owner. Tvat (zs) denotes the function of VAT paid on purchases, and (Tvat (s) − Tvat (zs)) denotes the VAT function of the value added (salespurchases) of the firm. Finally, δ(s∗ ) represents compliance costs related to the VAT system. The firm is not liable to report and pay VAT below a sales threshold s∗ . Therefore, [(Tvat (s) − Tvat (zs)) + δ(s∗ )] = 0 if s ≤ s∗ , and thus exceeding s∗ creates a discontinuous increase in both remitted VAT and compliance costs. However, above the threshold the firm can deduct the VAT on purchases from the remitted VAT. Below s∗ VAT paid on purchases is not tax-deductible.

8

First, we study the effect of a change in the VAT rate at the threshold. For now we ignore compliance costs, which we will come back to in Section 3.3 below. In Finland, there has been two kinds of changes in the VAT rate at the threshold: a VAT notch and a VAT kink. To start with the VAT notch, consider a VAT schedule where the value added is not taxed until the notch point s∗ . If sales exceed s∗ , the VAT rate will be applied to total value added. Thus the VAT liability jumps discretely at s∗ , as the firm needs to pay VAT also for the sales below s∗ if the threshold is exceeded. More formally, the VAT function in equation (1) in the notch schedule is of the form [Tvat (s) − Tvat (zs)] = [(s − s∗ ) − (zs − zs∗ )] τn + 4T (s∗ ), where τn is the flat VAT rate above s∗ , and 4T (s∗ ) denotes the VAT paid for the value added below s∗ which the firm needs to remit if the VAT threshold is exceeded. In addition, we denote the VAT paid on purchases as Tvat (zs) = zsτn . Next, let us characterize firm decision making within a small sales interval (s∗ − ,s∗ ). First, the marginal cost of additional unit of sales for a firm at (s∗ − ) is

c0 (s) = (1 − z) − τn z

(2)

For a firm locating exactly at s∗ , the cost of additional unit of sales is

c0 (s) = (1 − z) − τn − (s∗ − zs∗ )τn

(3)

where (s∗ − zs∗ )τn = 4T (s∗ ) denotes the fixed amount of VAT remitted on the value added below s∗ once the threshold is exceeded. This discontinuous jump in VAT liability creates a notch to the budget set, which we will further illustrate below. When comparing equations (2) and (3), we find that the firm has a notable incentive not to exceed the VAT threshold, as the marginal cost of additional sales is larger at the threshold than just below it

τn z < τn + (s∗ − zs∗ )τn

(4)

Intuitively, an additional unit of sales is less valuable at the VAT threshold because the firm needs to pay both VAT for the sales above s∗ and for the value added on sales below the threshold. Therefore, it is presumable that firms exceeding the threshold. This behavior would result in many firms locating exactly at the threshold in the sales distribution. This type of bunching behavior below the VAT notch is illustrated in the upper graph of Figure 3. The vertical axis denotes net-of-tax sales, and horizontal axis denotes sales before taxes. The straight blue lines illustrate the tax rates and the budget set. 4T (s∗ ) represents the VAT paid from sales below the threshold once the VAT threshold is exceeded, which creates the discontinuous notch in the budget set. The curvy red lines represent the indifference curves of different types of firms (type A and type B). A fraction of firms originally above s∗ will locate themselves at the VAT threshold after the intro9

duction of the notch. The extent of this bunching behavior depends on the elasticity of the value added with respect to VAT rate, which we will come back to in more detail below. Firms originally at s∗ or below the threshold will not change their behavior (type A firm). In the graph, s∗ + 4s denotes the hypothetical firm with the highest sales to bunch at the threshold (type B firm), which is indifferent between locating at s∗ and sB after the introduction of the VAT notch. In other words, s∗ + 4s marks the last firm bunching at the notch, which we refer to as the marginal buncher. More formally, the ´ s∗ +∆s fraction of firms located at s∗ in response to the notch is denoted as B(∆s) = s∗ h0 (s)ds, where h0 (s) denotes a smooth counterfactual density of sales in the absence of the notch.

10

Indifference curves

Net-of-tax sales

Type B

Type A Slope (1-τn)

∆ T(s*)

sB s*+∆s

s*

sales

Indifference curves Type B

Net-of-tax sales

Type A Slope (1-τk)

s*

s*+∆s

sales

Figure 3: Bunching at a VAT notch (upper graph) and a VAT kink (lower graph) In the VAT kink system, only sales exceeding s∗ are taxed at the VAT rate. The VAT function in equation (1) is then of the form [Tvat (s) − Tvat (zs)] = [(s − s∗ ) − (zs − zs∗ )] τk , where τk denotes the marginal VAT rate above the kink. Now, when comparing a firm just below and exactly at the threshold, we find that the marginal cost of additional sales is again larger at the threshold than below it

11

τk z < τk

(5)

Intuitively, the main difference between VAT notch (equation (4)) and VAT kink (equation (5)) is the size of the change in tax incentives at the threshold. Compared to a VAT notch, a VAT kink produces smaller incentives to respond, as the firm needs to remit VAT only for the value added above the threshold. Therefore, based on pure tax incentives, it is presumable that less firms will bunch at the VAT kink than at the VAT notch. Bunching at the VAT kink is illustrated in the lower graph of Figure 3. Type A firm which is located at s∗ before the introduction of the VAT kink will not respond to the kink, whereas a fraction of firms above s∗ will locate themselves at the VAT kink. As with notches, type B firm in the graph represents the marginal firm with the largest sales (s∗ + 4s) to bunch at s∗ after the introduction of the VAT kink. Figure 4 characterizes bunching at the VAT threshold in the sales distribution. In the Figure, the vertical axis denotes the number of firms and horizontal axis denotes sales levels. The solid blue line represents observed sales distribution, and the dotted red line is the counterfactual density in the absence of the threshold. The excess mass caused by the threshold is presented as a spike in the distribution at s∗ . The missing mass above the threshold is denoted as the area between the counterfactual distribution and the observed distribution within the region (s∗ , s∗ + 4s). Assuming smooth and heterogeneous sales elasticities with respect to the VAT rate across different firms and no extensive margin responses, the observed density gradually approaches the counterfactual density above s∗ . Thus s∗ + 4s represents the firm with the largest sales to bunch at the threshold. Intuitively, the larger the excess mass at the threshold is the further away from s∗ comes the last firm to bunch at the VAT threshold. We discuss the estimation procedure in Section 3.4 below. There are also circumstances in which a firm has no tax incentive to bunch at the VAT threshold. These situations occur when we relax the assumptions of our simplified model. Perhaps the main example of negligible incentives to bunch is substantial VAT paid on purchases stemming from, for example, large start-up costs. In other words, for some firms it could be that Tvat (s) < Tvat (zs) above the VAT threshold, and thus (marginally) exceeding the threshold does not increase tax liability. Second, it could be that z > 1 for some small firms above the threshold in the short run, which indicates no incentives to bunch at the threshold. However, small firms in our data are on average relatively profitable and have notably larger sales compared to overall expenses, which indicates that pure tax incentives to bunch at the VAT threshold exist for a large proportion of small firms in Finland. Overall, as the share of purchases in relation to output varies between different firms, there are firm-specific differences in the incentives related to the VAT threshold. In other words, exceeding the threshold induces VAT payments of different size for different firms. We study this heterogeneity in 12

Number of firms

Excess mass

Observed distribution Counterfactual

Missing mass

s*

s*+∆s

Sales

Figure 4: Bunching at the VAT threshold Section 5. Finally, large purchases related to sales increase incentives to voluntarily register to pay VAT even when sales do not exceed the VAT threshold. VAT registration might enhance the status of the firm and therefore increase business opportunities, which could make voluntary registration appealing even for firms with smaller purchases than sales. Within voluntarily registered firms, there are typically no incentives to bunch at the threshold. Therefore, as there exist firms with no incentives to bunch, it is presumable to observe a positive mass of firms just above the VAT threshold in the empirical sales distribution (Figure 4). We discuss voluntary VAT registration in more detail in the following sections.

3.2

VAT rate elasticities based on observed bunching

We approximate the sales elasticity with respect to the VAT rate at the VAT threshold using a similar approach as Kleven and Waseem (2013). We characterize the elasticity at the VAT notch and VAT kink by relating the sales response of a marginal buncher firm (s∗ + ∆s) to the change in tax liability caused by exceeding the threshold by ∆s. This reduced-form characterization of the sales elasticity offers an intuitive measure to scale the extent of the behavioral response by the change in the VAT rate under different VAT rate schedules. In other words, this approach enables us to approximate the sales elasticity under both the VAT notch and VAT kink regimes. Elasticity at the VAT notch is calculated with the following quadratic formula (for more details, see

13

Kleven and Waseem (2013)) eN ≈ (4s/s∗ )2 /4tN

(6)

where 4tN = [((4s − s∗ ) − (z4s − zs∗ )) + (s∗ − zs∗ )] τn /4s defines the relative increase in VAT payments caused by exceeding the threshold by ∆s. Importantly, when exceeding the VAT notch, the firm needs to pay VAT also for sales below s∗ . Sales elasticity associated with the VAT kink can be written as

eK ≈ (4s/s∗ )2 /4tK

(7)

where 4tK = ((4s − s∗ ) − (z4s − zs∗ ))τk /4s. Compared to the VAT notch, the firm needs to pay VAT only for sales above s∗ within the VAT kink system. However, as Figure 2 in Section 2 above shows, the average VAT rate increases above the VAT kink, implying a smoothly increasing marginal VAT rate. Therefore, τk is not constant in practice, and increases with 4s in the VAT kink system. We take this issue into account when calculating the elasticity. Overall, equations (6) and (7) imply that the change in the implicit marginal tax rate (4tN , 4tK ) is larger at VAT notch compared to VAT kink. In general, this creates larger incentives to bunch at the VAT notch compared to the VAT kink. However, within the Finnish context of increasing average VAT rate above the kink, this difference decreases when 4s gets larger. Nevertheless, assuming similar underlying sales elasticities of firms regardless of the VAT system, we should find larger excess bunching at the VAT notch compared to the VAT kink. Importantly, the sales elasticity formulas above ignore other costs related to exceeding the VAT threshold, such as VAT reporting costs. Therefore, the elasticity estimates with respect to the VAT rate are likely to be overestimated if these compliance costs induce notable sales responses. Thus our elasticity estimates represent the upper-bound of the pure VAT rate responsiveness of small firms. We discuss the implications of compliance costs in more detail below and in Section 5.

3.3

Compliance costs of VAT reporting

In addition to remitted VAT, a firm faces other costs when exceeding the threshold. For example, these costs include reporting and accounting costs. Once a firm becomes liable to pay VAT, it needs to file separate reports on sales and purchases subject to VAT to the Tax Administration. This procedure can be executed by the owner, or she can purchase an accounting service to conduct the VAT reporting for the firm. In addition to these direct costs related to the mechanical filing of periodic VAT reports, the owner faces costs related to understanding the details of VAT rules and regulations. For example, certain types of sales and purchases are subject to different VAT rates, which make the VAT system and VAT reporting relatively complex to comprehend. Furthermore, the firm might need to reprice its products, 14

and is legally required to separate the share of VAT from the selling price in all receipts and invoices. In addition, more extensive reporting could make tax avoidance and evasion more difficult, as firms need to report both sales and purchases in more detail when the threshold is exceeded. This can be seen as a cost induced by the VAT system for firms that systematically underreport their income for the tax authority. In our simplified framework, we assume that a firm faces a fixed compliance cost when exceeding the VAT threshold. The key insight behind the fixed compliance cost is that it induces incentives not to exceed the VAT threshold even if pure tax incentives are small. To illustrate this issue, we add compliance costs to equation (5) when studying the difference in marginal costs of additional unit of sales below and at the VAT kink,

τk z < τk + δ(s∗ )

(8)

In Equation (8), δ(s∗ ) represents the discontinuous increase in compliance costs when the threshold is exceeded. Intuitively, a fixed compliance cost above the threshold implies that even when z is close to one, firms still bunch below the threshold if the compliance cost is large enough. Furthermore, even firms with Tvat (s) < Tvat (zs) bunch below the threshold if compliance costs are large. Figure 5 illustrates the effect of fixed compliance costs in the VAT kink system. In the Figure, compliance costs create a notch to the budget constraint in a similar manner as VAT remitted from sales below s∗ in the VAT notch system. Figure 5 describes the effect of compliance costs within the VAT kink system, but the effect is similar also within the VAT notch system, in which the compliance cost increases the discontinuous drop in profits at the threshold. Therefore, compliance costs increase the incentives not to exceed the VAT threshold under both regimes.

15

Indifference curves

Type B

Π(s) Type A

Slope (1-τk)

δ(s*)

s*+∆s

s*

sales

Figure 5: Bunching at a VAT kink with compliance costs The significance of compliance costs is likely to be large for smaller firms. It is plausible that compliance costs include a sizable component that does not depend on the size of the firm. This implies that the share of this cost relative to the size of the economic activity of the firm is larger for smaller firms. This would lead to pronounced bunching at the VAT threshold especially among small firms.

3.4

Empirical estimation of excess bunching

The excess mass of firms at the VAT threshold is estimated by comparing the actual density function around the threshold to an estimated smooth counterfactual density. The counterfactual density function describes what the distribution of sales would have looked like without a change in VAT liability at s∗ . We follow the methods in Chetty et al. (2011) and Kleven and Waseem (2013) to estimate the counterfactual density. The counterfactual density is estimated by fitting a flexible polynomial function to the observed distribution, excluding an area around s∗ from the observed distribution. First, we recenter income in terms of s∗ , and group firms into small sales bins of 100€. We estimate a counterfactual density by regressing the following equation and excluding the region around the threshold [sL , sH ] from the regression

cj =

p X i=0

βi (sj )i +

sH X

ηi · 1(sj = i) + εj

(9)

i=sL

In equation (9), cj is the count of firms in bin j, and sj denotes the sales level in bin j. The order

16

of the polynomial is denoted by p. Thus the fitted values for the counterfactual density are given by Pp cˆj = i=0 βi (sj )i . The excess bunching is estimated by relating the actual number of firms close to the threshold within (sL , s∗ ) to the estimated counterfactual density in the same region. We calculate excess bunching as Ps∗ ˆj ) i=sL (cj − c ˆb(s∗ ) = P s∗ ˆj /Nj i=sL c

(10)

where Nj is the number of bins within [sL , s∗ ]. One important issue when estimating the counterfactual density is how to determine the excluded region [sL , sH ]. As in earlier literature, we determine the lower limit sL based on visual observations of the sales distribution. Intuitively, sL represents the point in the sales distribution where the bunching behavior begins, i.e. the density of firms begins to increase. We follow the approach of Kleven and Waseem (2013) to define the upper limit. We determine sH Ps∗ such that the estimated excess mass ˆbE (s∗ ) = ( i=sL cj − cˆj ) equals the estimated missing mass above PsH the threshold, ˆbM (s∗ ) = ( s>s ˆj − cj ). We apply this convergence condition by starting from a small ∗ c value of sH and increasing it gradually until ˆbE (s∗ ) ≈ ˆbM (s∗ ). This definition for sH denotes the upper bound of the excluded range, and thus the lower bound for estimated excess bunching (Kleven and Waseem 2013).2 Theoretically, this condition states that firms who bunch at the threshold come from the region directly above it. In addition, the convergence condition implies that we can intuitively define the sales response of the marginal buncher firm using the estimated excess mass and the upper limit sH . This enables us to approximate sales elasticities with respect to VAT rate for both the VAT kink and the VAT notch by relating the marginal sales response to the implied change in the VAT rate. When calculating the elasticity estimates, we use the bin-level average of the value added of the marginal buncher firm to define the VAT paid when exceeding the threshold by 4s (see equations (6) and (7) above). Following Chetty et al. (2011) and Kleven and Waseem (2013), the standard errors for all the estimates are calculated using a residual-based bootstrap procedure. We generate a large number of sales distributions by randomly resampling the residuals from equation (9) with replacement, and generate a large number of new estimates of the counterfactual density based on the resampled distributions. The bootstrap procedure takes into account the iterative process to determine sH . Based on the bootstrapped counterfactual densities, we evaluate variation in the estimates of interest. The standard errors for each estimate are defined as the standard deviation in the distribution of the estimate. 2 Kleven and Waseem (2013) apply this convergence condition to estimate the counterfactual density around individual income tax notches in Pakistan. For individual tax rate kink points in Denmark, Chetty et al. (2011) determine the upper limit visually, and then iteratively adjust the counterfactual density above the kink point such that it includes the excess mass at the kink. This makes the estimated counterfactual density equal to the observed density. These procedures are intuitively similar, but the convergence method of Kleven and Waseem (2013) typically provides a smaller estimate for excess bunching. In addition, the convergence method provides a more justified approach to define the upper limit of the excluded region when estimating the counterfactual density.

17

4

Data and descriptive statistics

4.1

Data

Our data are from the Finnish Tax Administration and covers the period 2000-2011. The data contain all businesses that operate in Finland, including firms that are registered to pay VAT and firms that are not included in the VAT register. The data include all information needed for tax purposes, such as sales, number of employees, taxable profits, total assets and organizational form. Importantly, data include accurate information on total sales also for firms below the VAT threshold. This enables us to analyze the effect of the VAT threshold on the distribution of sales. In addition, we can link owner-level variables to the firm-level data, such as personal taxable wage and capital income of the main owner of the firm. In the following analysis, we exclude all firms that operate in sectors that are not subject to VAT, such as financial and insurance activities, letting and operation of dwellings, education, and health and social work activities. Since these firms are not liable to pay VAT, it is not relevant to include them in the analysis of behavioral responses to the VAT threshold. In addition, we restrict the sample to include only firms with annual sales below 20,000 euros, since these firms can be though of as being affected by the current VAT threshold. Furthermore, we exclude firms that are taxed by assessment of the Finnish Tax Administration, as tax record information based on assessment does not provide evidence of behavioral choices of firms in response to the VAT threshold. The most common reason for assessed taxation is that a firm has not returned its tax forms in time.

4.2

Characteristics of small firms and their main owners

Table 1 shows the descriptive statistics of small firms and their main owners in Finland. The upper panel of the Table shows firm-level statistics. From firm-level statistics we can unsurprisingly see that most of the firms in our sample do not have any employees, and have relatively low taxable profits, expenses and assets. The Table shows that sole proprietor is clearly the most common organizational form among small firms in Finland, as approximately 70% of small firms in our sample are sole proprietors. One fifth of the firms in the sample are privately-held corporations, and 9% are partnership firms. In addition to firm-level characteristics, the data set includes tax record information on the main owner of the firm, which are presented in the lower panel of Table 1. Importantly, it is presumable that the underlying nature of small firms is relatively heterogeneous among different owners. Despite the relatively low level of sales, some small firms are the main source of income for their owner. In contrast, some small firms are simply side businesses, and the main source of income for the owners of these firms is not the taxable profit of the firm. Overall, the average total income of the main owner (the sum of taxable wage and capital income) is 18

relatively low, approximately 15,400 euros. However, there is heterogeneity with respect to the income level of the owner. Table 1 shows that over 50% of the owners in our sample have low personal taxable income (below 10,000 euros). Approximately 20% of the owners have personal income between 10,00020,000 euros, and roughly 25% of the owners have personal income above 20,000 euros. This implies that at least for 3/4 of the owners, the income from the firm represents an important share of personal income. However, there is a non-negligible number of owners for whom the firm appears to represent a side business activity. In order to more specifically describe the role of the firm in terms of personal income, we define owners as full-time entrepreneurs if the annual turnover of the firm is larger than the total income of the owner. It appears that most of the owners fulfill our definition of a full-time entrepreneur, as over 60% of all main owners in our sample have more annual turnover in their firm than they have total personal income. Therefore, it seems that side businesses do not comprise the majority of our sample. Nevertheless, potential heterogeneous responses to the VAT threshold could be important in terms of interpreting the results. For example, the implications of behavioral responses could be different if only side businesses respond to the threshold. In Section 5.1 we study responses to the VAT threshold separately for different types of firms and owners. Firm-level statistics Sales

No. of empl.

Profits

Expenses*

Assets

Sole proprietors

Corporations

Partnerships

Mean

8,942

.112

1,688

1,934

6,372

.688

.224

.0878

Median

7,962

0

778

362

155

1

0

0

SD

5,355

.451

4,974

3,291

21,042

.463

.417

.283

588,505

586,371

581,623

584,785

578,642

588,505

588,505

588,505

N

Owner-level statistics Age

Female

Total inc.**

Total inc.