The tax bases are generally narrow in developing countries, as various institutional deficiencies make tax enforcement difficult. The personal income tax, a dominant source of revenue in industrial countries, is of relatively minor importance in developing countries. Most corporate taxes are collected from large manufacturing corporations. In this regard, the value-added tax (VAT), a broad-based consumption tax with relatively less tax resistance, is a particularly attractive source of revenue in these countries.
While the base for the VAT is broad relative to other taxes, its coverage is still limited in developing countries. This is due to widespread exemptions and special schemes that have been introduced for distributive and other social policy objectives. Basic necessities and meritorious goods mostly escape the VAT net, and goods supplied by small businesses tend to receive lenient tax treatment. Such exemptions not only erode the tax base, but also cause various distortions. Thus, the primary task for most developing countries with VATs is to expand the base in a way so as to raise revenue as well as reduce distortions.
One possible option is to bring financial services into the VAT net. Currently, most financial services are exempt from the VAT in most countries having this tax. There are few economic and social justifications for exempting financial services, however. In theory, a VAT is a tax on the incomes of all primary factors that enter into the production of final consumption goods. Neutrality requires that income accrued on labor and capital used in providing financial services should be taxed just as nonfinancial goods and services are. In addition, exempting financial services from taxation would not mitigate the distributive concerns related to the VAT, since financial products and income associated with them are often thought of as progressive sources of revenue.
The most convincing case for exempting financial services is made based on the operational difficulties of identifying and valuing the appropriate tax base when payments for financial services take non-explicit forms. Often, charges for financial services are buried in margins, such as the differences between interest on loans and deposits or between the rates for buying and selling of foreign currencies. Fees can also be included in the purchase prices of financial products, instead of being explicitly charged as fees. In such cases, it is difficult to identify and value the services on a transaction-by-transactions basis for the purpose of applying the credit-invoice VAT.
For such administrative reasons, financial services provided for margin-based and other implicit charges are mostly treated as exempt supplies under the VAT. Furthermore, even most explicit fee-based financial services are exempt from taxation in most countries having a VAT. This is because financial institutions may otherwise have incentive to substitute margin-based charges for explicit fees. As a result, most financial services are exempt from the VAT, regardless of whether they are rendered for explicit or implicit fees. In this way, major conceptual and administrative issues surrounding VAT application to financial services have been circumvented in most countries having a VAT.
However, it has turned out that the exemption approach generates various economic distortions. Most notable of these is tax cascading. Financial institutions that supply exempt financial services are denied input tax credits for their taxable purchases made to supply these services. When exempt financial services are purchased by taxable businesses, tax on tax or tax cascading occurs. Such over taxation of financial services can erode the competitive advantages of domestic financial institutions relative to foreign providers of similar services. It can also cause a self-supply bias for domestic service providers seeking to reduce unrecoverable input taxes. In addition, financial institutions with both taxable and exempt products must calculate their allowable input tax credits based upon the compositions of their supplies and the uses of purchased inputs. The associated administrative and compliance costs could be large, and thus partially offset the operational advantages of the exemption approach stemming from the fact that margin-based services do not need to be valued. Changing market conditions, due for example to deregulation and globalization, have been an additional source of pressure on the current system.
To address the limitations of the exemption approach, a number of alternative methods have been suggested. However, none of these new methods seems dominating in terms of revenue effects, economic distortions, and administrative and compliance burdens. Some have been successfully tried in specific country settings, but their applicability in other countries are yet to be tested.
While the difficulties with the VAT treatment of financial services have been mostly discussed from the perspective of industrial countries, developing countries with a VAT or plans to introduce them face similar problems. Developing countries also have additional considerations to take into account, e.g.:
- Their market structures and tax systems do not resemble those of industrial countries.
- The focuses of their tax policies can also be different, as growth objectives typically dominate distributive concerns in the early stages of development.
The optimal solutions for taxing financial services in developing countries, therefore, may deviate from the ones chosen for industrial countries.
This paper reviews approaches to taxing financial services under a credit-invoice VAT regime, with particular attention to their applicability in developing countries. Section 2 examines the revenue, efficiency, and administrative implications of alternative treatments of financial services within a VAT, while Section 3 discusses policy options available to developing countries. Section 4 contains a brief conclusion.
2. Alternative Treatments of Financial Services Under a VAT
This section provides a brief assessment of the current exemption approach and its major alternatives to taxing financial services under a VAT. These methods are compared to each other in terms of revenue effects, distortions, and administrative and compliance costs.
The Exemption Method
Most countries with credit-invoice VAT systems have adopted the exemption approach. While the scope of exemption varies across countries, margin-based services are exempt, certain explicit fees are taxed, and exported services are zero-rated. Since the EU mandated this approach in the Sixth Directive, with the aim of harmonizing the VAT among member states, most other countries with a VAT system have adopted this model.
As noted earlier, financial institutions that supply exempt financial services are denied input tax credits for their taxable purchases made to supply these services. This has been considered a major limitation of the exemption approach. When exempt financial services are purchased by final consumers, these services are under-taxed relative to other taxed goods and services. On the other hand, when exempt financial services are purchased as inputs by businesses which make taxable sales, denial of tax credits for inputs used in supplying exempt financial services can cause cascading or over taxation. In either case, the exemption method deviates from full taxation, which becomes the source of various distortions.
Facing over taxation of their exempt supplies, financial institutions may have an incentive to supply inputs internally. Such self-supply practices may help lower their tax burdens, but they may also generate inefficiencies by distorting production decisions. In addition, domestic suppliers may lose some competitive advantage vis-à-vis foreign suppliers providing similar services, if the products of the latter do not include the VAT elements. As world financial markets are becoming more integrated, the practice of foreign outsourcing may pose a serious threat to domestic providers of financial services.
The administrative simplicity of the exemption approach can also be compromised by the burden of allocating input tax credits for financial institutions with both taxable and exempt services. The associated administrative and compliance costs could potentially be large. In addition, tax planning opportunities can occur related to changes in use of fixed assets between taxable and exempt supplies. Rules concerning change-of-use entail extra compliance efforts by the relevant businesses.
The revenue implications of the exemption method are not clear, since financial services are partly under-taxed and partly overtaxed relative to full taxation. Moving to full taxation could increase revenue to the extent that under-taxation outweighs over taxation. This is an empirical question, concerning which the related evidence is sparse. Furthermore, some countries impose ad hoc taxes on financial services, to make up for the revenues foregone under the exemption approach.
Taxation of Explicit Fees
Technically, financial services rendered for explicit fees can be taxed just like nonfinancial goods and services, since their values can be measured on a transaction-by-transaction basis. If explicit fees are taxed, input tax credits on taxable business inputs can be recovered by financial institutions providing explicit fee-based services. Since the scope of exemption is limited to services with margins or implicit fees, this approach can reduce cascading and other distortions associated with the EU type of exemption. Since these days the majority of financial charges take the form of explicit fees and commissions, this method can be an attractive choice in modifying the current system.
However, this scheme creates incentives for financial institutions to substitute nontaxable implicit fees for taxable explicit fees, or to bundle them together so that service charges can be buried in nontaxable margins. The substitutability between margins and explicit fees may vary across countries, depending upon their market conditions and relevant rules. As financial markets become deregulated and globalized, financial services tend to be unbundled and more individualized. As financial services become unbundled, financial charges are less likely to be buried in margins or to take other implicit forms. Such market pressures can be less severe in developing countries, since their financial markets are generally regulated and concentrated. However, tax policy needs to be forward-looking. In addition, where the effective VAT rates are not high, as in many developing countries, substitution of margins for fees may not be worth the associated transaction costs. Accordingly, bringing financial services with explicit fees into the VAT net appears a more feasible option than previously thought, in both industrial and developing countries.
The option of taxing explicit fees might reduce a significant portion of the distortions associated with the exemption approach, but it cannot remove them completely. The issue of determining creditable input taxes still remains for financial institutions having both exempt and taxable sales. The substitution between margins and fees can still be a problem.
If reducing economic distortions is the primary policy concern, financial services can be zero-rated instead of being exempt. This will allow financial institutions to recover full tax credits for taxable inputs used in supplying their exempt services. In this case, there will be no cascading or tax-driven self-supply biases, while competitiveness pressure from abroad can also be reduced. Since all financial services are now taxable, the issue of input credit allocation does not arise either.
In a developing country where the size of the underground financial market is large, zero-rating could further reduce inefficiencies by inducing businesses into the formal sector. By eliminating taxes on intermediate goods in the value-added chain, zero-rating can lead financial institutions to lower the prices of financial products sold in the formal sector. In contrast, the providers of financial products in the informal sector cannot claim credits for taxable inputs produced in the formal sector, and thus come to face competitive disadvantages. This point has not been adequately discussed before, but it might have important efficiency implications in many developing countries.
Zero-rating has its own share of limitations, however. Under this approach, there will be no tax burden on financial services provided to final consumers and businesses making exempt sales. While eliminating the over-taxation of exempt supplies to businesses with taxable sales, zero-rating aggravates the under-taxation problem associated with supplies to households and businesses with exempt sales. Often, this aspect of zero-rating is cited as a source of distortion between financial services and nonfinancial goods and services. In general, however, financial and nonfinancial products are from the standpoint of consumers not quite substitutable.
A more serious concern about the zero-rating approach is its revenue costs. The size of the revenue foregone will depend upon the institutional details and the behavioral responses by taxpayers. One possible remedy for this problem is limiting zero-rating to margin-based services while taxing fee-based services, as discussed earlier. In this way, cascading can be eliminated with relatively small revenue costs. Of course, this approach may as previously discussed cause substitution between margins and explicit fees.
Another compromise is to limit zero-rating to business-to-business supplies of financial services, as has been attempted in New Zealand. The recipient of financial services must make mostly taxable sales in order for the service provider to qualify for this scheme. Financial supplies made to consumers and businesses with exempt sales are thus ineligible for zero-rating. Similar to the case with the option mentioned above , cascading can be removed without large revenue costs. The problem with this approach is the potentially large administrative and compliance burdens. To apply for zero-rating, a business must have information about the registration status of the recipient and her ratio of taxable-to-total sales.
Exemption with Input Tax Credits
If the main policy aim is to reduce cascading, reducing the scope of exemption for financial services supplied to taxable businesses (i.e. treating more financial services as taxable supplies) could be the appropriate choice. The preceding two approaches – taxation of explicit fees and zero-rating – are mainly designed to address this issue. Singapore and Australia have adopted yet another approach to mitigate the problem of blocked input credits, allowing input tax credits for exempt supplies used as business inputs in certain situations.
In Singapore, financial institutions can claim input tax credits under either the “special method” or the “fixed input recovery method.” Under the special method, exempt financial supplies made to registered persons are treated as taxable supplies at a zero rate. This approach is similar to the one used in New Zealand (i.e. zero-rating of business-to-business supplies), although the exact details differ. Since financial institutions have to separate eligible services out of total services to receive this benefit, they face compliance burdens in allocating input tax credits based upon the type of services. In contrast, the fixed input recovery method allows a financial institution to claim input credits for a fixed percentage of total input taxes. The percentage depends upon the type of financial institution involved.
While both methods can reduce cascading, the fixed input recovery method is administratively more simple than the special method, since there is no need to distinguish among service recipients based on their registration statuses. However, the special method, which zero-rates business-to-business transactions, is more faithful to the principle of full taxation. In developing countries where tax administration is relatively weak, the fixed input recovery method can be an attractive first choice. In countries with mature VAT systems, variants of the zero rating approach can be serious options.
Australia also allows financial institutions to claim partial input tax credits for exempt supplies, although the focus is more on reducing the self-supply bias. For a defined list of services purchased by financial institutions making exempt sales, 75% of GST paid on these inputs can be recovered. This method is comparable to the fixed input recovery method used in Singapore, although the latter is more flexible in that the recovery rate varies across industries.
Unless drastic reform is planned, the preceding approaches appear the most promising and feasible choices for countries contemplating modifications of their current exemption schemes. There are, however, other methods more in line with full taxation of financial services. The cash-flow approach is a conceptually simple and superior method of eliminating all of the distortions associated with the current exemption practice. As discussed earlier, margin-based services are not easily valued on a transaction-by-transaction basis, so that it is difficult to apply the credit-invoice VAT scheme to them. Use of the cash-flow method can overcome this problem, since all transactions can be identified on transaction-by-transaction and customer-by-customer bases. Unlike with the current treatment of financial services, there will be no problem in allocating input tax credits to individual transactions under this method. Although it is theoretically superior, no country has so far adopted this method. This is because the associated administrative and compliance costs could potentially be quite high.
While virtually all countries with VATs use the subtractive credit-invoice method, the measurement of value added can be done on an addition basis as well. Under the addition method, the sum of wages and profit constitutes the value added. In this way, the full value of financial services can be taxed. Israel currently uses a variant of this approach. However, while this method provides an administratively convenient way of taxing a significant proportion, if not the full value, of financial services, it is not free from cascading and other distortions.
To compensate for revenues foregone from exemption of financial services under their VATs, some countries impose ad hoc taxes on gross receipts of financial institutions. This method is simple but leads to overshooting of the full taxation of the value added. For example, the gross interest on a loan includes the cost of the funds in addition to the value of financial intermediation. Under this method, administrative and revenue gains are made at the cost of efficiency as well as horizontal equity vis-à-vis other goods and services subject to a credit-invoice VAT. Low-margin-high-volume businesses will particularly suffer under this approach.
3. Policy Options for Developing Countries
Tax policy toward financial services cannot be separated from the overall tax system. Tax instruments can be either combined or substituted in order to meet specific policy goals. The economic and political structure can impose constraints on the availability of tax tools. While revenue, efficiency and equity considerations are relevant to any discussion of tax policy, the specific details vary from place to place, and especially between developed and developing countries. The VAT treatment of financial services has received considerable attention in the EU and other industrialized countries, but its applicability in developing countries should be assessed with caution. This section examines some key factors underlying the tax structure, and their implications for the taxation of financial services in developing countries.
The Revenue Implications of Financial Markets
The structure of taxation is influenced by the nation’s tax policy objectives, economic structure, and enforcement capacity. In the early stages of economic development, growth objectives are typically a dominant force shaping the tax system. Raising of revenues to finance public infrastructure and education has been an integral part of the growth strategies in developing countries. Due to enforcement problems with the self-employed and with smaller firms, however, their income and consumption tax bases are narrow. The personal income tax, a dominant source of revenue among rich countries, is of relatively minor importance in developing countries. Corporate taxes are paid primarily by large manufacturing firms. Taxes on consumption and trade typically account for major proportions of the tax revenues, although the importance of the latter has been declining as a result of trade liberalization. While efforts have been made to adopt broad-based consumption taxes like the VAT, excises remain an important form of consumption taxation in most developing countries. Many of these excises are earmarked taxes aimed at mitigating tax resistance.
Difficulties with tax enforcement in developing countries have a lot to do with the lack of information available on the earnings of firms and individuals. The government is very unlikely to have exact knowledge of the taxable sales and incomes of businesses operating in the informal sector or the cash economy. Tax enforcement hinges critically on the use of the financial sector by firms and individuals, since cash-only transactions leave no paper trail necessary for the collection and checking of relevant information. The design of a country’s tax system is likely to be influenced by the size of its cash economy, since unobservable transactions are not taxable.
The informal sectors are typically much larger in developing than in industrialized countries. In particular, informal financial markets have contributed importantly to investment financing for small businesses, since such businesses tend to rely more on cash transactions. In contrast, the formal banking sectors, largely dominated by the state-controlled banks, have been more focused on state-owned enterprises or large manufacturing corporations. Therefore, one immediate task for base broadening in developing countries could be development of their financial markets and provide better financial services to households and firms. Once individuals and businesses realize that the benefits of using financial markets exceed the costs (both the financial service charges and the taxes stemming from financial service use), they will make greater use of the formal financial sector.
The base-broadening effects of financial market development are not limited to the reduction of evasion possibilities due to shrinking of the informal sector. The government can also save tax expenditures. Tax preferences have often been criticized as a source of distortions and revenue losses. In an economy with large evasion pressures, however, the government might be using these incentives, along with certain nontax benefits, as means of preventing erosion of the tax base into the informal sector. The government might well perceive these base protection effects as large enough to justify the associated costs. Examining Korea’s tax structure, Jun (2006) argues that many of the tax incentives provided to small businesses in Korea might have served not so much the purpose of promoting the targeted activities as that of keeping taxpayers from disappearing into the informal sector. Generous nontax compensation has also been available for heavily-taxed large corporations in Korea. The above observation would seem to imply that if the informal sector shrinks as a result of improved financial services, it could be easier to reduce tax preferences aimed at fending off evasion pressures. In this case, a developing country government could more readily adopt the traditional ‘rate-cut-base-broadening’ approach.
In summary, two important channels have been identified here, through which financial market development will lead to revenue and efficiency gains. First, improved financial services could bring more businesses into the formal sector, making tax enforcement easier. Second, shrinking of the informal sector could lead to government saving on tax expenditures related to evasion pressures.
The Costs of Taxing Financial Services
The total opportunity cost of providing a dollar of public service is the tax dollar plus associated costs. The costs of taxation – administrative, compliance and efficiency costs – are often ignored in tax policy debates, since they are not visible. However, they are real costs to the economy, in that they divert resources from their most efficient uses. Taxes with high compliance costs will likely cause evasion or resistance motives. Taxes with high efficiency costs will be detrimental to economic growth.
The financial sector is not just another industry. It links the savings decisions of households and the investment decisions of businesses. Development of financial markets and intermediaries can accelerate economic growth by reducing information, transaction and enforcement costs. Through its linkages with the other sectors, disturbances in the financial sector can easily spread through the economy. Thus, the proper functioning of its financial markets is an essential test of the growth potential of an economy.
In this respect, the over taxation of financial services under the exemption approach could be very costly. This is particularly true in developing countries where financial markets are at the emerging stage and the cash economies are still large. Emerging markets are also likely to be regulated and concentrated. Overall, financial markets in developing countries are already suffering various inefficiencies, while their financial institutions face competitiveness pressures from abroad. In this situation, financial sector taxation could be very costly, since tax-induced inefficiencies tend to be much larger in the presence of existing distortions.
Administrative and compliance costs are generally large in developing countries, due to the complexities of their tax systems and their weak administrative capabilities. Their unstable macroeconomic environments can also give rise to uncertainties about tax liabilities, since tax codes in developing countries are not usually indexed. One troublesome aspect of compliance costs is their regressive nature. Smaller businesses find the compliance burden to be relatively more painful, which might be a reason causing them to avoid use of the formal financial sector.
Based on the preceding discussion, several guidelines for taxing of financial services in developing countries can be derived. While the standard prescriptions suggested in industrialized countries represent a useful starting point, there will be additional factors to consider:
- Concerning the revenue potential, emphasis needs to be placed on the long term implications of improved financial markets, rather than on short term revenue gains. This is especially relevant to countries with large informal sectors. Financial market development will reduce the size of the informal sector. This will make tax enforcement easier and allow the government to reduce tax subsidies designed to prevent tax base erosion. Of course, financial market maturation will accelerate economic growth, which in turn has positive revenue implications.
- As for the costs of taxation, reducing the efficiency costs should receive primary attention in the introduction and modification of financial services taxation. Distortionary taxes could retard the development of financial markets and economic growth, without resulting in large revenue gains. In general, the efficiency costs of taxation are estimated to be much larger in developing countries, due to their more narrow tax bases, than in industrial countries.
- Simple methods are preferable, in order to reduce administrative and compliance costs. High compliance costs could not only push firms into the informal sector but also affect taxpayers in a regressive manner.
- While the full taxation of financial services requires elimination of both over taxation and under-taxation, priority in developing countries should be given to the former. Tax cascading could be very costly in respect of the financial sector’s competitiveness. In contrast, the costs associated with under-taxation of financial services are mostly short-term revenue losses.
- Regardless of whether financial services have been overtaxed or under-taxed, lowering the tax burdens on them can help financial institutions in the formal sector maintain competitive edges against foreign competitors as well as businesses operating in the informal sector. Zero-rating of even financial supplies made to final consumers or exempt businesses (i.e. non-taxation of these products) could lead to net efficiency gains if the informal sector is very large.
- Ad hoc taxes need to be replaced by a broad-based consumption tax like the VAT. A revenue neutral switch would reduce the resultant distortions in the financial markets.
Considering these guidelines, the existing approaches to taxation of financial services are assessed as follows:
- The exemption approach creates cascading and the self-supply bias, but it is administratively simple and allows for the collection of some revenue. Thus, it is a compromise solution that has been adopted in most industrialized countries with a VAT. In developing countries, however, competitiveness distortions associated with tax cascading are likely to be more pronounced than in industrial countries. Considering the revenue implications of developing financial markets, methods with smaller distorting effects appear to be preferable.
- A serious option could be to extend the VAT to financial services provided for explicit fees by financial intermediaries acting as agents, such as brokers and dealers, as under the Singapore VAT. This approach will have a base-broadening effect as well as reducing many of the distortions associated with the exemption method. Under this method, there might be an incentive to substitute margins for fees. However, as financial markets are deregulated and globalized, financial services tend to be unbundled and more individualized. Under this trend toward disintermediation, financial charges are less likely to be buried in margins. While financial markets are generally regulated and concentrated in developing countries, tax policy needs to be forward-looking. In addition, where the effective VAT rates are not high, as in many developing countries, substitution of margins for fees may not be worth the associated transaction costs.
- For margin-based services, some variants of the Singapore methods of allowing input tax credits or New Zealand’s zero-rating of business-to-business supplies could be used. These methods can reduce cascading while the operational simplicity of the exemption approach is maintained. In developing countries where tax administration is relatively weak, the fixed input recovery method can be an attractive first choice. In countries with mature VAT systems, zero rating appears to be a desirable solution, since it could offer clear efficiency gains relative to the exemption approach. If the revenue costs are worrisome, zero-rating can be limited to margin-based services, while explicit fees are treated as taxable. Another option could be zero-rating of business-to-business supplies only, as done in Singapore and New Zealand.
- A more drastic approach with respect to zero-rating is to extend it even to financial services provided to final consumers and businesses making exempt sales. This approach has been shunned, since it could be a source of distortion between financial services and nonfinancial goods and services. In general, however, from the standpoint of consumers, financial and nonfinancial products are not quite substitutable. Furthermore, in countries where the sizes of the underground markets are large, zero-rating of financial services provided to consumers or small businesses with exempt final products might result in net efficiency gains by inducing their greater use of the formal financial sector.
- Cash flow or addition methods are not recommendable, due to uncertainties related to the costs of taxation.
Which of these recommendations should be given more weight depends upon the configuration of the existing tax system as well as the economic structure and market conditions in the country concerned.
Financial services could be an attractive source of tax revenue. Their tax base is potentially large and can be expected to expand. This paper has assessed alternative approaches to taxing financial services under a VAT, from the perspective of developing countries.
The revenue implications of developing financial markets and the benefits of reducing tax distortions are stressed. Considering competitiveness pressure, over taxation of financial services is likely to be much more costly than under-taxation. Distortions associated with the informal sector imply that premature taxation of the financial sector could result in high revenue and efficiency costs.
The arguments presented in this paper indicate that taxation of explicit fees combined with zero-rating of margins can be a promising solution in the long run. Some variants of the methods used in Singapore and New Zealand could be alternative choices, depending upon the current tax structure and market conditions.