India follows a version of the `classical system' of dividend taxation, where companies pay corporate tax of up to 34.61% on their income and a dividend distribution tax ( "DDT" ) of 20.47% on distributable post tax profits.
In the last week of February, the Finance Bill, 2016 ("Finance Bill" or "Bill") introduced an additional dividend tax ("ADT") over and above these levies. The ADT applies for resident individuals, partnerships and trusts if they receive dividends in excess of Rs.1 million in a year. The rate is 10% calculated on a gross basis.
In this article, we examine the ADT and DDT from the wider lens of taxation of corporations and dividend tax policy in India.
We start by analysing the economics: the rate applied, arguments for or against the levy and its impact on different categories of actions. We go on to legal analysis: where does the new provision fit within the legal framework, how is it drafted and will the legal form enable the levy to meet its economic objective?
My primary argument is that dividend tax policy in India is based on insubstantial economic rationale in how it determines tax rates and is fraught with legislative inconsistencies. We need to think more deeply about the rate and method of taxing corporate profits and dividends, and urgently address the legislative issues with the ADT and DDT, irrespective of whether or not we reevaluate our dividend tax from an economic perspective.
The Economic Critique
The economic analysis of taxation of dividends focuses on three questions:
- What is the appropriate aggregate tax rate for profits earned through a corporate vehicle?
- How should the rate derived under (a) be split between undistributed corporate profits at the company level, versus distributed corporate profits?
- What structure should be adopted for application of the rate split derived at under (b), particularly in relation to distributed corporate profits?
Let's start with the first two questions. What is the appropriate aggregate tax rate for profits earned through a corporate vehicle? How should the rate derived under (a) be split between undistributed corporate profits at the company level, versus distributed corporate profits?
When modern day corporate income tax systems were first introduced, one of their primary objectives was to act as prepayment of personal income taxes due by the shareholders, also referred to as the "gap-filling" function (Bird, 2002). Therefore, when corporate tax was levied, the company was chosen as the taxable unit in order to prevent potentially indefinite deferral by the individual taxable units i.e. the shareholders who existed behind the corporate veil.
To put this differently, corporate tax systems were intended to maintain neutrality between individuals investing through a company or directly, by treating both sets of persons similarly. The implications of this were twofold: firstly, profits were intended to be taxable irrespective of whether they were generated individually or through a company. Secondly, the aggregate rate on profits received through a corporate vehicle was intended to approximate the personal slab rate to the extent possible.
Table 1 provides a flavour of international experience with corporate and personal rates. This shows that the dividend tax rate is often structured in a manner that adjusts for the difference between the effective corporate tax and personal tax rates, so as to achieve neutrality between an incorporated versus unincorporated structure.
|Country||Highest personal tax||Highest corporate tax||Dividend tax band|
|Australia||45%||30%||Credit allowed for taxes paid at company level|
|UK||45%||20%||7.5 - 38.1%|
|Greece||42%||29%||Dividends from Greek domestic companies are exempt|
|Canada||33%||38% subject to a federal tax abatement of 10% on domestic Canadian income||Credit allowed for taxes paid at company level|
|Brazil||27.5%||34%||Dividends from Brazilian domestic companies are exempt|
What structure should be adopted for application of the rate split derived at under (b), particularly in relation to distributed corporate profits?
The third question relates to the specific issue of designing dividend taxes. As shown above, most countries apply some form of corporate tax on undistributed corporate profits. In addition, most provide some relief for dividends, either by taxing them at a lower rate and/or allowing credit for corporate tax. This calls for a dividend tax framework which will bridge the divide between corporate and personal tax rates. The following mechanisms are the design choices:
- Reduced rate of tax: This method applies a lower rate of tax to dividends than the maximum marginal rate applicable to individuals. Japan (personal income tax rate of 5-45% & dividend tax at 20%), China (personal income tax rate of 3-45% & dividend tax at 20% on 50% of dividend income) and South Africa (personal income tax rate of 18-40% & dividend tax at 15%) are examples of countries adopting this system.
- Imputation system: This method allows shareholders the benefit of corporate taxes paid by the company. Since imputation credit mechanisms are complex to administer, countries apply them either wholly or partly depending on what they find workable. Australia and Canada are examples.
- Exemption system: Dividends are exempted from personal tax, on the basis that corporate taxes have already been paid. The exemption can either be whole or partial. The Indian dividend tax system is not an exemption system even though it exempts shareholders, because it still imposes a DDT at the company level. True exemption systems, such as the ones followed in Greece and Brazil, allow for an exemption when distributions are made out of post-tax profits.
- Deduction system: The company's taxable profits are reduced to the extent that distributions are made to shareholders. This option evidently involves timing issues i.e. a reconciliation of the taxable year of the company and the shareholder, which is why it is not as common. Iceland currently follows a version of this system.
- Full integration: In its most extreme form, an imputation system would disregard the tax form of the company entirely and only levy taxes at a single level (similar to how we tax partnerships, for example). This method, known as the "full integration" approach is understandably not commonly used, due to the differences in characteristics of partnerships and companies. Its application is typically limited to companies with partnership like features, such as US S-Corps or limited liability companies (LLCs). One variant of this is the exempt-exempt-tax system for the treatment of corporate profit and dividends proposed by Kelkar & Shah, 2012.
Each of these systems has its pros and cons. Exemption systems are a blunt tool and generally preferred for administrative simplicity only. Reduced rate systems are unable to consider the effective tax rate at the corporate level. Full integration systems are practically impossible to apply. A partial integration system such as imputation is therefore preferred to the extent that it is administratively feasible to match credits. One must also emphasise here that:
- Most of the relief systems described above are in the context of distributions to individuals. Countries typically allow more relief for intercorporate distributions, recognising the cascading effect of the double tax as distributions move up a corporate chain.
- The reliefs are likely to perform differently in a cross border context. Dividend distributions to non-residents are often subject to a higher withholding tax (as in the case of Russia or the US) while dividends received from offshore sources are sometimes exempt (as in the case of Singapore). Further, there is some discussion around how a classical system may perform better and create fewer distortions in a cross border context as compared to relief based systems.
- Some countries may apply one or more versions of these relief systems
The Indian situation
With this backdrop, let's look at the Indian dividend tax system. Unfortunately, there is no clear answer to any of the three questions.
- Fixing the aggregate tax rate: Until the introduction of the DDT in 1997, our corporate and dividend taxes seem to have been designed in a manner that factored in their connection to personal tax rates. However, since 2000, dividend tax rates have been increased (2000), reduced (2001) and increased (2007) again, with the ADT being the most recent addition. Their autonomous movement suggests a lack of strategy on the aggregate tax rate. Corporate and personal tax rates have remained broadly unchanged at 30% since 1997, with some variations to surcharge/cess and a couple of outlier years such as 2001.
- Choosing a dividend tax structure : The dividend tax structure should not be seen distinctly from the decision on aggregate rates. Our dividend tax structure followed a coherent policy so long as our aggregate tax rates paid heed to personal tax slabs, although we have tended to emphasise the importance of administrative concerns. We went from an imputation based dividend relief system prior to 1959, to a withholding credit system in 1959, to a reduced rate dividend distribution tax in 1997, each time citing administrative reasons - see the Explanatory Notes to the Finance Act 1959 and the Explanatory Notes to the Finance Act 1997 for details.
In comparison, the Explanatory Notes to the Finance Act 2016 justified the levy of an ADT by referring to the vertical inequity of taxing shareholders at 15%, where those with high dividend income would have otherwise been subject to the 30% slab. The notes do not examine the issue in further detail, and the logic leaves us hanging as it does not consider the broader issues relating to corporate tax structure.
The Legislative Critique
We now turn to legal issues associated with the structure of the DDT. Since it is a company level transaction tax which exempts shareholders, it does not mesh well with the larger structure of our income tax act and international tax provisions. It results in incompatibility with treaty provisions, denial of foreign tax credit to non-resident shareholders on DDT paid in India and disallowance issues under section 14A of the Income Tax Act. These issues exacerbate the problems created by an ill thought out corporate tax rate (for example, by increasing the preference for debt over equity). Some of these issues have led people to argue that the DDT should be replaced by a dividend withholding tax at the company level.
Let's look at the incompatibility with treaty provisions, for instance. At a time when international cooperation on tax matters is taken seriously, the DDT unilaterally overrides the international tax treaties India has committed to, since none of these treaties were written with a company level transaction tax in mind. This is a demonstration of bad faith and may be also considered an abuse of pacta sunt servenda. We should note that Estonia and India are the only two countries amongst the OECD and BRICS countries that apply a dividend distribution tax instead of a traditional form of shareholder taxation. This means that we no longer fall within the purview of most double tax treaties, and have unilaterally taken a bigger piece of the pie. Estonia does not levy a tax on undistributed corporate profits though, unlike India which does. Also instructive, is that South Africa used to have a version of the DDT, known as the secondary tax on companies (STC) which was done away with to align the taxation of dividends in South Africa with the international norm.
The stacking up of the ADT over an already imperfect DDT makes matters worse -- it has lead practitioners to argue that the form of the new levy results in economic "triple taxation" (See
here, here and here). It may be tempting to pose a counter that, if the dividend tax was paid at the shareholder level, and high income shareholders were asked to pay an aggregate tax higher than personal rates, it would still result in economic double taxation of corporate profits and not "triple taxation". However, this is not a valid counter because the difference in taxable units results in legal anomalies that effectively convert the ADT into a third layer of tax. For example, a shareholder receiving dividends subject to DDT would face disallowance issues under section 14A, while the shareholder paying ADT may not necessarily be, since ADT distributions are not "exempt" under section 10(34). This difference doesn't seem to be grounded in a considered or logical basis.
A host of questions remain open. If the DDT was introduced to bring about administrative simplicity, we need to ask whether the benefit still holds with the introduction of the ADT? If not, shouldn't technologies such as e-filing and comprehensive equity ownership records at NSDL/CDSL enable us to shift to a dividend withholding mechanism without significantly adding to administrative complexity? We follow a withholding mechanism for other forms of income such as interest on securities, and there doesn't appear to be a compelling reason why it should not be possible in the context of dividends.
As a broader level comment, our tax laws seem to suffer most from myopic drafting, brought about through these piecemeal annual changes that Nani Palkhivala referred to as "precipitous tinkering". Their amendments often ignore the larger conceptual framework of tax policy, creating more confusion than an introduction of a levy should involve. We need to find ways to defend the conceptual integrity of tax policy in each year.
RM Bird, 2002 c.f. OECD. (2014), "Fundamental principles of taxation", in OECD., Addressing the Tax Challenges of the Digital Economy, OECD Publishing, Paris. DOI
M. Govinda Rao & Kavita Rao, "Trends and Issues in Tax Policy and Reforms in India", in Bery, Suman, B. Bosworth and A. Panagariya, (eds.) India Policy Forum 2005-06, Sage Publications, New Delhi, pp.55-121.
Vijay Kelkar & Ajay Shah, "Indian social democracy: The resource perspective", NIPFP Working Paper, 2011.
I thank Jonathan S. Schwarz for his comments on the UK and South African position, and Sriram Govind for valuable discussions. I also thank the anonymous referee for thought provoking comments.
Shreya Rao is a lawyer who lives in Bangalore.