ISSN (Print) - 0012-9976 | ISSN (Online) - 2349-8846

A+| A| A-

Administering Goods and Services Tax in India

The long process of introducing a comprehensive Goods and Services Tax at the national level has thrown up many complex challenges. The question of the nature and structure of the agency for administering and collecting the GST is one of the most crucial ones. It appears that the existing institutional and organisational structure with multiple departments (one at the union and one in each of the states) would be maintained for collection of what is essentially a single tax (though with two components), thereby severely eroding most of the expected benefits of a national GST as well as putting a heavy burden on taxpayers. The influence of the bureaucratic complex and power politics of the organised civil services as well as a lack of political will are the major reasons for such a possible outcome. The need is to move towards a harmonised, modern and professionally managed national revenue agency for GST with a proper safeguard mechanism for preserving the fiscal federal nature of India and its states.


Administering Goods and Services Tax in India

Reforming the Institutional Architecture and Redesigning Revenue Agencies

Praveen Kishore

The long process of introducing a comprehensive Goods and Services Tax at the national level has thrown up many complex challenges. The question of the nature and structure of the agency for administering and collecting the GST is one of the most crucial ones. It appears that the existing institutional and organisational structure with multiple departments (one at the union and one in each of the states) would be maintained for collection of what is essentially a single tax (though with two components), thereby severely eroding most of the expected benefits of a national GST as well as putting a heavy burden on taxpayers. The influence of the bureaucratic complex and power politics of the organised civil services as well as a lack of political will are the major reasons for such a possible outcome. The need is to move towards a harmonised, modern and professionally managed national revenue agency for GST with a proper safeguard mechanism for preserving the fiscal federal nature of India and its states.

Praveen Kishore ( is currently director, Integrated Child Development Services, Bihar.

ndia has been working to introduce a comprehensive Goods and Services Tax (GST) for almost a decade, though the wait now seems perennial. However, one thing that appears reasonably certain at this point of time is its inevitability despite so many roadblocks. GST as a tax policy and economic reform package is in itself a huge area and is not the focus of this study.1 Instead, this article is more about tax administration and (re) organisation of revenue agencies for effective and effi cient operationalisation of the proposed GST as it is widely recognised that “tax administration is tax policy”, especially for developing countries. This applies even more so in the case of diverse, federal and competitive democracies like India. Fiscal federalism in India is tilted towards the union with important tax powers lying with the union government and with a constitutional mechanism for mandated rule-based devolution of tax revenue to states, which share major expenditure responsibilities. The institutional framework of revenue administration consists of two large union revenue agencies as well as a large number of state- and local-level revenue agencies which coexist, more of less oblivious of each other’s existence. Against this backdrop, an integrated, modernised and harmonised revenue administrative architecture is a sine qua non for successful execution of an ambitious tax reforms agenda, especially GST. However, the emerging administrative architecture for GST appears to be a severe compromise due to the overarching landscape of the bureaucratic complex and power politics of the Indian civil service as well as a lack of political will. If not tackled, it could lead us to a situation where most of the envisaged benefits of the proposed GST would be severely eroded.

1 Fiscal Federalism and Revenue Collection – Union and States

The world over, tax systems have undergone signifi cant changes during the past two decades as many countries, including India, have undertaken comprehensive tax reforms. Though the evolution of the Indian tax system was motivated by international concerns, yet in some ways it is different and even unique (Rao and Rao 2006: 4). Unlike most developing countries, which were guided in their tax reforms by multilateral agencies, Indian tax reform attempts have largely borne a domestic brand (ibid). They have been calibrated in response to changes in the development strategy over time while keeping in tune with the institutional arrangements in the country (ibid). The reforms of the past two decades have

april 28, 2012 vol xlvii no 17

Figure 1: Share of Major Taxes in been remarkably succeGovernment’s Total Tax Revenue in 2008-09

ssful in many respects.



Corporate tax 23% Sales tax/VAT 21% Other central tax Excise 2% 11% 13% state/local taxes 13%
income tax Yet, a lot more still needs to be done, especially on the institutional and organisational front. It would be instructive to have an idea of the shares and trends in revenue colle ction by the union and state gov ern-

Service Customs ments separately. tax 6% 11%

Figure 1 gives the per-

Source: Compiled from Government of India (2010):

centage share of major

Economic Survey 2009-10, Ministry of Finance, Chapter 3, and Reserve Bank of India (2010): State Finances: A taxes in the national Study of Budgets of 2009-10, Chapter 3 and Appendix.

revenue collection for 2008-09. It can be seen that six taxes yield the bulk of tax revenue forming around 85% of the total. Out of these, five are levied and collected by the union government contributing around two-thirds of the total tax revenue collected by all levels of government (union, state, local/municipal). In 2010-11, income taxes (individual income tax and corporate income tax), which are levied by the union government have contributed to 37% of total revenue (and this proportion is increasing continuously) whereas the remaining 63% is contributed mainly by “consumption taxes” levied both by the union (excise duty, custom duty, service tax – 28%) and the states’ (value added tax (VAT), entertainment tax, etc, 35%) (GoI (2011), Economic Survey 2010-11, Table 3.3, 3.11). In Figure 1 “other state/local taxes” include local and muni cipal taxes like property and house tax, entertainment tax, etc. Looking at the data of the past 30 years, we notice that the relative shares of union and state government taxes in total tax collection has remained almost stagnant at around two-thirds and one-third, respectively. Though largescale systematic reforms of union level tax policy and structure started in 1991, the share of the states has only marginally changed. This has happened due to the dynamic nature of taxes levied by the states, mainly being retail sales tax/VAT which are dependent on economic activity and growth.

States spend more than half of the total combined expenditure while collecting around one-third of revenue. The difference is largely met by mandated transfers of union tax revenue to states. Thus, the states’ own tax revenue constitutes only around 60-64% of their total tax revenue, the rest being transfers from union government tax revenue as per the formula of the finance commission. Out of their own tax revenue, the major source of tax for states is sales tax/VAT, which has been contributing around 60-65% of total states’ tax revenue over the years. Due to the predominance of sales taxes, any attempt to improve the revenue productivity of the states’ tax system, therefore, is inextricably intertwined with the reform of the sales tax system and in this respect, the recent reform of moving towards a destination-based VAT has been extremely important (Rao and Rao 2005: 39). The gradual move from sales tax towards VAT has been the groundwork for eventual move towards a comprehensive nationwide GST.

Economic & Political Weekly

april 28, 2012 vol xlvii no 17

Disaggregating state tax collection data state-wise gives us interesting insights especially on institutional and organisational capacity. There are considerable variations among states in terms of their revenue productivity and effi ciency. Whereas on the one hand, we have states collecting between 10% and 12% of state GDP as taxes (Karnataka, Delhi, Tamil Nadu), on the other there are states which collect less than 5% of their GDP as taxes (Bihar, West Bengal, Assam, north-eastern states) (RBI 2010 – Statements 18 and 19; GOI 2009 – Table 4.13). On the whole, the economically more advanced states have high tax-GsDP ratios, whereas poorer and less developed states have low tax-GSDP ratio. One factor behind this could perhaps be that underdeveloped states also lack the institutional and administrative capacity to tax revenue resources. However, making such simple conclusions may be naive. There could be other important socio-economic and political reasons behind such a diverse performance. In any case, it is widely believed that state-level tax departments are on the whole less effi cient and effective than union agencies.

In such a scenario, the widespread differences in administrative capabilities of different states’ revenue agencies and the difficulty of carrying out institutional reforms of these agencies due to various political-bureaucratic reasons can lead to a situation where they become a bottleneck for any wide-ranging tax policy reform at the national level like the envisaged GST, which requires considerable harmonisation between union and state governments not only in the tax base and the tax rate, but also in tax administration.

2 Attractiveness of Goods and Services Tax in India

There are many differences in the way VAT (or GST, as we now tend to call such taxes) are understood and implemented around the world. Nevertheless, there are certain common characteristics of such taxes which can help us in understanding and analysing various related issues. VAT is a tax on consumption, paid ultimately by the final consumer. The tax is generally levied on a broad base (as opposed to, for example, excise duties that sometimes cover specific products). The system is generally based on tax collection in a staged process, with successive taxpayers entitled to deduct input tax on purchases and account for output tax on sales. In general, many countries with VAT impose the tax at all stages and normally allow immediate deduction of taxes on purchases by all but the final consumer (OECD 2006: 6). The main feature of VAT/ GST is their neutrality, irrespective of the nature of product and services, the structure of the distribution chain and the technical means used for its delivery.

It is widely felt that for operation of a common market, India needs a unifi ed GST levied nationally. Further, administratively a centrally administered VAT/GST is highly suitable for federal nation states. Within the constitutional assignment of tax powers in India and the current political environment, however, a purely federal VAT is not considered feasible even though it may be considered desirable (Rao 2008: 3-4). The power to levy sales tax/VAT constitutes the most important tax powers of Indian states. Further, the power to levy this tax is a


“power” in itself and not just revenue power and taking this away will grievously reduce states’ fi scal autonomy.

It is widely agreed that GST is a major improvement over the present system due to many reasons. The excise at manufacturing level gives rise to definitional (“what is manufacturing?”) and valuational issues (cost of production or the value at which the tax to be levied) and makes its effective burden dependent on the supply chain (Poddar and Ehtisham 2009: 3). As per the constitutional arrangement, the state government cannot levy taxes on services. This makes it difficult to tax goods supplied in a composite form bundling both goods and services. Further, with high buoyancy in service taxes, there has been demand from states that they should also have express power to tax services, which today forms almost 50% of India’s GDP. Another problem in the present system of indirect taxation is tax cascading. The whole structure has become quite complex, with high compliance cost, inefficiency and administrative lacunae (ibid: 13).

Due to all these difficulties, the basic objective of GST reform has been to establish a tax system that is economically effi cient and neutral in its application, distributionally attractive and simple to administer (ibid: 11). The task force on GST of the Thirteenth Finance Commission noted that a conventional GST cannot be implemented without the states losing their fi scal autonomy (Government of India 2009a: 3). However, this does not appear to be politically feasible since revenues from state VAT account for substantial proportion of the states’ revenues. Therefore, a solution has been found within the existing federal framework where both levels of governments would have concurrent powers to tax domestic trade in goods and services (ibid: 4). Thus, it has been agreed that India will have a dual GST – a concurrent levy to be imposed by the centre/union (to be called the Central GST – CGST) and state governments (to be called the State GST – SGST) separately on the same base and implemented by multiple statutes. However, the basic features of law such as chargeability, definition of taxable event and taxable person, measure of levy including valuation provisions, basis of classification, etc, would be uniform across these statutes as far as practicable (GoI 2009b: 14). GST would be paid to the accounts of the union and state government separately, which will also require utilisation of input tax credit paid for union or states separately (ibid: 14). Within this broad framework, further progress on the actual design of process, structure, rules and regulations as well as institutional and organisational mechanism has been proving quite diffi cult.

3 Institutional Structure of Revenue Agencies in India

The Indian federal structure has uniquely shaped the way revenue organisations have taken root in India. At the foremost, the division of tax powers between the union and the states has made it imperative to have three layers of organisation, first at the union level, second at the state level, and third at the local level, though the last is the least developed. Further, states in India being large independent politico-administrative units as well as sociocultural entities with democratic form of provincial governments, each state has its own institutional structure, administrative machinery and large bureaucracy.

86 Thus, we have union government tax departments which function under overall control of the Union Ministry of Finance as well as tax departments under each state government which function under the ministry/department of taxation of respective state governments with their own structures, processes and bureaucracy. In addition, there are local-level revenue agencies (mainly urban municipalities) which have been entrusted with the responsibility for collection of some specifi c local taxes mainly house property, entertainment, etc, by the state governments. Thus, the overall macro-institutional structure is somewhat akin to “type of tax” model,2 where different levels of government are responsible for different types of taxes and have established separate agencies for this purpose. The broad concepts of taxation economics and the suitability of different types of tax instruments for different levels of government3 can be found broadly applied in India.

The upshot is a multiplicity of revenue agencies in India, where taxpayers often have to deal with different revenue agencies for different types of taxes. It has often the effect of unnecessary fragmentation of tax management systems, duplication of functions and inconsistent treatment of legal as well as accounting issues, increased compliance cost, diffi culties and even increased harassment of taxpayers. The need for a modern and taxpayer-friendly revenue agencies for GST assumes all the more importance against this backdrop. From a taxpayer’s point of view, it is always desirable to have a single revenue agency taking care of all tax liabilities. In not for all taxes, one agency for GST would be the minimum which most of us would agree to in principle. A consideration of the present system and institutional structure are important because it broadly delineates, in some sense, a landscape in which any redesign idea can be constructed administratively and politically within the larger sociopolitical, institutional and organisational environment.

At the union level, there are two large government organisations, each separately responsible for administration and collection of “direct taxes” (being taxes mainly on income and wealth) and another responsible for management and collection of “indirect taxes” (being taxes mainly on consumption and production). There two departments function under the overall control of two boards called the Central Board of Direct Taxes (CBDT) and Central Board of Excise and Customs (CBEC). These boards are responsible not only for overall supervision and control of the fi eld departments, but are also entrusted with all policymaking and decisions in the areas of union taxation.

These agencies function through their line organisations, fi eld offices, and innumerable specialised bodies often called directorates. The line departments are organised in matrix form along (territorial ×functional × taxpayer) axes in differing fashion and at different levels. The CBDT and CBEC together are one of the largest organisations of the union government employing as many as 1,30,000 people with a presence in more than 1,000 locations having more than 1,200 offices across the length and breadth of India.4 These organisations are an integral part of the Government of India and are staffed by permanent government employees. The highest levels of these

april 28, 2012 vol xlvii no 17

organisations are manned by professional permanent bureaucrats of the Indian Revenue Service (IRS).

At the state level there are generally a single tax department entrusted with management and collection of states’ taxes. For all state governments, as sales tax/VAT forms the single largest source of tax revenue, the departments are generally called sales tax or VAT or commercial tax department and are responsible not only for sales tax/VAT but other taxes like the entertainment tax, entry tax, luxuries tax, and professional tax, etc, as and when levied. However, some specifi c activity taxes like excise on alcoholic liquor, taxes on vehicles, stamp duties, land revenue, etc, are assigned to the respective line departments which are otherwise in charge of such functions. In many states, some of the taxes belonging to the state list like taxes on land and property have been assigned to local level/ municipalities and are thus administered by them.

State-level tax departments are an integral part of the state government machinery and are staffed with permanent government employees of the respective states. The middle levels of these organisations are filled with state civil servants and the higher level are managed by both state civil servants as well as professional bureaucrats of the Indian Administrative Service (IAS). The state-level tax departments are comparatively smaller with an obviously limited jurisdiction within the physical boundaries of the respective states. As the states vary in size, the state agencies do as well considerably, employing anywhere from 1,000 to 13,000 employees. The organisational structure of state-level departments is simpler, with basically a (territorial × functional) structure. Generally, the department is headed by a commissioner or secretary/principal secretary who has a secretariat with middle and lower management level support officers and staff. Then, there are fi eld offi ces, quite similar to that of union government offices. Often the designation and hierarchies are also similar like commissioners, additional/joint commissioners, deputy and assistant commissioners, tax officers, inspectors, assistants, etc, but the span of control and reporting relationships may vary. Further, these do also vary across different states.

4 Question of Efficiency and Harmonisation among Multiple Revenue Agencies

An appropriate strategy for tax reform would fi rst involve studying the tax structure and setting appropriate policy goals, and then modifying them in the short term by taking cognisance of the associated administrative problems. If the ordering is reversed, and administrative consideration became the binding constraint in tax reform, which by its very nature is a longer term process, the tax system is likely to play only a very limited role in achieving economic policy objectives (Shome 1995). As often happens in India, it is the administrative bottlenecks which start playing the dominant role in directing tax policy.

Organisational structure and operational requirement as well as reorganisation of the two union government agencies have always been on the agenda of the various tax reform initiatives. Though it has mostly been tackled as part of larger tax reform initiatives, started by the union government in the

Economic & Political Weekly

april 28, 2012 vol xlvii no 17

early 1990s, in themselves the agencies have seen major changes in structure, process and organisation of not only human and infrastructure resources, but in almost all the areas of tax administration. As a result, there have been some significant changes in tax administration machinery, with some far-reaching consequences. The historical change in relative share of direct and indirect taxes is a case in point. If we

Table 1: Staffing and Revenue Collection of Union Government Revenue Agencies

Organisation Staffing/Number of Employees Tax Tax Total Senior Junior Support Collection Collected Staffing Officers Officers Staff (Rs billion) Per (2008-09) Employee (Rs million)

CBDT/income tax department 61,300 3,000 6,000 52,300 3,382 55.17 CBEC/central excise and customs 67,400 2,500 10,000 54,900 2,695 39.98

Total 1,28,700 6,500 16,000 1,07,200 6,077 47.22

The number of employees has been rounded off to nearest hundreds. The numbers are those of “sanctioned posts” as against the “working strength” being the number of employees actually working/employed at a particular point in time. Over past few years, there has been severe shortage of staff in CBDT going to the extent of as much as around 20% for some of the cadres. The situation of CBEC has also not been much different. Source: Compiled from Government of India, Ministry of Finance:Economic Survey 2009-10, Annual Report of Ministry of Finance 2009-10 and Comptroller and Auditor General of India: Audit Report of Income Tax and Central Excise and Customs of various years, and personal knowledge of author.

look at the relative figures for tax collected per employee for these two agencies, we find the direct tax agency to be more efficient, as shown in Table 1 but again, this has happened only recently and may not reflect the whole story. There has been a wide-ranging technical upgradation programme, increasing use of information technology for improving the service to taxpayers and for taxpayer management and automation of routine processes, etc, in both the agencies. As a result, these two agencies are better managed, more efficient and effective than most of the state-level tax agencies, with few exceptions, where reform efforts have not been very effective due to various reasons.

There are wide-ranging differences among revenue agencies of different states not only in terms of their potential for revenue collection but also in terms of size and operational effi ciencies which can roughly be measured in terms of tax collected per employee, though there are other measures also, like cost of collection, which we will examine briefly. Although data in this respect is not readily available, the calculation has been made for seven medium/large states for which some information could be accessed as given in Table 2 (p 88). We can see wide variations in the tax collected per employee from as little as Rs 1.8 crore to as much as Rs 4.8 crore. It must be noted that the highest figures are lower than the figures for the all-India average of the Income Tax Department (ITD). It is indicative that most of the state-level tax agencies are less efficient than union agencies. However, some states like Gujarat, and Tamil Nadu can be considered as better managed. The short comparison also points in the direction that the economically more developed states are able to generate more taxes per employee, which may not always be linked to the efficiency of their tax departments.

To have some more analytics, we can look at another important and popular measure of administrative effi ciency and operational effectiveness – the “cost of collection”. It is roughly


defined as expenditure incurred on collection of taxes as percentage of tax collection. Table 3 gives these fi gures for central departments and two state agencies.

For the income tax department, the cost of collection has been decreasing continuously and significantly for the past decade. But all this may not be due only to an improvement in administrative and operational structure. On the CBEC side, which is much more relevant here for the GST question, there has been an improvement in the cost of collection but its precise reason

Table 2: Staffing and Revenue Collection of Few State Government Tax Departments

States Staffing/Number of Employees Tax Tax
Total Senior Junior Support Collection Collected
Manpower Officers/ Officers Staff (2008-09) Per
Managers (Rs Crore) Employee
(Rs Crore)
Andhra Pradesh 9,500 130 2,090 7,280 35,700 3.76
Bihar 3,020 80 440 2,500 6,300 2.08
Gujarat 5,000 180 490 4,330 24,200 4.84
Tamil Nadu 10,400 150 2340 7,910 34,500 3.32
Uttar Pradesh 13,000 29,100 2.24
Madhya Pradesh 4,700 130 650 3,920 8,500 1.81

The number of employees has been rounded off to nearest 100. The numbers are those of “sanctioned posts” as against the “working strength” being the number of employees actually working/employed at a particular point in time. Source: Compiled from websites of state governments’ commercial tax or sales tax departments, tax collection figures from Reserve Bank of India (2010): State Finances: A Study of Budgets of 2009-10, statements 18, 19.

Table 3: Cost of Collection for Union Government Revenue Agencies and Some State Tax Departments

Organisation Cost of Collection (as % of Tax Collection) in the Year 1999-2000 2005-06 2009-10

CBDT/income tax department 1.83 0.86 0.71

CBEC/central excise and customs 1.71 1.37 1.21

Tamil Nadu commercial tax department <1 0.71 0.83

Madhya Pradesh commercial tax department 0.85 1.00

Sources: (1) Compiled from Government of India, Ministry of Finance, Annual Report of Ministry of Finance 2009-10 and 2010-11, Administrative Handbook of Income Tax Department-2011, and Comptroller and Auditor General of India: Audit Report of Income Tax and Central Excise and Customs of various years.

(2) Compiled from Government of Tamil Nadu, Commercial Tax Department, Demand No 10, Policy Note 2011-12 and of 2006-07, as well as Government of Rajasthan, Commercial Tax Department, Annual Administrative Report, 2010-11.

needs to be further researched. A lot of reduction in the cost of collection may be due to policy improvement in tax design and the resulting growth in tax collection as well as buoyant economy. The data for states are difficult to find, and only two states could be compared. The fact that Tamil Nadu is a bettermanaged state is again reflected here with its cost of collection always below 1%. However, it has almost been stagnant over the past decade. Madhya Pradesh could be a typical state with an archaic institutional and operational structure. Its cost of collection has, in fact, increased during the past five years. Further, the cost of collection may also be misleading in one sense. The low cost of collection may point towards a low level of administrative and infrastructural capacity of the agencies which has the effect of increasing the cost of compliance for the taxpayers as has been pointed out in many studies (Chattopadhyay and Das-Gupta 2002). Nevertheless, the cost of collection fi gures as well as collection per employee definitely point towards a larger administrative malaise as well as procedural and operational issues which the nation faces in moving towards a harmonised GST administration: How to coordinate and design a wellintegrated operational and administrative structure with so many diverse institutions across the nation having widely varying capabilities, structure, organisational cultures and histories?

As of now, there is minimal coordination and cooperation between union and state-level revenue agencies as well as among various state-level agencies, all functioning independently in virtually their own silos. Here, a pertinent question would also be as to whether there is actually a need for multiple agencies to administer and collect a comprehensive nationwide GST? As the excise/service tax wing of CBEC and statelevel revenue agencies (VAT/commercial tax departments) would have a similar tax base and would require similar processes and even organisational structure, it would not make much sense to have two agencies for the same function, leading to duplication on many fronts.

5 Single or Multiple? Envisaging Revenue Administration for Indian GST

Although the task force on GST of the Thirteenth Finance Commission has come up with an administrative architecture for a proposed dual GST, it is too sketchy and has only concentrated on a few issues like registration of taxpayers, the GST invoice, periodicity of GST payments and on the administrative structure of GST (Government of India 2009a: 49-53). The thrust of the administrative recommendation is on maintaining two separate entities for the operationalisation of GST at the union and state level. Thus, it has been recommended to have two departments/agencies, one for the union and one for each state GST under the control of union and respective state governments. These agencies would be separately responsible for functions like assessment, enforcement, audit/scrutiny, etc, of their respective portion. However, it has also been recommended that all the processes and procedures would be the same under both CGST and SGST. This appears to be a contradiction. It becomes salient when we closely examine some of the harmonised operational structures envisaged and discussed.

  • The jurisdiction between the union and state agencies needs to be properly divided and operationalised. If it can be done in such a manner that a taxpayer has an interface with only one agency, it will be ideal. Instead of having one agency, the alternative – though much more difficult to operationalise on various other considerations – is to have a jurisdictional division between the two agencies on some economic criteria like turnover threshold, etc, (ibid).
  • All the procedures under CGST and SGST need to be uniform. It is important to have a common information technology infrastructure, and this can be done only at the national level. It would also be desirable to have taxpayers’ information network (on the lines of ITD’s TIN) and it should be shared between the union and the states. All the information should be stored in a common data base. Some progress has been made in terms of devising a GST network (GSTN) to be set up on the line of ITD’s TIN (Government of India 2011). There is also a plan to provide three core services to taxpayers through a common web portal – registrations, tax returns and tax payments (ibid).
  • april 28, 2012 vol xlvii no 17

  • Various processes and activities like tax payment, transaction reporting, etc, should ideally be made through combined payment and transaction reporting statements, and such documents should again be common.
  • Tax return filing should also be common. Further, audit being an important activity for ensuring compliance, it would be ideal if both levels of agencies have a structure whereby the same taxpayers are not subject to a simultaneous audit by both the agencies.
  • There should be a common taxpayer’s registration and identification base. The taxpayer identifi cation number of ITD (called PAN) is already being used by the central excise and it can easily be accepted and modified for the dual GST model. This appears to be the plan also under the GSTN (ibid).
  • There is need to have a common and standardised GST invoice system as an invoice forms a primary source of information and is essential for effective control and enforcement as well as audit of the GST.
  • As the tax base would be common, it would make sense to have a common appellate structure and similar other authorities like for advance ruling, etc.
  • All the above signposts/recommendations suggest that if these types and levels of harmonisation are achieved, why then have two separate agencies? It effectively means administering and collecting a GST through a single department.

    Thus, the best organisational design for administration of the GST would be to have a single agency/department at the national level for this “tax type”, with further operational divisions on functional (taxpayer management and bulk operations, assessment and audit, enforcement, etc), regional (for different states, group of small states), and/or taxpayer segment (companies, partnerships, individual traders, manufacturing, service, etc) criteria in essentially a matrix structure spread all over the country. This will ensure the maximum level of operational ease and efficiency, as well as better taxpayer services through a single agency structure for what obviously is a single tax (though it is shared between union and state government).

    Nevertheless, there are some other administrative and policy issues, like the establishment of an apex level GST council for policy and high administrative level coordination between different states and the union government, etc, which are essential to preserve the fiscal federal nature of our country. What would be the structure, functions, membership, roles and responsibilities of such a council has not yet been agreed upon, this obviously being a contentious political issue. Furthermore, how and in what manner should interstate transactions be treated and taxed is still not fully resolved despite a broad agreement on the IGST. The rates of taxation, the revenue sharing between union and states, framework for exemption, thresholds and composition are some of the issues still being debated at the highest level. Despite all this, the attractions and merits for a single agency in charge of collection of the GST do not diminish. In any case, the projected roll-out date of the GST regime from 1 April 2012 may have to be extended in light of such contentious issues and the still elusive consensus from

    Economic & Political Weekly

    april 28, 2012 vol xlvii no 17

    some state governments (most of which are governed by parties other than the one in power at the union level).

    6 GST/VAT Administration in Other Countries

    There has been a worldwide trend towards some sort of VAT/ GST and it would be instructive to note as to how some of the countries have managed the administration of a reformed tax regime. In countries with two or more layers of governance, various political economy factors come to play an important role in such decisions and some compromise has to be achieved. China has recently moved towards a centralised VAT regime where revenue is shared with the provinces. It has been ensured that provinces get as much revenue as under the previous arrangement, plus a share of the increment (Poddar and Ehtisham 2009). In effect, the levy as well as collection are both administered by a central authority. Australia also has introduced a GST. There GST is a single national levy and all GST revenue is collected by the centre. Subsequently, a state’s share is returned to the respective states. In Canada, we have a system called the harmonised sales tax, where the tax is levied at a combined federal and provincial rate in the three participating provinces. Tax design and collection are controlled by the centre, but provinces have some flexibility to vary their tax rate. The revenues collected are shared among the participating provinces on the basis of consumer expenditure data for the participating provinces (Rangarajan and Srivastava 2007). In Germany and Austria, the tax design is controlled by the centre, but states collect the taxes. This has led to incentive problems and a tax administrative measure has been found to be used for realising tax policy goals (Poddar and Ehtisham 2009).

    In Mexico, the establishment of a VAT at the centre replaced the state sales tax, but had to be part of a political economy compromise that assured the states an automatic share of the revenue generated from all federal taxes (ibid: 17). VAT was introduced in Mexico as a federal tax mainly to harmonise taxes between federal and local governments by substituting more than 300 local taxes with VAT (Jose et al 2008). Brazil is a highly fiscally decentralised state and, historically, VAT has been the most important tax source for state governments in Brazil. However, it has posed difficult horizontal tax coordination problems. It has been observed that though state and local sources of revenue must be preserved, some limitation on sub-national government taxing powers is unavoidable (Burki et al 2000). The various problems related with decentralised VAT in Brazil have often generated demands for its federal administration.

    All these examples show that there is a tendency towards a centralised administration/collection machinery for GST/VAT, which is important and required for achieving a national-level efficiency and common market objectives as well as for minimising all sorts of distortions. A national GST holds a great appeal from the perspective of establishment and promotion of the common market in India but the political realities of India are different and it is now certain that India is not going to have a single national GST, but a dual GST. Due to this very factor, the need for administrative harmonisation at the revenue agency level assumes even more importance as


    harmonisation at the tax base and the tax rate level would be hard to come by in the Indian federal structure.

    7 The Bureaucratic Complex and Politics of GST Administration

    Despite these strong indicators pointing towards the desirability of a single GST administrative agency at the national level, the actual movement has been in the opposite direction. It is well understood that the important question of role and responsibility of union and state governments on policy issues related to the GST base, tax rate, etc, are signifi cant for a truly federal structure, its relation with an effective and effi cient collection mechanism and collection agency is quite weak. Therefore, it is quite possible to have a single, professionally managed GST agency/department at the national level responsible for only administration and collection of GST (theoretically, both state and union components, which in practice is merely a numeric division). However, the final proposal is for maintaining two sets of agencies responsible for collection of union and state GST and administration of related activities separately.

    The reasons clearly have to do much with political imperatives rather than issues of technical soundness and effi ciency. The GST policy is being discussed and debated at the highest political level and obviously the question of fiscal autonomy of states and their ability to raise revenue independently has played a decisive role. Therefore, though it may technically have been more attractive to have a unifi ed GST, the consensus has emerged for a dual-GST to be “levied” both by union and state governments separately. However, on the administrative architecture issue, question of how separable are the power to “levy” GST and its operationalisation for “collecting” this tax has not got due attention. It has happened largely due to the presence of a strong state-level bureaucracy and its rivalry with the union bureaucracy. As we have noted earlier, state-level tax agencies are dominated by IAS offi cers at the highest level and state-level civil servants at the junior level. On the other hand, union excise and custom duty administration is dominated by IRS of officers at the senior level and junior officers (of the union government) at the lower level. Therefore, without basically examining what could have been the most effi cient and effective way of “collecting” a dual GST, the question has been approached on the presumption that both these wings of the government would continue to exist (in the background, the existential threat to entrenched bureaucracies, the threat of their erosion of power and privileges, etc, has played its part), thus making impossible any meaningful examination of this crucial issue. However, implementation of GST through two departments in charge of simultaneously administering it over the same base may substantially nullify any envisaged benefi ts of this ambitious tax policy reform.

    This basically means that although there are considerable gains to be made by having a single agency to collect/administer a dual GST, it may not be politically feasible. Despite efforts to the contrary, this may ultimately lead to a situation where there would be significant administrative ineffi ciencies, higher

    90 compliance cost and where taxpayers would be forced to deal with two departments for compliance of even a single tax. Consider, the worst situation where a GST payer would be paying GST twice, once to the union agency and then to the state agency, maintaining two sets of records, fi ling two GST returns and dealing with two agencies at each and every stage of compliance – a truly disappointing scenario! Therefore, the issue of having two separate agencies for collecting what is essentially a single tax has to be faced squarely.

    Highly harmonised dual-layer separate revenue agencies at the union and state level often offered as the way out though certainly an improvement over the status quo is only a second best solution. Even to achieve this, the level of commitment and efforts required are considerable. The extent of harmonisation required to achieve a satisfactory level of operational results is truly significant. This has to be in terms of resource sharing and common processes and procedures – not only shared infrastructure and information technology platform (obviously there should be a single tax return form and a tax payment system, and preferably a joint audit and scrutiny mechanism), sharing of information and tax base through common architecture, cooperation and coordination at different levels, but also sharing and utilisation of human resources across the agencies especially at the higher level so that professional expertise in tax administration in compliance and enforcement operations, etc, can be effectively utilised. However, this is easier said than done. Obviously, it would face resistance from bureaucracies, more from the IAS cadre. Being generalists, their claim to be able to manage state revenue agencies as expert tax professionals is not easy to digest. Despite this, as they are powerful not only in state but also in union governments, it would not be easy to get the tax professionals to manage all tax agencies.

    An important step towards a modern, effective and effi cient tax administration is the management of revenue agencies by professional civil servants, who have specialised knowledge, experience and expertise in handling revenue functions. Therefore, it makes perfect sense to have a cadre of profe ssional civil servants which will be responsible for running revenue agencies not only at the national level but also at the state as well as local/municipal levels. In revenue agencies, like in many other organisations, at the higher management and strategic levels, the role and responsibilities of revenue administrators start getting complex, wider and involve more of supervision, mentoring, communication across levels and agencies as well as across boundaries of agencies, departments and ministries. On the other hand, the core operational level involves more of specialised (and some routine) functions where such a broad approach and exposure may not be so crucial. Accordingly, the personnel structure of the revenue administration should be designed in a manner so that the top managerial level can move across agencies, tax types, functions, and most importantly across types of governments (the union, state and local/ municipal) whereas the operational core remains more focused on its core-competency. The present structure at the union level is some what like this: IRS officers are able to move across

    april 28, 2012 vol xlvii no 17

    functions and regions (though limited across tax-types) and junior and lower level employees specialise in one or few functions and are assigned to a single region/state.

    The practice in states at present is different for higher level officers, who move from, for example, the public health department to the tax department, two totally unrelated functions which often require different levels of expertise, exposure and outlook. This system does not make much sense and has been prevalent only due to political reasons and the historically dominant positions of the IAS in government. The often cited claim that the highest level functionary should be a “generalist” rather than a “specialist” does not carry much weight in the modern world where governments have come to perform highly complex functions. The world over, it is the “specialist” manager who is in charge of overall supervision and direction. The need for functional specialisation of different “branches” of services has been recognised and recommended by the Second Administrative Reform Commission which has come out in favour of “domain” specialisation by different branches of civil service (Government of India 2009c).

    8 Conclusions

    GST being a single tax to be shared between the union and the states, its best operationalisation could be achieved through a single national agency/department in charge of collecting GST and implementing policies of the state and union governments. The GST so collected can conveniently be apportioned and distributed among the respective governments. The tax policy and other coordination mechanism can be taken care of by establishment of the GST council and other institutional structures. Any harmonisation with multiple agencies would not mean much unless it is of a very high level, which will effectively make multiple agencies redundant. However, looking at the complexity of issues involved, a start could be made with the second best option of harmonisation and coordination in some crucial areas. That this is not an end in itself but only the first stage in a gradual move towards a single revenue agency through one higher benchmark to another which should be clear from the outset. In this endeavour, some of the union territories could be the starting point for having a unifi ed agency for collection and administration of GST.

    The design of appropriate institutional and organisational structures is crucial for the success of GST reforms. In such a vision, the lynchpin of success would be a highly professional and motivated human resource, nationwide information technology architecture, processes and procedures, work fl ows and systems compatible with the GST structure and a high level of harmonisation with other revenue agencies especially ITD. Another important issue would be to integrate, align and reform the revenue administration and civil services structure with more efficient forms of organising government bureaucracies on professional lines. Only then the seemingly confl icting objectives of all revenue administration – effi cient, convenient and citizen-friendly taxpayer service delivery, on the one hand and effective compliance, enforcement and deterrence mechanism, on the other – can be achieved.


    1 There are a amount of studies and literature available on GST/VAT as a tax instruments, its ideal and optimal design feature, its effect on economic activities, its desirability vis-a-vis other instruments of taxation, etc. We are surely not entering that territory.

    2 Generally, three models for design of revenue agencies are recognised. Under the “type of tax” model there are separate endogenously multi-functional, self-suffi cient agencies/ departments for each major type of tax. Under “functional model” tax administration organises its work in logical groupings of core functions that encompasses all taxes. Under “taxpayer segment” model the revenue administration is organised around important segments of taxpayers on the assumption of different characteristics, need and compliance beha viour of taxpayer segments. These three models have been discussed in detail in Chapter 2 of OECD (2008).

    3 A detailed analysis of such technical issues can be found in Parthasarathi Shome, ed. (1995): Tax Policy Handbook, Fiscal Affairs Department, IMF, Washington DC; especially part II.

    4 Annual Report of Ministry of Finance, Government of India 2009-10 and CAG of India: Audit Report on Direct Taxes, Audit Report on Indirect Taxes (various years). The numbers are only approximate. Strictly speaking, they would be the third largest department, the first two being the Indian Railways and Indian Posts.


    Burki, Shahid Javed, Perry Guillermo ed (2000):

    Annual World Bank Conference on Development

    in Latin America and the Caribbean 1999, The World Bank, Washington DC.

    CAG, Comptroller and Auditor General of India (2010): Audit Report on Direct Taxes, 2009 and of previous years, New Delhi.

    – (2010): Audit Report on Indirect Taxes, 2009 and of previous years, New Delhi.

    Chattopadhyay, Saumen and Arindam Das-Gupta (2002): “The Compliance Cost of Personal Income Tax and Its Determinants”, National Institute of Public Finance and Policy, New Delhi, December.

    Government of India (2009): Report of the 13th F inance Commission, Ministry of Finance, New Delhi.

  • (2009a): Report of the Task Force on Goods and Services Tax, Thirteenth Finance Commission, Government of India, New Delhi.
  • (2009b): “First Discussion Paper on GST in India”, The Empowered Committee of State Finance Ministers, Ministry of Finance, New Delhi.
  • (2009c): Report No 10 of 2nd Administrative Reform Commission, Refurbishing of Personnel Administration – Scaling New Heights, New Delhi.
  • (2010): Annual Report of Ministry of Finance, 2009-10, New Delhi.
  • (2011): The IT Strategy for GST – Report of the Empowered Group on IT Infrastructure for GST,
  • Ministry of Finance, New Delhi.

    Government of India: Economic Surveys of 2010-11, 2009-10, 2005-06, 2001-02, 1996-97 and 199192, Ministry of Finance, New Delhi. These and Economic Surveys of other years are available online at

    OECD, Centre for Tax Policy and Administration (2006): International VAT/GST Guidelines, Organisation for Economic Cooperation and Development, Paris, France.

    – (2008): Tax Administration in OECD and Selected Non-OECD Countries: Comparative Information Series – 2008, Organisation for Economic Cooperation and Development, Paris, France.

    Jose, Pagan, Soydemir Gokce and Tijerina-Guajardo Jose (2008): The Evolution of VAT Rates and Government Tax Revenue in Mexico, Contemporary Economic Policy, Vol 19, No 4, October 2001.

    Poddar, S and Ahmad Ehtisham (2009): “GST Reforms and Intergovernmental Considerations in India”, Ministry of Finance, Government of India, Working Paper No 1/2009-DEA, March.

    Rangarajan, C and D K Srivastava (2007): “Fiscal Transfers in Canada – Drawing Comparisons and Lessons”, Working Paper, National Institute of Public Finance and Policy, New Delhi.

    Rao, Govinda and Rao Kavitha (2005): “Trends and Issues in Tax Policy and Reform in India”, Working Paper, National Institute of Public Finance and Policy, New Delhi.

    Rao, Kavitha (2008): “Goods and Services Tax in India”, National Institute of Public Finance and Policy, New Delhi, Working Paper 2008-57, November.

    Reserve Bank of India (2010): State Finances: A Study of Budgets of 2009-10, Mumbai.

    Shome, Parthasarathi, ed. (1995): Tax Policy Handbook, Fiscal Affairs Department, IMF, Washington DC.

    Economic & Political Weekly

    april 28, 2012 vol xlvii no 17

    Dear Reader,

    To continue reading, become a subscriber.

    Explore our attractive subscription offers.

    Click here


    To gain instant access to this article (download).

    Pay INR 200.00

    (Readers in India)

    Pay $ 12.00

    (Readers outside India)

    Back to Top