How to build tax administration

by Shubho Roy and Ajay Shah, Business Standard, 8 August 2016

Tax policy in India is moving forward with the first milestone of the GST. There are a few more building blocks required, most notably the Direct Tax Code. Comparable work on setting up the tax administration has not begun. The World Bank `Paying Taxes' study, 2016, placed India at rank 157. This came as no surprise to Indian taxpayers, who know about the travails of reassessment and excessive scrutiny. Private citizens dread a tax notice, and companies are often involved in endless litigation, negotiations, and tax discussions with courts, authorities, tax advisors and accountants. A search for ``harassment by income tax officers India'' gives 8,18,000 hits on Google.

The courts have started noticing the misbehaviour of Assessing Officers (AO)s. The Allahabad High Court, in `Commissioner of Income Tax v. Intezar Ali (2013)', asked CBDT to specifically investigate an AO for harassment and conduct driven by malice. The tax authorities' own tribunal, the Income Tax Appellate Tribunal, has questioned the department for mindlessly filing appeals when there is no merit in the department's claim.

The government is placing pressure on tax officers, and the Prime Minister has asked the tax administration to remove fear of harassment from the mind of taxpayers. Tax officers have protested the Revenue Secretary's increased scrutiny, and some quarters of the tax administration are demanding more autonomy for tax officers. One common element in these demands is separation between tax policy (at the Ministry of Finance) from tax administration (at CBDT/CBEC). These demands draw parallels to other jurisdictions (such as the US and the UK) where communication and interaction between the tax office (CBDT) and the policy setting executive (MOF) is governed by well specified protocols. Memorandums specify each tax authority's domain, and require them to communicate exclusively through formal channels. This blocks politically motivated tax harassment.

There is merit in this line of thought. It is similar to the thinking in India about how to setup the Principal-Agent relationship between the Ministry of Finance and RBI, or how to improve the working of CBI. However, as with RBI or CBI, setting up a sound agency requires a full perspective of the Principal-Agent problem. There are five key issues.

First, while measures protecting the tax administration from political interference are important, we should not merely hand out independence of tax administration. Focus on independence alone with no consideration of accountability mechanisms simply replaces political interference with executive abuse. While political interference can at least be addressed at the ballots, there is no way for a citizen to protest against abuse by unelected officials.

Second, tax targets in India are usually a simple rupee value target per geographical area under a Commissioner of Income Tax. A tax officer can, therefore, either pressurise maximum collection from a few high net-worth/income tax-payers, or maximise the number of persons under their tax net. As the latter is harder to achieve, the administrator favours the former. What is required is low rates and a wide tax base, but flawed tax administration has created incentives for the opposite.

Other jurisdictions (where tax authorities enjoy more independence) do not impose such simplistic tax targets. They undertake detailed statistical studies examining the nature, volume and distribution of economic activities in an area. These studies are conducted by bodies independent of the tax administration, and targets are set for each income tax bracket with specific emphasis on types of economic activity and expected revenue. In contrast, India is replete with stories of tax authorities harassing large formal sector businesses while a parallel informal/underground sector freely operates untouched.

Third, in other jurisdictions, there are consequences when tax officials are unable to defend their assessments in independent courts. These are marked against the department and, as an accountability measure, the win/loss ratio of the department is tracked. The officer conducting the initial assessment, the officer recommending appeal after initial loss, and every officer in between is held accountable for each decision made. In India, the win/loss ratio is ignored.

Fourth, information requests in other jurisdictions are efficiently managed. In India, a person is required to produce information or documents that are already submitted to the tax authorities in regular filing. Most jurisdictions specifically prohibit tax authorities from requesting the same information twice.

In India, taxpayers often travel to meetings with assessing officers only to find that the relevant officer is not present on the appointed date. There is no system of recording the absence of the officer, nor any electronic recording of previous meetings between that officer and the tax-payer being assessed. Questions are not always written down or communicated before meetings. As a result, if a case is transferred from one officer to another in India, the entire information gathering starts afresh.

Fifth, the lack of separation of powers in the tax system works in favour of tax officers. When the tax administration faces a large loss in court, the response too often is a retrospective amendment to the tax legislation. Fairness of such amendments is not considered and tax law is often vaguely drafted in favour of the department.

What is the way forward? CBDT must be seen as a mini-state akin to RBI, and held at an arms length from the Ministry of Finance. The Ministry of Finance must draft tax policy (i.e. the law that defines what is taxable and by how much), without involvement by CBDT. CBDT must be governed by a new `Tax Administration Code' which establishes sound governance procedures. This must include separation of powers, and detailed procedural law, for the legislative, executive and quasi-judicial wings of the CBDT. There must be accountability procedures embedded in the working of the board. Under these conditions, there must be independence of the CBDT staff from political interference. The mechanisms adopted for FSLRC and BLRC must be used to set up an expert Commission to draft this law.

We thank Sanhita Sapatnekar for useful discussions.

Back up to Ajay Shah's 2016 media page
Back up to Ajay Shah's home page