magazine article big title

Interview with Pawan K. Sinha

Pawan K. Sinha holds Masters in Economics from Delhi School of Economics (India), Post Graduate Diploma in Intellectual Property Rights and a Diploma in Financial Management. He belongs to Indian Revenue Service and currently holds the position of Additional Director General of the National Academy of Customs and Indirect Taxes in India. He has expertise in the fields of anti-corruption and internal control, training, compliance, tax assessment, container scanning, and intelligence and investigation. Pawan graduated from IACA MACS 2015-2017 on 7th December 2017. He earned his Master of Anti-Corruption Studies (MACS) 2015-17 degree with summa cum laude honours and his thesis “Exploring Effectiveness of Automation in Controlling Corruption in Customs: Theoretical and Empirical Evidence: A Case Study of India” was adjudged as the Best Master’s Thesis.

What brought you to the idea of choosing the research topic reflected in the title of your thesis “Exploring effectiveness of automation in controlling corruption in customs: Theoretical and Empirical Evidences: A case study of India”? Can you briefly summarize your research topic and main findings?

The ubiquity of corruption in taxation and customs services, though in varying degrees, is a common phenomenon in many countries, and India is not an exception. Being in charge of tax administration during different stages of my civil service career spanning nearly three decades, controlling corruption remains a very intriguing and daunting task. The measures taken to control corruption in India met with limited success. When I joined IACA’s MACS programme, I was introduced to its unique and well-structured inter-disciplinary and integrative approach to understand the complex spectrum of corruption. I could naturally relate various theoretical expositions and class lectures conducted by specialist academics and practitioners to my real-life experiences within my own social, economic, and political eco-system. Most importantly, I could relate my learnings to corruption-related problems faced by organizations I dealt with in my work, including the customs department where I worked for so long.

As part of India’s change in policy stance, a modernization process started in 1993. Being involved in this process, I recall that it was taken for granted that automation of customs would not only make processes quick and simple, but also ‘completely’ stamp out corruption. Customs administration in India, like in many other countries, embarked on a gradual process of simplification and modernization, including automation. Twenty years hence, while efficiency increased in the customs processes, corruption remained almost intractable, flagrant and only marginally abated. This was a counter-intuitive and bewildering outcome.

Having to choose a topic for the master’s thesis was an opportunity to explore the plausible answer to the puzzling and lingering questions in my mind, in particular the obfuscated relationship between efficiency, automation, and corruption. I also felt that a theoretical and empirical framework was required to better understand this relationship. Essentially, my master’s thesis challenges the axiomatic belief of officers, policy-makers, and scholars that by enhancing efficiency in customs through trade facilitation measures and automation, one can control corruption (bribery) in core processes effectively. I used the ‘single country - single sector’ case study method to understand the different shades of bribery (instead of the whole gamut of corruption) in core customs processes of goods clearances, which are partly automated. The master’s thesis highlights that there is tremendous potential in automation, as a tool, to control corruption, if automation is designed with a view to also prevent corruption by adequately addressing the causal mechanism of the drivers of corruption and inefficiency simultaneously. After measuring automation, efficiency, and corruption in the customs context as key variables and determining the nature of the relationship between them through paired linear regression, the paper explores causal mechanisms through process-tracing. It identifies different drivers such as those which neutralize the effectiveness of automation in controlling corruption. The master’s thesis uniquely theorizes and explains how automated Risk Management Systems, syndicated corruption, and intermediaries could be prime drivers of corruption in customs. It concludes that partial automation aimed primarily at improving ‘efficiency’ alone may trivially control corruption. It argues that if unique anti-corruption policies based on this contextual study were hardwired with efficiency-enhancing measures while designing automated customs processes, and were complemented by the preventive and repressive anti-corruption policy instruments, bribery in customs could be more effectively controlled.

You conducted interviews to collect data in order to “support longitudinal analysis of the state of corruption in customs”, can you tell us how was your research topic perceived by stakeholders who were being interviewed by you?

This was the most challenging part of my research. Finding stakeholders who could offer their experience-based responses to the state of corruption in customs, while maintaining randomness and pan-India representative sampling for ensuring reliable and valid data, was a daunting task. There were four main categories of stakeholders (respondents): customs officers, customs brokers, exporters and importers, and citizens with some knowledge about customs. Interviewing them individually and interacting with them in focus groups on the research topic was a unique experience. Generally, stakeholders across all categories welcomed the study. Many respondents initially expressed surprise on being approached for such an interview about corruption among senior government customs officers. There was an element of awkwardness in most of the people interviewed for an apparent reason: corruption is a taboo subject and many people prefer to remain silent. Most of them stated that they had never participated in this kind of study and required nudging to talk freely on various aspects of corruption (bribery). The majority of respondents agreed to share their experiences only under the condition of anonymity. The use of a consent form was very useful to diminish self-censorship due to palpable fear among respondents of any possible ‘self-incrimination’ and punitive actions which they feared may follow after publishing the survey’s outcome. While most customs officers and customs brokers candidly shared their experiences and observations, the exporters and importers could not provide any meaningful responses, and sample data concerning exporters and importers was discarded. The younger citizens perceived the research topic as most useful for the country and wanted to know whether the outcome of the study would be translated into any new anti-corruption policy measure(s).


Do you think that automation should be either increased or reduced in specific areas in order to fight corruption?

The master’s thesis challenges the efficacy of the current state of automation, which primarily focuses on ‘efficiency enhancement’ by reducing transaction costs and time in the import and export processes. There is also a proven endogeneity problem (of simultaneous causality) between inefficiency and corruption, and existence of confounding variables, such as automation, simplification of laws and procedures, audit and review system etc., that affect these two variables (inefficiency and corruption) in different ways, and lead to different outcomes. This would require finding a point of equilibrium through rigorous optimization process between these two variables, instead of seeking mitigation of corruption by merely enhancing efficiency. A theoretical analysis based on the concept of ‘routine tasks’ and ‘non-routine tasks’ suggested by Levy and Murnane (2003) has been undertaken, where automation can become ‘capital substitute’ for the former and ‘capital complement’ for the latter. While automation as ‘capital substitute’ may curtail opportunities of corruption, automation as ‘capital complement’ requires a design to enforce accountability (not merely transparency). Automation is a potent tool, if, in the design of automation, it is ensured that anti-corruption requirements are hardwired along with efficiency-enhancing features.

The research does not subscribe to reducing automation, but suggests improving its coverage and design, and proposes implementing it together with better capacity-building of officers and other stakeholders. In other words, there should be ‘smarter and holistic automation,’ and not merely automation to enhance efficiency. The research, therefore, underscores the need for organizational changes and targeted logistics to remove face-to-face contact between decision-makers and beneficiaries or service recipients, where automation could be an enabling tool, instead of expecting automation to achieve this objective by itself. The research also admits that the quality of automation has an important role to play in the control of corruption, and that there could be disruptive technologies in the future like the advent of artificial intelligence and quantum computing in custom processes. It suggests that there could be better opportunities to achieve a balance between efficiency and anti-corruption, provided such balance is prioritized at the design stage itself.  

What governmental policies should be prioritized to improve automation in India to better tackle corruption?

Automation is a high-cost, high-dividend investment. The cost involved is not only in terms of software and hardware costs, but involves costs of capacity development of officers, institutionalization of digital operations, and compliance of its users, such as exporters, importers, and other stakeholders in the supply chain. The Government of India has already launched the ‘Digital India’ initiative and has taken measures for expanding internet usage in automation. However, as the present case study relating to customs has revealed, while investing in modernized automation of service delivery is important, it is essential to revisit laws and procedures to simplify these in order to reduce ‘rent-seeking’ opportunities and red-tape, improve internal audit and oversight mechanisms to enhance accountability, and invest in modern capacity-building systems which should focus not only on operational trainings, but generate ‘esprit de corps’ among officers to create an organizational unity against corruption. Creating ‘nudge units’ to desist officers from indulging in corruption could be another innovative measure. The capacity of automation to control corruption would increase exponentially when other such measures are complemented.

Last but not least, how do you foresee the applicability of your research outcomes?

My research is a bold attempt to establish the ‘single - country single - sector researchmethodology with successful diagnostic capabilities to identify the type of corruption in an institution or organization, to measure the important variables like automation, efficiency and corruption, to identify specific drivers of corruption, and to understand their relationships using statistical techniques like correlations and linear regression. Corruption-related data are mostly non-parametric. The analysis used non-parametric statistical tools with 95% significance to identify the links. Realizing that such links might run into the ‘post hoc propter hoc’ fallacy, I successfully conducted the process tracing technique to understand the causal mechanism of the confounded variables. Thus, the soul of the research lies in establishing a methodology for theoretical and empirical evidence-based diagnostic systems, rather than merely being a study of efficacy of automation in Indian customs in controlling corruption. In other words, the research has much wider applications for the meso-level exploration of corruption in any sector and in any country rather than being limited to automation in Indian Customs. It also underscores that the case study of customs in India cannot be applied directly to design anti-corruption policies in other sectors in India or even customs in other countries. By using the established methodology in this master’s thesis, however, authorities may undertake cogent diagnosis processes which may help in designing the customized policy prescriptions specific to a sector, but this should be duly complemented by other usual anti-corruption micro and macro level preventive and punitive instruments. It is argued that such an approach would have a high probability of success in controlling corruption in any specific sector of a country. Hence, I am very optimistic about high prospects of applicability of my research in the real world scenario.

End Notes

Levy, F., & Murnane, R. J. (2003, November). The Skill Content of Recent Technological Change: An  Empirical Exploration. The Quarterly Journal of Economics, 1280-1281.

Post hoc propter hoc fallacy’ refers to thinking process „after this, therefore, because of this“, without true
causalities; for example, ‘the rooster crows before sunrise, therefore the crowing rooster causes the sun to rise’. Refer to for more details.

“Process tracing” is the cause-effect link that connects independent variables and outcomes viewed in fragmented form, allowing the researcher to view observable evidence at each step leading to fi nal outcome or overall consequence. Evera, S. V. (1997). Guide to Methods for Students of Political Science, Cornell University. Retrieved May 23, 2017, from Masaryk University :

A confounding variable is an outside infl uence that simultaneously changes the effect of dependent and
independent variables, causing confusions in experimentation. (? Psyhcology in Action,

Current issue

iacalumnus issue xvii 2018 cover

Past issues

DMC Firewall is a Joomla Security extension!