It was famously said by Benjamin Franklin, “In this world, nothing is certain except death and taxes”. Perhaps Benjamin was unaware of the lagging efficient administration, shallow policy orchestrations, and bad governance, a byproduct of rent-seeking prevailing in our country. Here, businesses and mogul individuals, while conniving with the higher-ups, strive to remain in the grey area to circumvent the certainty of taxation’s axe.
The available economy (documented economy) is 64 per cent smaller than the addressable one (undocumented). This cul-de-sac of formalism is not a naturally occurring ailment; rather, it is the product of a symbiotic relationship between the wealthy and the power-holders – if they are distinctive anyway – shattering the fiscal account of our country aeon upon aeon.
The recent IMF delegation visit and revenue shortfalls (Rs386 billion in the first half of the fiscal year), despite a considerable increase in collection vis-a-vis the previous year, signify albatross structural failures and greater lacunae in our fiscal canvas. According to the International Tax Competitiveness Index 2024, Estonia ranks as the best-performing country, while Colombia ranks the worst. However, Colombia has a Tax-to-GDP ratio of 22.2 as per the OECD Revenue Statistics 2024. In comparison, Pakistan lags even further behind in terms of its tax-to-GDP ratio, at a mere 9.2 per cent.
Although Pakistan is not included in the index, one can assess its deviation from the OECD’s general principles of good taxation by examining the proportion of indirect taxes (including the withholding tax regime, as also highlighted by the AGP) to total tax collection, which has consistently hovered around 85 per centover the past five years. Two problems ensnaring our revenue mobilisation functions are the narrow base, which results in mostly taxing the already taxed, and the disproportionate reliance on presumptive regimes, commonly referred to as ‘withholding-isation’. This is evident from the fact that, while the latest Economic Survey reports a labour force of 72 million, only around five million taxpayers are active (with 70 per cent of them showing zero taxable income).
Second, there is a lack of adequate taxation. Over the years, there has been no commendable growth in collection after factoring in inflation, and it has lagged behind economic growth. Furthermore, despite repeated calls from MLDAs and local experts, sectors like agriculture, retail, and real estate have consistently evaded taxation. Even now, as per the proverbial National Fiscal Pact, only Punjab has adequately legalised the taxation of agricultural income, while other provinces remain in a state of trance – underscoring the incumbents’ reluctance to risk their political capital by addressing these entrenched interests.
There needs to be a radical transformation; without it, dramatic re-engineering of FBR processes and dealings is not possible. Here, technology could be the enabler of the much-needed reparation. Since the launch of ChatGPT in November 2022, it has caused a stir, propelling artificial intelligence (AI) to the forefront of innovation. AI adoption has more than doubled in five years, contributing $4.4 trillion annually to the global economy, according to McKinsey.
A recent IMF technical note, titled ‘Understanding Artificial Intelligence in Tax and Customs Administration’, outlines how AI is transforming tax administration by enhancing risk management, investigations, enforcement, and taxpayer services. Established use cases, such as case selection for audits, collections, and refunds risk assessments, leverage machine learning (ML) and expert rules (ERs), offering high potential value and proven effectiveness.
For taxpayers, AI promises to simplify tax filings through platforms that integrate with financial systems to calculate liabilities and generate returns. AI-powered solutions can also assist in alerting users to new regulations and tax rate changes, thus avoiding non-compliance. Virtual assistants can interpret laws and answer questions in real time, reducing the complexity of compliance. These bots have the potential to demystify the otherwise labyrinthine taxation system in an easy-to-grasp manner, thus facilitating compliance among potential taxpayers.
AI applications like social network analysis (SNA) and predictive models can also expand the tax base by identifying non-compliant entities and spotting patterns in the informal sector, thereby bringing new taxpayers into the fold. Further, AI can undo pathological morality during audits by reducing human interactions, curbing bias, and enhancing objectivity. For tax collectors, emerging use cases promise significant advancements in areas like desktop productivity, content creation, and staff training. These tools can automate routine tasks, enhance staff education, and streamline administrative functions, ultimately improving the overall efficiency of tax administrations. AI can also prioritise high-risk anomalies for scrutiny, reducing manual audits and ensuring fair, data-driven decision-making.
Spain and Greece serve as examples of AI integration in tax administration. The Spanish Tax Agency employs AI-driven virtual assistants to address common taxpayer queries, such as filing deadlines, VAT, and personal tax issues. AI also pre-fills tax forms using employer-reported data, reducing errors and saving time, while aiding in fraud detection through pattern analysis.
Similarly, Greece’s Independent Authority for Public Revenue uses AI-powered virtual assistants to provide real-time, reliable taxpayer support and streamline administrative processes. AI in Greece also performs automated error checks and identifies discrepancies in filings, enhancing compliance. Both countries showcase the global trend of AI transforming tax systems by improving efficiency, reducing errors and prioritising high-risk anomalies for scrutiny.
Effectively leveraging AI in Pakistan’s taxation system is crucial for expanding the tax base, improving enforcement, and addressing the inefficiencies of extractive regimes. It can work as a double whammy: roping in licit businesses from the undocumented economy into the fold and curbing illicit businesses, thereby optimising tax collection and promoting fairness.
However, a holistic adoption of AI in taxation also demands significant investment in digital infrastructure, workforce training, and robust data privacy laws to build trust among taxpayers and ensure long-term effectiveness. While recent efforts, including the establishment of high-level committees on AI and data analytics, are promising, the real test lies in implementation. To ensure success, these initiatives must move beyond bureaucratic exercises and deliver tangible reforms.
The integration of AI could stabilise the economy and help repair the fiscal account, but only if accompanied by a commitment to transparency and reform at every level of governance. The question remains: will we rise to the challenge or remain mired in our fiscal inertia?
The writer is a Peshawar-based researcher who works in the financial sector.
The writer is a retired ambassador and author of ‘Pakistan and a World in Disorder – A Grand Strategy for the...
“Our adversaries are intensifying their efforts in the entire under-developed world. Those who oppose their advance...
Pakistan’s return from deep economic slumber will continue and succeed, mainly with the private sector assuming its...
On January 8, 2025, Indian Foreign Secretary Vikram Misri met with Mawlawi Amir Khan Muttaqi, acting foreign minister...
The classic coup d’état has guns. Uniformed men run wild seizing government agencies and claiming control over what...
President Donald J Trump likes teasing out the unmentionable, and the Israel-Palestinian situation was hardly going to...
‘The latest contribution by Ali Usman Qasmi to the existing scholarship on the history of Pakistan and how religion...
History has shown that education, particularly the presence of quality teachers, plays a crucial role in national...