Page 47 - CMA Journal (July-August 2025)

Focus Section

Table 1: Tax-to-GDP Ra os and Data Integra on Status (2023–24)

Country Tax-to-GDP (%) Data Integra on Level Key Tool(s)
US 27 Advanced IRS Data Matching, AI anomaly detec on
UK 33 Advanced Connect, mul -source risk profiling
China 18 High Real- me e-invoicing, AI compliance
India 17 Medium–High GSTN, e-way bills, invoice matching
Bangladesh 11 Developing EFDs, VAT–Income Tax integra on
Pakistan 9 Low POS, Track-and-Trace, par al NADRA data

Pakistan: Milestones, Bottlenecks, and Potential tax-relevant datasets, harmonize standards, and
improve integration. The UK’s HMRC Connect
Pakistan has expanded e-filing platforms. In 2023, over
platform merges over 30 datasets, showing how
4.5 million taxpayers filed electronically. While modest in effective centralized governance can be. Pakistan
percentage terms, this reflects a functional digital base.
could potentially close compliance gaps worth 0.5%
Point of Sale (POS) integration in retail records of GDP through such reforms.
transactions in real time. Though initially limited to large
retailers, it sets a model for broader expansion. 2) Data Quality: The taxpayer registry contains
Track-and-Trace for tobacco, sugar, and beverages duplicates, outdated addresses, and incomplete
creates production and sales trails, reducing corporate ownership information. India’s GSTN
underreporting. Frameworks are in place for wider sector cleaned over 1.8 million forged registrations through
application. Initial agreements with NADRA, SBP, and systematic validation. Pakistan could replicate this by
SECP allow cross-verification of taxpayer data. These form integrating with NADRA and SECP, unlocking 0.5–1%
the legal base for real-time integration. of GDP in gains.
Bottlenecks
3) Third-Party Integration: While Pakistan has
FBR platforms (POS, IRIS, Track-and-Trace) operate agreements with NADRA, SBP, and SECP, these lack
separately. There is no central data warehouse real-time execution. Integration with utilities,
consolidating information from NADRA, SBP, SECP, and provincial tax systems, and telecom operators is also
provinces. Duplicate CNIC-linked records, outdated inadequate. In the US, the IRS processes over 5 billion
addresses, and incomplete corporate ownership data third-party information returns annually, making
reduce the reliability of analytics. Limited analytics evasion structurally difficult. Pakistan could generate
training means advanced tools are underused. Manual 1.5–2% of GDP by developing secure APIs for
audit selection remains dominant. Integration faces seamless integration.
pushback from vested interests resisting transparency.
Taxpayers perceive FBR as inefficient, discouraging 4) National Tax Intelligence Platform: A centralized
voluntary compliance. warehouse could consolidate FBR, NADRA, SBP, and
SECP data. China’s State Taxation Administration
Data-Driven Tax Reforms for Pakistan
integrates VAT invoices, customs, and banking data in
To overcome these bottlenecks, Pakistan must transition real time, supporting an 18% tax-to-GDP ratio. A
from fragmented initiatives to a cohesive, data-driven similar BI system in Pakistan could generate 0.5–0.8%
ecosystem. Global experience illustrates that integrated of GDP.
data governance, real-time analytics, and risk-based
5) Risk Profiling: AI can replace manual audit selection
enforcement can expand the tax base and improve
compliance without raising tax rates. with targeted, high-accuracy enforcement. HMRC’s
AI-driven risk engine flags most high-risk taxpayers
1) Data Governance: A Corporate Data Office (CDO) before audits, improving efficiency. Pakistan could see
within the FBR could unify the management of 0.8–1% of GDP in gains by adopting this model.

Table 2: Pakistan’s Reform Status

Area Current Status Gap
Data Governance Par al No Corporate Data Office
Third-Party Integra on Par al No real- me APIs with NADRA, SBP, SECP
Data Quality Weak Duplicate and outdated taxpayer data
Risk-Based Audi ng Low Manual audit selec on dominates
Advanced Analy cs Minimal Limited AI or machine learning tools

ICMA’s Chartered Management Accountant, Jul-Aug 2025 45