Business, News

Meet the Man who Brings Big Data to Indonesia’s Tax Office

ShareMeet the man who brings big data to Indonesia’s tax office. Two years ago, a district tax office (KPP) in Surabaya, East...

Written by Erwin Prasetyo · 3 min read >
Big Data

Meet the man who brings big data to Indonesia’s tax office. Two years ago, a district tax office (KPP) in Surabaya, East Java, would need four days just to retrieve a taxpayer’s data from servers at the Jakarta headquarters because it had to wait for the information to be manually extracted, processed, then flown-in on compact discs.

Big Data
Tax office director for digital transformation Iwan Djuniardi. Meet the man who brings big data to Indonesia’s tax office (JP/Norman Harsono)

“The headquarters couldn’t email most of the data either. It was unsafe. And much of the data’s file sizes were huge so it couldn’t be sent via email,” said tax office director for digital transformation Iwan Djuniardi during an interview with The Jakarta Post at his workplace in Jakarta.

The headquarters was, in fact, forbidden from sending tax data through any public file sharing vendor due to the 1983 Tax Law Article 34 that stipulates tax office staff cannot share such data with “outsiders”, whom directors interpret to include vendors.

Thus, when Iwan took office in 2015, he decided the tax office needed to build its own big data system that would automatically compile, analyze and report data for tax officers nationwide.

He said the system, dubbed Social Network Analytics (Soneta), was rolled out nationwide last year and reduced the KPPs’ waiting time to a matter of seconds.

Iwan’s ultimate goal is to raise Indonesia’s tax-to-gross domestic product (GDP) ratio, which has been steadily declining from 14.3 percent in 2013 to 10.7 percent in 2017 before rebounding to 11.5 percent last year, based on Finance Ministry data.

Indonesia’s tax ratio in 2017 was also the lowest among Southeast Asian peers Malaysia (13.1 percent), the Philippines (14.2 percent), Singapore (14.2 percent), Thailand (14.8 percent) and Vietnam (19.1 percent), according to World Bank data.

According to the International Monetary Fund, developing countries like Indonesia need to sustain a minimum 12.75 percent tax-to-GDP ratio for 10 consecutive years to increase GDP per capita by 7.5 percent.

The Finance Ministry, which oversees the tax office, would come to write in its 2016 annual report–a year after Iwan became director–that modernizing the office was one of five key initiatives to improve tax collection at the helm of Minister Sri Mulyani Indrawati.

“But first, we had to learn,” said Iwan, whose team comprised mostly trained accountants while he himself was a mechanical engineer. “So I sent five of my staff for training with Solusi247 [a big-data company]. After they finished training, I put them to the test.”

In May 2015, the reskilled staffers tested a prototype data warehouse built from 10 second-hand computers connected to a single monitor. It ran on open source big data software Hadoop and had a humble storage capacity of 10 terabytes but laid the foundation for the office’s technological back-end for the years to come.

In addition, the prototype cost zero rupiah in hardware expenses yet uncovered Rp 1.3 trillion (US$90.9 million) in potentially unpaid taxes just by better sorting data compiled from tax return forms and company invoices, Iwan recollected.

Armed with positive results, he convinced the House to quintuple his department’s annual budget for the following year to Rp 1 trillion, most of which would be spent on new hardware and software.

In 2016, Iwan’s department received the bigger budget and Minister Sri Mulyani announced a plan to develop a complex application, dubbed the core tax administrative system, that would automate most tax office administrative work.

“Our taxpayers have more than tripled in number and our offices have increased. Considering the number of registrations, payments and management, we need to upgrade our IT system,” she said back then, as reported by media outlets.

Iwan told the Post that the development of the administrative system, which will serve as the tax office’s technological middle-end, was outsourced to an external company while his department focused on developing back-end systems (the data warehouse) and front-end systems (taxpayer-facing services).

With financial support endorsed by the House, Iwan and his team pushed the warehouse prototype through three evolutions over the past three years. At each stage, it would acquire faster processing power, more functions and a catchy acronym.

Stage two, for example, was named Cendol, short for Centralized Data for Online Analytics and a reference to a traditional dessert. This technology, which was rolled out nationwide two years ago, ended the need to deliver compact discs, said Iwan.

Soneta, a reference to Rhoma Irama’s 1970s dangdut band, is the latest evolution, which appears as a dashboard application that not only shows taxpayers’ data but also any suspicious activity between it and other taxpayers.

Iwan said Soneta could detect 30,000 suspected fraud cases each week whereas manual checks could only detect a hundred each year. The cases, if all proven fraudulent, would be worth Rp 2.9 trillion in missed tax revenue.

Interestingly, Soneta also tracks the search activity of tax officers. The system shows how often each officer accessed the data of taxpayers who were not within their jurisdiction. If the frequency was high, they may be a mole.

Yustinus Prastowo, the executive director of the Center of Indonesia Taxation Analysis, said such big data technology was a critical tool to improve tax compliance.

“The tax office has a lot of data ready for use yet its usage is still minimal. The tools available to tax officers today are still insufficient to boost tax collection on the ground. Developing these tools should be the tax office’s task,” he told the Post in a short message.

One of Iwan’s deputies, who joined later in the interview, said the department was working on adding publicly available information from Facebook, Twitter and Instagram into the system. Such information could help verify relationships between fraud suspects.

But Iwan and his team have much more in mind for the system. They, for example, plan to introduce new data sources and develop sharper data analytics until the system can not only detect but also predict fraudulent behavior.

“My vision is to make the tax office a digital native institution,” Iwan said.  Meet the man who brings big data to Indonesia’s tax office (Norman Harsono, The Jakarta Post)

Leave a Reply

Your email address will not be published. Required fields are marked *