The Troubled History of the E-tazkera (Part 2): Technical stumbling blocks

Original link (please quote from the original source directly):
https://www.afghanistan-analysts.org/the-troubled-history-of-the-e-tazkera-part-2-technical-stumbling-blocks/

 
Author: Jelena Bjelica and Martine van Bijlert
Date: 26 January 2016
 


The introduction of electronic ID cards – or e-tazkera – in Afghanistan remains haunted by delays, obstacles and poor planning. Although CEO Dr Abdullah approved the e-tazkera pilot phase on 3 August 2015, the process has yet to begin. Since then, President Ghani has questioned the very feasibility of this oft-delayed project, while main donors have suspended funding. In this second issue of a two-part dispatch, AAN’s Jelena Bjelica and Martine van Bijlert take a closer look at the technical stumbling blocks.

A short history of the e-tazkera project

Two ministries

The Ministry of Communications and Information Technology (MoCIT) started work on the project in February 2009. Six months later, it formally announced the introduction of an ‘electronic tazkera’ containing biometric information. The aim of the project is to create a modern, uniform and unique identity card for Afghan citizens that would replace the six versions of tazkera (ID documents) currently in circulation. (1) Because the paper-based tazkera system lacks proper security features, it is prone to manipulation and falsification. Moreover, there is currently no centralised system to gather and store population data.

The project is divided between two ministries, which, over time, have competed for a leading role: the Ministry of Communications and Information Technology (MoCIT), which handles the e-tazkera’s software development, design and production; and the Ministry of Interior (MoI), responsible for registration and distribution. In 2011, the E-Tazkera Authority (ETA) was established as the lead agency (based on the Council of Ministers’ resolution no 37) and placed within the MoI. It is also referred to as the E-Tazkera Department.

The contract for the original project, the National Electronic ID Card (e-NID), was won in December 2010 by the private company Grand Technology Resources (GTR). GTR is a multinational, Afghan-Malaysian company, specialising in Information and Communication Technology (ICT) and registered in Malaysia. (2) It was contracted to implement the project within three years, which was an optimistic timeframe. As soon as the contract was signed, the lead ministry at the time (MoCIT) informed the public that distribution would start within eight months – by September 2011.

The introduction of a standardised ID document that could also be used as a voter card or to help establish reliable voter lists, has long been seen as a prerequisite to effectively reducing election fraud. As a result, in 2013, discussions surrounding the process were superseded by a wish to accelerate both the project and distribution of the cards in the run-up to the presidential elections. On 17 September 2012, the Afghan cabinet rejected a proposal by the Independent Election Commission (IED) to launch a new voter registration process, preferring instead the countrywide distribution of the e-tazkera. The registration phase for the ID cards was scheduled to begin in Dalw 1391 (January/ February 2013), although it was clear that, even under the best of circumstances, only part of the population might be issued with an electronic ID-card before the 2014 election. (3) Other delays and controversies stalled the process (see part 1 of this dispatch, published on 25 January 2016).

Following the inauguration of the National Unity Government (NUG) and in light of the upcoming parliamentary elections (now scheduled for 15 October 2016), discussions regarding the e-tazkera and the elections were revisited. However, they were laid to rest when, on 19 June 2015, President Ghani extended the parliament’s term, until then scheduled to end on 22 June 2015. With discussions regarding electoral reform underway and the realisation that the two processes had very different timelines, talks regarding the e-tazkera and elections were pushed into the long grass.

Partners involved

The e-tazkera project is worth over 200 million US dollars, jointly funded by the Afghan government, the European Union and the United States. From this amount, a contract worth 101 million US dollars was awarded to GTR to set up the system and to procure all necessary equipment and software. GTR had previously been awarded two other data-centric contracts by the Afghan government. (4) The remaining funds were allocated to the distribution of the cards and administrative costs.

According to a report by the Independent Joint Anti-Corruption Monitoring and Evaluation Committee (MEC), the Ministry of Finance (MoF) allocated five million US dollars of its own funds for 2015 through the annual national budget. As of September 2015, the Afghan government had spent approximately 64.5 million US dollars on the project. (7.5 million US dollars were spent by the MoI on salaries and 57 million US dollars by the MoCIT to establish the system.)

The exact details of GTR’s contract are not known. In a recent report on the e-tazkera project’s susceptibility to corruption (more below), the Independent Joint Anti-Corruption Monitoring and Evaluation Committee (MEC), which independently monitors and evaluates national and international efforts to fight corruption in Afghanistan and it reports to the public, Parliament, President, and international community, remarked that the committee could not report on the GTR contract “due to the technical aspects of it.” However, it was clear that the MEC had been unable to obtain all the information required to produce a balanced report. This included, for instance, a full list of IT equipment that the MoCIT had contracted, along with necessary specifications. The MEC recommended that the GTR contract “be examined by a professional team of IT experts from an independent third party organisation.”

In its bid for the contract, GTR included a number of technical partners. These included the company IRSI , which specialises in poly-carbonated cards with security features and their integration into the overall system (so far 15 million cards have been ordered, of which seven million have already been delivered); the Korean company HYUNDAI-IT, which specialises in system development methodologies (HSDM) for applications development, system integration and system
security; the company DongDO for the system integration and application development with HYUNDAI-IT; and a company called ENTRUST for the “public key-infrastructure.” According to GTR, the company’s tasks are “to lead the overall project, to provide system integration, client mentoring, coaching and capacity building and to ensure responsible transfer of the project to the government.”

The pilot phase: approved, announced but yet to get underway

The e-tazkera project is divided into four key activities: (a) public awareness and data collection (for an illustration on the challenges of data collection, see this article (5)); (b) citizens’ identity/data verification; (c) biometric registration in enrolment centres; and (d) printing of IDs. Work on the first two activities – data collection and verification – has started at a low level (data for 400,000 citizens in Kabul were collected and some 100,000 verified). In consultation with the International Organisation for Migration (IOM), which provides technical support to the project, the ETA has now proposed a start-to-end pilot project, in which all four key activities in the process will be tested on a small scale.

When planning the pilot project, IOM suggested three options: either to focus on a limited number of areas in Kabul; or, for the project to be implemented in Kabul and four other urban centres; finally, to implement it in Kabul and three districts in different provinces, in order to ensure that the system be tested in different geographic and social settings. In the end, the ETA opted for the first (and least complicated) option. The pilot project would, in theory, take place over 150 days: 60 days for its preparation (although the MoI initially believed 30 would suffice) and 90 days for its practical implementation. The estimated cost is three million US dollars, to be shared equally between the Afghan government and international donors.

The proposal for the pilot project was presented to the president in June 2015 and approved by CEO Abdullah and the Council of Ministers on 3 August 2015. It was initially scheduled to start on 19 August 2015, which coincided with Afghanistan’s Independence Day. However on 18 August 2015, during a meeting at the palace, the president insisted that the day’s focus should be “on commending and decorating the national defence and security forces.” According to a palace statement, it was decided that the e-tazkera pilot be launched at a later date in the presence of officials and notables (English version of the palace statement here;  the clearer Dari version can be found here ). This has not yet happened.

On 18 August 2015, the E-Tazkera Authority organised a ceremony in which dozens of sample IDs were handed out to civil society representatives in an attempt to launch the project.

Although the pilot has been announced, it remains unclear whether the project is technically ready and, more importantly, whether it has the political buy-in it needs. The President himself appears to have reservations, given several orders he has issued to assess the feasibility of the project. Moreover, he has still not formally given the go-ahead for the pilot phase of the project. Setbacks and politicking have hindered the project from the beginning (see part one of AAN’s e-tazkera series available here).

Logistics, weaknesses and concerns

Three assessments

Since last summer, at which point there was still enough pressure to propel the project forward, the President has tasked at least three commissions to advise him on the e-tazkera’s feasibility and its potential pitfalls: in June 2015 he ordered a rapid technical review of the project; in September 2015 he requested an assessment of the project’s susceptibility to corruption; and in October/November 2015 he ordered a more political review of the implications and sensitivities of the e-tazkera, as well as a look at the main technical concerns raised in earlier reports.

The technical review in June 2015 was conducted by a three-member committee and resulted in a long report that flagged various concerns relating to issues including data processing and data security. According to an abridged copy (seen by AAN), concerns included the securing of data transmission and data storage, the possibility of data loss, issues of connectivity (and the possibility that this might exceed the budget) and the lack of a robust testing of the system. However, the scope of the conclusions was limited to technical aspects of the project, as the review focused on what needed to be done before embarking on the pilot.

Secondly, the president instructed the country’s two anti-corruption bodies – the Independent Joint Anti-Corruption Monitoring and Evaluation Committee (MEC)  and the High Office of Oversight and Corruption (HOO)  – to jointly conduct a Vulnerability to Corruption Assessment (VCA). A public summary of the findings, released on 21 October 2015, flagged two key concerns: the nepotism, discrimination, political interference and ethnic conflicts within the ETA; and financial corruption in staff recruitment and procurement. The MEC report concluded with two recommendations: reform of the administrative structure and human resources procedures of the ETA through the introduction of a transparent, competitive and merit-based hiring process; and an independent audit of the GTR contract. (6)

On 10 January 2015, ToloNews released details of the confidential version of the VCA report that they had had access to, stating that:

The report reveals that scores of high school graduates and unprofessional individuals have been appointed in the department without considering their abilities and skills for the posts. “Nearly 27 employees of the e-NIC [department] have been appointed by direct order of former interior minister [Muhammad Umar Daudzai] and his deputies – without going through the employment process. In addition, the leadership of the e-NIC department has been appointed based on the order of the Interior Ministry and the Chief Executive Office,” the report reveals. Only 320 out of 1,046 employees in the e-NIC department have been appointed through free competition, but the remaining officials have been appointed based on direct orders by a number of MPs and former leaders of the Interior Ministry, the report says. The leaked documents also reveal that 519 employees of the e-NIC department were first appointed directly by the MoI and then they were provided the grounds for a competition process and finally they were hired as permanent staffers in the organization. However, e-NIC officials have said they have done this because of the pressure on them.

The third review, which began in October/November 2015, looked into concerns that had been raised in both the technical review and the VCA, to see whether they had been addressed and to advise on the political sensitivities surrounding the e-tazkera (see part one of AAN’s e-tazkera series available here). The commission was headed by Abdul Salam Rahimi, former head of the Office of Administrative Affairs, currently head of Public Policy Formulation in President Ghani’s office. The findings of the commission are not known to AAN. The commission has apparently advised the president that its main concerns had been addressed, however the president appears to remain sceptical.

The slow dissolution of the ETA

E-tazkera staff lose jobs

The EU provided funding for 300 of the MoI employees in the E-Tazkera Authority (ETA), which came to an end in June 2015 (although some 100 e-tazkera employees remained on the payroll until 15 August.) The EU’s commitment of funding until the end of 2015 was based on the condition that it would not cover operational costs. When EU funds ran out, the Ministry of Finance took over the payment of the 300 e-tazkera staff salaries, in addition to the other salaries it was already covering. This came to an end on 20 December 2015 (30 Qows) when contracts ended and were not extended. The e-tazkera project had employed over 1000 staff in the MoI over three years. (7)

Many e-tazkera staff continued to show up for work, hoping to retain their jobs. Some argued they had continued to work despite the fact that the project had long since come to a halt. Zmaray Baher, head of the registration (also termed enrolment) department at the ETA, published an article on the Hasht-e Sobh website on 28 October 2015, in which he described how staff continued to gather bio-data and confirm existing identity documents for the e-tazkera and to digitalise the records of the population registry department (a project separate from the e-tazkera). (8) “Since a few months we have not received a salary,” he lamented. “We face shortages of stationary and other supplies; we face a shortages of forms, which are at the heart of our job; and most important of all, each day the interest of the citizens for this process is waning.”

Once a start date for the pilot project is announced and work resumes, the IOM will be tasked to screen all ETA employees and ensure that future recruitment be transparent (recruitment had initially been one of IOM’s responsibilities, however by the time the organisation got involved, they found all staff had already been hired). The screening will take place during the pilot’s two month-long preparatory phase. This is particularly important, given the findings of the MEC’s susceptibility to corruption report:

Among the problems MEC identified were a lack of written procedures for recruitment; a tendency to ignore the Civil Service Law in the recruitment process; a lack of oversight; and the interference in the process by powerful individuals from the government, parliament, and elsewhere—all of which have created corruption vulnerabilities and facilitated the improper hiring of approximately 70% of the authority’s staff.

E-tazkera staff push for resumption of project

In the meantime, the ETA and in particular its head, Homayun Mohtat, had attempted to continue the project. In September 2015, the Minister of Interior, Nur-ul-Haq Ulumi, sent a letter to President Ghani, saying that if the e-tazkera process, which had ground to a halt in August 2015, was not restarted, over 1000 people would be left without a job.

On 15 December 2015, a local media outlet leaked another letter from the Minister of Interior that was sent to the President on 28 November 2015. It stated that the ETA had lost funding and the MoI faced difficulties paying ETA staff salaries: “Luckily with the exceptional help from the Ministry of Finance and reduction in salaries the ETA managed to find the resources to keep the staff and provide them with salaries until the end of 1394 [21 March 2016]. Unfortunately practical work on e-tazkera has not started and the MoI would not be able to extend the contracts of their staff after the end of December 2015.” This appeared to be another attempt to force the issue on to the agenda.

Additionally, in November 2015, ETA’s head, Mohtat, sent a nine-page document to the President (to which AAN had access), addressing e-tazkera challenges and offering recommendations. He suggested a new management structure, addressed issues concerning staff recruitment raised by the MEC report and in general appealed to the President not to scrap the project. On 20 December 2015, Mohtat also told the press that a new draft plan to improve the identification of individuals through biometric data had been sent for government approval.

On 26 December 2015, ETA employees in a public protest called on the government to start the distribution process. They blamed political disagreements within the NUG leadership for the delays, including the inclusion (or lack thereof) of ethnicity and nationality on the ID cards, and said that getting rid of the employees was a waste of money already spent. Another demonstration was held on 6 January 2016 (see here).

On 28 December 2015, Mohtat wrote on his Facebook page that CEO Abdullah, in coordination with President Ghani, had approved the MoI’s proposal to begin with the biometric data collection, the third key activity of the e-tazkera project (this Facebook page is no longer available.) There, Mohtat explained that in phase three, the families covered under phase one and two would be invited to enrolment centres to record their biometric data (iris scans, fingerprints, digital photographs and signatures) and enter their information into the system. “When the legal and political issues are solved, the data printed on the front of e-tazkera will be finalised and we will print the e-tazkeras.” He also wrote that, “due to the ambiguous circumstances around the start of the project and the lack of sufficient funds, we closed four of our nine field offices and plan to close our remaining offices.” He pointed out that each of the field offices had up to 150 staff, that the equipment had been moved to warehouses, and that the citizen information forms had been shifted to archives. “When the political and legal issues are solved and the contents of the e-tazkera is finalised, we will be able to start issuance without waiting on the other phases. Then attracting international funds will not be a big issue,” he wrote in his closing line.

By this time the ETA leadership’s frustration was palpable. On 28 December 2015, Mohtat told AAN:

This is a dead-end… The e-tazkera is the second of the six commitments of the National Unity Government. It is the mother of all other reforms in the country. Now it seems to be the victim of the government’s mismanagement at the top.

Technical problems, possible glitches

Political issues, delays and public clamouring aside, the project has been plagued by a variety of technical problems and pitfalls.

A current concern is the system’s lack of rigorous testing. Although GTR claims that the system has been tested and is fully operational (GTR tested the system in August 2015 after the Council of Ministers had decided to launch the issuance of the e-tazkera), AAN has been told that this test was done in optimal conditions but only for small volumes. Moreover, the bandwidth was provided by GTR and not by state-owned Afghan Telecom (which does not yet have a contract with the MoI for internet connectivity). It is therefore unclear how the system would perform in actual conditions, or indeed how it would hold up when all enrolment centres begin feeding data simultaneously. The IOM has been pushing for an end-to-end test since the beginning of 2015, but until now GTR has not consented to this. (9) There were some initial glitches in the design of the registration system – for instance, it was not possible to register women as household heads – an issue that may or may not have been sufficiently addressed.

There are also concerns that some of the devices procured by GTR are now out of date. According to the June 2015 technical review of the project, “since the project took considerably longer than anticipated, some of the products that were initially considered were discontinued by the manufacturers, and replaced with newer models. However, there are still some devices, which are announced end-of-life by the manufacturer and will stop providing support, for some devices, as early as May 2016.” The report concludes that “there is no immediate need to replace any devices in the data centre for the pilot,” but it does not mention what might be needed after the pilot is concluded.

ETA’s head Mohtat also informed AAN that some of the equipment had been gathering dust in a basement at the MoI for three years and that the warranty for these machines may soon expire. He said the four printing machines that have been placed in the MoI are supposed to have an overall printing capacity of 32,000 ID cards per 24 hours. This has, however, never been tested.

Finally, GTR was supposed to train e-tazkera employees in software and hardware handling. However, AAN was told that 75 per cent of contracted hours of training have already been used. Suggestions that GTR train a number of master trainers in order to ensure a smooth and continuous transfer of skills have never been acted upon. It seems the impetus for this was lost somewhere between the MoI, the MoICT and GTR. This means that the continuation of the e-tazkera program is currently unlikely to survive the end of GTR’s contract.

Aside from the political challenges analysed in part 1 of this dispatch, the e-tazkera project thus also straddles a large number of technical ones. Although donor interest in a modern and secure national identity card in Afghanistan remains (and is likely to grow) in the face of increasing refugee movements to Europe, it has waned in the face of the many problems surrounding the project. It would be a major feat to revive the technical process, calm the political controversies and rekindle donor interest. While not impossible, the challenges are considerable.

 

 

(1) The different tazkeras in circulation were issued under different regimes. Before the Karzai regime, both 
the Daud Khan (1973-78) and PDPA governments (1978-92) issued versions of the tazkera that are still widely circulated today. Although less common, tazkeras issued during the Taliban regime (1996-2001; two versions), the mujahedin era (1992-96) and under the monarchy (pre-1973) are also still in circulation and are generally considered valid forms of identification, as long they contain photographs and are still readable. A majority of the people (77%), however, appear to hold tazkeras of a post-2001 vintage. These tazkeras have the form of a one-page document in an A4 format, while most pre-Karzai tazkeras took the form of a 16-page booklet. (Source: An Exploratory Study of Afghan Tazkera Ownership, The Liaison Office (TLO), June 2013.)

(2) According to its website, the Grand Technology Resources (GTR) has been registered in Kuala Lumpur since 2006. Mirwais Alizai, the young Afghan businessman who heads GTR, was born in 1979, either in Kabul, Kandahar or Helmand (there are several versions of his biography on the GTR website; the Kabul-born version here , the Kandahar-born version screen shot, and for Helmand-born version see this website). Apart from GTR, Alizai has also established the crude and refined oil company Globix that works in various countries, mainly across Asia and the Middle East, and he has “been investing in the fields of aviation” since 2005. Mirwais Alizai also “supports and contributes to the culture and arts programs and events around the world.”

(3) The Ministry of Interior (MoI), at this point the ministry with the lead on the project, clearly stated that the distribution of electronic ID-cards was “too tough a task to be fully completed ahead of the elections.” The MoI estimated that, realistically, 70 per cent of the distribution could be completed within the next three years, and that for the remaining 30 per cent of the population, mainly living in insecure or remote areas, it could take another three to six years. This was a little over a year before the elections. For more details see this AAN dispatch from January 2013.

(4) GTR’s earlier contracts with the Afghan government were for the establishment of the Afghanistan National Data Centre (ANDC) with the same ministry and for the establishment of the AFMIS (Afghanistan Financial Management Information System) Data Centre for the treasury department of the Ministry of Finance.

(5) This New York Times report from December 2014 erroneously uses the word “census,” while in fact the data collection described in the story is part of the e-tazkera process (national identity cards distribution and census are often used interchangeably, even though they are two separate processes).

The phenomenon of choosing one’s family name, and the confusion it can bring, as described by the New York Times, is not new in Afghanistan. As described by Louis Dupree in the late 1960s (Afghanistan, Princeton University Press: 1980, 2nd edition):

Aside from some literates (mainly Western trained), few Afghans have family names, but call themselves “son of so and so.” Family names, however, relate to the necessity to identify oneself beyond the extended family group for some – but not all – bureaucratic purposes… Family names become necessary, however, to those Afghans who leave the country for overseas educations. Others, particularly writers and scholars, choose to elect personal identifications. Some adapt geographic names: Panjsheri, Ghaznavi or Kohzad (“of the mountains”); some open a book with eyes shut point to a word; still others adopt an adjective which relates to their occupations or interests, e.g., Tarzi (“stylist.”)

(6) The MEC also hinted to AAN that it had suspicions with regard to GTR’s procurement methods, as the MoI never provided MEC with technical specifications for the equipment (serial numbers, part numbers), claiming that “everything had been installed and could not be disassembled again.”

In addition, the MEC report raised concerns that the procedure for obtaining an e-tazkera – after certification by two Afghans and approval of the NDS – made the process vulnerable to corruption and abuse.

(7) According to this article, the organisational structure of the ETA is composed of a general directorate and five sub-departments (human resources, technical, financial, operational, and registration). Out of the 1046 personnel, 194 worked for the general directorate and 852 for the various sub-departments.

(8) According to Baher, e-tazkera mobile teams had visited homes and offices and had already finalised and collected the identity forms of 400,000 citizens in Kabul. He said his teams had started approaching students and staff at Kabul universities and encouraged them to cooperate with the process, despite the uncertainty over its timing and fate. He also wrote that, to date, more than 100,000 Afghans’ identities had been confirmed and that the personal information of half a million (500,000) citizens had been digitalised.

(9) The ETA mobile teams collect personal data from citizens in provinces, after which ID cards will be printed in the capital and sent back to the provinces. Enrolment centres collect the data in urban areas.