Fellow Ghanaians, we are in the midst of the apocalyptic novel coronavirus 2019 (COVID-19/2019-nCoV). Soon, most of our energies may well be absorbed into the task ahead to contain and suppress the virus from becoming a major disease outbreak in this country.
But the issues surrounding the Electoral Commission’s (EC) decision to throw away 7,500 ultra-expensive laptops and accessories plus 72,000 expensive hand-held gadgets, bought and maintained with tens of millions of our hard-earned money, haven’t gone away.
If anything at all, the announcement of the first two cases in Ghana throws into sharp relief whether the “social distancing” directive which discourages large social gatherings can be compatible with the EC’s mass registration calendar, which shall see on average 2200 people cluster in the 8000 registration areas over a period of 43 days beginning on the 18th of April 2020.
That this timeline was technically impossible to meet from the get-go because the EC is yet to even complete its procurement activities for the hardware and core software, much less sign contracts with the relevant vendors, complete specifications, and receive the equipment for customisation, merely compounds the matter.
A pall of uncertainty now hangs over the timeline to the elections, with its strict sequence of activities: registration, exhibition, adjudication, and compilation before distribution of the register to the political parties for their review. Coronavirus, rather than displacing this EC new biometric register matter from our priority list, has in fact reinforced the urgency with which we must handle this situation.
We at IMANI have taken our time to compile this Questions and Answers (Q&A) document as a quick and ready reference for all those who have tracked the outcomes of our reinvigorated advocacy this week.
As we transition from the public phase of our educational effort to the stakeholder engagement phase, we would like to clear any doubts still lingering in the minds of the public because of the EC’s sponsored propaganda, clearly stepped up in recent days. We also acknowledge many of the questions and issues raised by respected and highly experienced public figures in recent days in the line-up of responses we have produced below.
Question: This country has a history of mass registrations (registering every eligible voter/elector afresh) during election years. 2004 and 2012 come to mind. So why the concern about the timing?
- Actually, since 1992, those are the only two electoral cycles during which fresh mass registrations were conducted. In 2004, the EC registered all voters in March and provided them with photographic ID cards for the first time at the total cost of $10 million. There were no biometric procurement issues. The only “sophisticated” gadgets needed then were digital cameras. The population of eligible voters was only about 57% of current numbers. The process thus required only two weeks.
- In 2012, once again, the registration was undertaken in March of that year, but this time round there was indeed a biometric system in place. However, the biometric system procurement process was completed on 4th October 2011, with the award of a contract to STL, nearly six clear months before the commencement of voter registration. And even so, the 2012 elections saw the highest number of technical breakdowns at polling stations since 2000, facts attested to during the election petition hearing. The reason can be attributed to the novelty of the system in Ghana at the time of implementation and the knowledge limitations related to its operation. In short, any sense that the EC is rushing the process should give serious cause for concern.
Question: The EC says the existing system isn’t in a good state for the general elections. They are the ones close to the situation. Shouldn’t we just give them the benefit of the doubt?
- Answer: The EC provided a very specific reason why the system is not fit for purpose. Here is it in their own words: “The BVR kits are more than 7 years old and End-Of-Life (EOL) and can hardly be supported.“ They described all the different parts of the complex network of systems making up their technology platform in a similar way (See: https://www.graphic.com.gh/news/politics/election-2020-new-voters-register-new-biometric-voter-management-system-coming-electoral-commission.html). IMANI has carefully probed this very claim and found it to be untrue. What is shocking is that, to date, no media house has been brave enough to take the EC on for this totally false claim. The evidence of the EC’s mendacity has been provided on multiple occasions and shall be reproduced here too.
Q: Wait, hold it, the EC has produced a letter from STL, the former contractors who worked on the system to prove the obsolescence theory. Do you know better than STL?
- Well, the EC says STL was milking Ghana, charging $3.9 million per year in annual fees for doing very little. We tend to agree, like most people, that STL took advantage of lax oversight at the EC to overcharge. The previous Chairperson of the EC made the same arguments during her procurement-related impeachment proceedings. She talked about renegotiating several contracts signed by subordinates on her arrival at the EC etc. The current EC Chair and her deputies have also indicated that they came to see some STL-related contracts, which they eventually abrogated.
- These are among the various reasons why STL’s contract was abrogated by the current EC bosses four months after coming into office. It is strange that STL’s unexecuted contracts, already dismissed as extortionate, are now being used to justify fresh overspending. STL has been gone for 15 months now. During that period, the EC never bothered to go to the actual companies who were providing the technology – HSB and Genkey – and ask them directly what it would cost to continue upgrading the system.
- STL was, and has always been, a mere “middleman” in this context. It is not surprising that they will “pad” any quotations. HSB and Genkey, the actual Intellectual Property (IP) owners of the core platform are experienced biometric hardware and software providers whose products are certified to the highest standards according to the standards of the International Biometrics and Identity Association (IBIA) and are conformant with the CBEFF ISO standards. Evidence coming to light shows clearly that the EC made no effort to deal directly with HSB and Genkey to maintain and enhance the existing system after they had sacked STL.
Q: STL may have padded their quotations and may have been milking Ghana as many have accused them over the years, but that does not still mean that the biometric system is not obsolete.
- We will repeat once more for emphasis: the EC’s system is a “network of multiple equipment and software”. It is akin to a banking system (core software, ATMs, POSs, teller desktops etc) or a television production system (cameras, mixers, microphones, editing suites etc.). Different parts of the system tend to be overhauled at different paces. The system as a whole DOES NOT DATE FROM 2011. That claim was a blatant untruth. The country has spent over $60 million at least buying new equipment, software and licenses to upgrade, maintain, refresh and keep it up to date. Some of this equipment was even delivered after the present EC bosses came to office.
- According to a very connected journalist, this new cadre of EC bosses have taken delivery of as many as 150 of these very expensive devices. If we use historical procurement patterns at the EC as our guide these additional devices and accompanying software would have cost not less than $500,000.
- In 2016, the entire set of 7000 laptops-with-accessories (called BVRs) was refurbished (note “refurbished”, not repaired). And 500 brand-new BVRs were purchased. 30,000 BVDs (the handheld gadgets used for identification) were refurbished and 40,000 brand new ones purchased. In that year, spending on technology hit roughly $50 million according to the Auditor General of Ghana. In 2018, according to the Ministry of Finance, 2000 more BVRs were purchased, and some old machines apparently retired.
- If one conducts a portfolio lifecycle analysis, one can only conclude on the face of the evidence that the system is very up to date. Because BVRs come with licensed software installed already and the EC has also been updating the central software packages (the VMS) regularly, the system currently in place is pretty new overall.
- Once this point sinks in, the fallacy in the EC’s foundational argument becomes all too clear. If the system is not obsolete, then there is no basis in replacing every single equipment. One may replace a subset of the portfolio periodically because breakages do happen so one can buy some equipment over time to replace truly broken-down ones.
- The EC’s projected cost for refurbishing the system, which it says it got from STL, $74 million is wholly flawed because it is based on assumption of having to replace every single equipment and that STL shall intermediate the purchases. Since that is not necessary, and only a portion of the system needs replacement, the real cost of preserving the existing system for this electoral cycle is “only” $15 million, and not $74 million. This would mean 2000 more BVRs, 5000 more BVDs, data centre, and auxiliary service.
- On the other hand, the EC’s claims that buying a brand-new technology platform will only cost $56 million has been exposed to be untrue by the outcome of its just ended sham tender, which produced $72 million as the winning bid.
- The EC’s allies claim that renegotiation shall bring this amount to close to $60 million. Nevertheless, there are still squabbles, per our sources, over incoterms and other auxiliary costs, and the $59 million purportedly agreed is thus merely the floor price. At any rate, this is just the hardware cost. Adding the software procurement, datacentre and fibre-optic network costs may well balloon the amount to over $80 million. This is the real cost of the decision to completely replace the technology system.
- Though the EC refuses to acknowledge that a limited voters’ registration exercise would be much less costly than a full, mass, registration, the evidence suggests otherwise. A limited registration exercise shall involve only the two million max of voters who have turned 18 or couldn’t register in last year limited registration. A limited registration on the other hand should cost less than $20 million, instead of the roughly $70 million the EC plans to spend on the mass registration.
Q: Well, this is rather surprising. Which official documents are you relying upon to make these conclusions?
- : Page 18 of the Ministry of Finance’s 2019 Program Based Budget analysis of the EC. Page 12 of the 2017 version of that same document. Pages 8 and 9 of the 2019 report of the Special Budget Committee, a parliamentary oversight body that reviews the EC’s spending, and page 17 of the 2020 version of that same committee’s report. Pages 145, 146 and 153 of the Audit Service March 2019 report to Parliament covering the EC and other major institutions.
- Triangulating all these various sources lead to plain and incontrovertible verbs that the EC’s claim of the system dating from 2011 is a COMPLETE UNTRUTH. If the media were to hold the EC to account for deceiving the public so blatantly, the notion that we should give them the benefit of the doubt wouldn’t even arise. How do you give the benefit of the doubt to an organisation insistent on denying the public the truth?
Q: Okay, but surely, this is probably just a mere case of misunderstanding and confusion. Maybe, some monies have been listed as released to the EC that never got disbursed or maybe wasn’t used?
- A: In reviewing the records we paid attention to indications that a particular expenditure is “past and complete” and not merely an “allocation” or even “disbursement”. For example, the Audit Service’s report captures specific dates on which the EC transferred monies to vendors, sometimes complete with payment voucher numbers. The evidence is virtually fool proof and ironclad.
Q: What about the claims that the EC mentioned to parliamentary committees several times that the biometric system was obsolete and that even the just exited Chair of the commission signed some contracts which have not been honoured? Surely, that can get captured and picked up by even auditors to create an erroneous impression of procurement?
- Yes, in 2018, the EC appeared before the relevant parliamentary committee and insisted that it needs additional BVRs. But note carefully: it confirmed on pages 8 and 9 of the resultant parliamentary (Special Budget Committee) report that all 7500 BVRs in its portfolio have been refurbished but that it still needs an additional 5000. In addition, it wanted to procure 5000 BVDs.
- This was clearly in view of the expansion of the number of polling stations and the need to maintain an excess margin in respect of the “Two BVDs per Polling Station” policy. Hence as late as 2018, according to the parliamentary record, the EC very much intended to keep the existing equipment. They just wanted to buy more items. In fact, in 2019, when they returned to the committee, they assured the members that the biometric system in its current state was fit to manage the upcoming elections. The only part of the EC’s system that it has been consistent in calling for a replacement is the datacentre. The contract for that component has already been awarded.
Q: And the claim that the previous EC Chair signed contracts that were abrogated by the current management?
- : Not sure what that argument is meant to achieve, except to mire this whole discussion in distraction. We have indicated clearly that equipment has continued to be bought and to be received at the EC. So it is obvious that the previous EC Chair bought quite a quantity of these machines.
- In fact, due to the budget shortfall issues that other commentators have mentioned in recent days, the EC started to repurpose operational funds for use on capital expenditure. In fact, in the last electoral cycle, the EC spent almost 11 times its budgetary allocation for capital expenditure by diverting operational funds which, clearly, had been over-allocated.
- They had also planned to increase the number of devices for sure, but the controversy today is not about the propriety of buying more machines to augment the present portfolio. It is about throwing away perfectly functional machines in order to redesign and rebuild the whole system from the ground up based on false claims of obsolescence.
Q: Could it be that the recent EC is just distrustful of any system left behind by the previous administration and wants something she can be fully confident in as the procurer?
- : Please understand that the EC terminated the contract of the old contractor, STL, four months into office. She then paid new consultants to take over the system. That is what she has used to run a range of elections so far including the referendum. There is no validity to any such paranoia about an unreliable legacy system. Mere sentiment cannot be used to throw away a system valued at well over $60 million in order to spend $150 million buying a new one and collecting fresh data. $150 million is 85% the total value of recent annual allocations to the Road Fund, 70% of the value of the GET Fund, and 50% of the value of the National Health Fund. We are talking about money that can have massive impact on health, education and infrastructure.
Q: A major criticism is that you keep flitting from issue to issue, and changing the goal posts, all in a desperate bid to scuttle an activity of a major state institution.
- That logic is strange. We have been very clear from day one what our issues are. The matters that divide major contributors to the ongoing debate are very clear. For even more clarity, we have illustrated the divergence in opinion below.
- It should be easy to see that IMANI’s position has stayed very much the same.
- B. We believe that there are two main dimensions here: the EC’s plans will cause a loss of nearly $150 million to Ghana, but beyond monetary losses the plans also put our elections at risk because the EC is rushing so much it is doing shoddy procurement work. On top of that, it is rigging tenders to achieve a predetermined outcome.
- C. Both dimensions – cost and risk - have the prospect of damaging the health of our elections. We have already talked about the compression of timing. A compression predominantly due to the bungling of tender processes that began as far back as April 2019, but have been aborted and restarted several times with a plain end of achieving some preferred outcome. The sham tenders merely compound the risk factors.
Q: The EC and its allies deny any procurement issues. They say they acted within the law.
- The fact that the law gives an institution power to perform some action or activity does not mean that it can undertake it anyhow it wants. The constitution, for instance, imposes a reasonability threshold in Article 296. Furthermore, all institutions are subject to public audits and financial responsibility statutes.
- The EC’s hardware tender that was eventually awarded to Thales suffered the illegitimacy of the Chairman of the Tender Committee, himself a longstanding member of the Electoral Commission, in fact one of the seven commissioners, writing to the Central Tender Review Committee (CTRC) and the Public Procurement Authority (PPA) to disown a tender evaluation report bearing his name and the names of the other panel members. He apparently refused to participate any further in the process in protest.
- The EC then reconstituted a new committee, which rushed through a purportedly brand-new process to award the $72 million contract within a week. The classic safeguard of technical evaluation taking place without the evaluators knowing the financial offers of the bidders was breached because the financial offers had been opened one week, i.e. 3rd January 2020, before the new committee was created. The new committee went ahead to disqualify one of the world’s leading biometric solutions providers, Idemia, and then set about discrediting the other highly qualified bidder, Smartmatic, thereby paving the way for Thales, and Rhema Systems, their local partners to win. Remember this was exactly the outcome the EC Chair had tried to force in her 18th December 2019 memo. (See the full chronology of events posted above and below to better appreciate this point).
- A most peculiar revelation in recent days, by a member of the EC’s Entity Tender Committee is that the process of scoring the technical bids in the EC’s recent hardware tender somehow managed to take place without factoring in any due diligence on the bidders.And that the reputational risk defects of one bidder was indicated in the report but room could not be found to do same for the other. The final report used to recommend Thales for the contract award incorporated references to allegations against Smartmatic in the Philippines but refused to contend with far more damaging, and conclusive, records of wrongdoing by Thales, confirming the absence of an equal playing field.
- Thales (and its subsidiary, Gemalto) are being sued for bungling the national ID system of Estonia; were once blacklisted by the World Bank; and are the subject of an active indictment in the ongoing Guptagate scandal in South Africa, concerning bribery at the highest levels of the South African government. A history of credible findings by competent and authoritative bodies, not mere allegations, of corrupting African elections officials was also ignored during the scoring.
- In all this, it is important to bear in mind that the software tender, also published in October 2019, ended up being cancelled. Our sources tell us of attempts to sole-source the contract in the custom market, ruling out the prospect of being able to develop the relevant software in less than 6 months, and thus forcing the eventual subcontracting of the software job to Thales, a proprietary systems developer that can simply not hand over its source code to the EC, in the fashion contemplated by its inexperienced IT procurement expert. Thus, a major basis for this whole full replacement strategy, the notion of root control of the system, turns out being worthless.
Q: The argument has been made that Thales prevailed in an open and fair process and that none of the other bidders, Smartmatic or Idemia, was graded on a scheme not applicable to the others.
- On 18th December, the EC management tried to force the original tender evaluation panel to eliminate Smartmatic and Idemia by way of a memo providing comments to the committee’s draft report. When that failed, they dissolved the committee.
- The new committee then found a way to eliminate Idemia on January 13th 2020 on technical grounds thus ensuring that Idemia never proceeded to the financial evaluation round. Idemia manufactures the Morpho devices recently used in the Kenyan elections and others around the world. In our pricing analysis, we found the devices to be one of the most cost-competitive in the world.
- In all honesty, it is not imaginable that in a rational evaluation, Idemia will be eliminated in a technical evaluation round also involving Miru and Buck Press, which have far limited elections technology deployment experience. Or that Miru will obtain higher marks than Smartmatic, which has 10 times Miru’s experience, and three times the elections technology deployment experience of even the eventual winner, Thales. Miru’s only comparable previous project was the one in DR Congo. An earlier attempt in Argentina was ended when the government found serious vulnerabilities in their systems.
- When the EC says that they threw away the first report because the marks awarded bidders were too lopsided and unbalanced, one has to ask how a non-technical EC management can claim to have more insights than technical panel properly constituted to evaluate offers for several weeks. And whether an organisation that cannot even run a tender without confusion and rancour can smoothly implement a radical overhaul of a complex technology system within five weeks, the timeline to the EC’s own self-imposed deadline to commence voter registration.
Q: Your comments have been interpreted as being in favour of some of the vendors that lost out, perhaps Idemia and Smartmatics.
- Those inferences are plainly ridiculous. IMANI begun its opposition to the procurement of a new system in December when all these companies were busily bidding to win the contract. How can IMANI then be working for some of the companies, when our actions have been to deny them any money in the first place? Let’s repeat: the procurement issues are related to the risk factors. If the EC is mismanaging the procurement, it is also mismanaging the elections. If the EC is trying to rig a tender, then its motives for pushing for a brand new system becomes suspect.
Q: But the EC claims that these matters that you are busily canvassing did not even feature in the evaluation?
- That is also ridiculous. No $72-million-dollar project can be awarded without considerable due diligence involving background checks, reputational risk evaluation and track record research. So, the notion that these areas were merely explored in jest for some of the bidders and ignored for others because they are inconsequential for all makes no sense.
- On the scoring sheets, there are multiple areas where the vendors are instructed to show that the equipment comes from REPUTABLE suppliers. On page 31 of the final tender committee report, it is plainly stated that due diligence of Smartmatic indicates an “unfavourable business profile”. The notion that this major finding in a section clearly labelled: “Weaknesses of Bid” on page 33 did not play any further role in scoring the bids is too ridiculous to even deserve further examination. And on page 28 of the 94-page report, the writers are emphatic that they did “NOT SIGHT ANY ADVERSE FINDINGS” on Thales. This was a categorical statement.
- Because the qualitative inferences must influence the quantitative scores, we must look to them as providing us with a picture of how the evaluators saw each bidder. That is the only way to interpret the even sparsely annotated scoring marks. In 75% of cases, no remarks are given to explain scores. In one instance a bidder was scored zero because their solution did not “go beyond minimum requirements”. What does this mean? One bidder had their biometric verification scanner penalised for not matching the 4-4-2 configuration, yet that fact had already been addressed by penalising them through a modification of the financial offer to accommodate the preferred spec.
- The entirety of the section 5 of the scoring scheme is dedicated to reviewing the quality of each bidder’s track record. Had the proper thing been done, Thales would have lost virtually all the marks it earned there due to the widely reported adverse findings related to its electoral technology deployment record in places like Gabon and Benin, where an investigative finding confirmed that its Gemalto unit had been paying monies majority owned by a member of that country’s Electoral Commission and his relatives. Adjusting its marks that way though would have lost it the bid. Which clearly could not be countenanced by the powers driving the tender.
Q: You have mentioned the term, “conflict of interest”. Where and how?
- We have it on very good authority that Fairgreen Limited, whose CTO has been serving as the EC’s technical expert on IT procurement and who has sat on the EC’s tender evaluation panels, has been suppliers to the EC and continues to pursue datacentre subcontracting contracts. It is frighteningly mind-boggling that a procurement expert that stands to benefit from the direction and content of its advice also becomes a supplier or subcontractor to the same entity it is advising.
Q: There are some that argue that all this talk about technology obsolescence is besides the point, the real issue is the “bloated character” of the current Voters’ Register.
- Whilst this is a popular view, there is no evidence, statistical or otherwise, that the ineligible names on the Register are anyway particularly skewed towards any particular political party. Without evidence of skewedness, any ineligible names can be considered as randomly distributed. The fact that victories in Ghanaian elections have vacillated over the last two decades reinforces the notion that there is no skewedness. Also, the trends in turnout numbers do not show in anyway a peculiar characteristic of this current register.
- The strong pattern discernible in turnout stats analysis is the tendency of turnout to decrease in elections in which the incumbent political party loses. So whilst the ~64% turnout observed in the 2012 elections was lower than in the previous year, it was higher than the 2000 figure of ~61%. And, as further evidence that low turnout is not some distinct attribute of the current calendar, we can point to the referendum figure which was in the region of ~85%.
Q: So why doesn’t IMANI go to court?
- There are public agencies resourced from our national purse to enforce the law and avert wrongdoing. CHRAJ, EOCO, CID, Attorney General’s Department, Auditor General etc. etc. The law and order dimensions of this matter must be directed to them, and we are happy to join the media and other CSOs make the case why they should act fast. But our primary mandate as an organisation is public education. It is even in our name. We do not have the resources to be junketing from law court to law court forcing public institutions to do the right thing. But we can educate the public and media and hopefully increase the pressure for the institutions set up and funded to do that work to sit up and act.