A Wasteful Enterprise
On Friday, August 7, the Electoral Commission of Ghana called a press conference to pat itself on the back for completing the latest in the long line of registration exercises conducted in this country.
As if mesmerised by its own achievement, the EC lunged into a series of misguided statistical commentary about the numbers it recorded, in obvious response to critics who argue that the millions of dollars spent on this needless mass registration exercise have mostly been effectively wasted.
Yet, evidence that the EC has wasted more than $185 million of the country's money (over $70 million for new hardware, software and datacenters to Thales and others; over $60 million of near- brand new equipment discarded; and at least $65 million for goods and services) for no clear gain whatsoever is firm and solid. As we have persistently argued, at most $35 million would have been required had the EC listened to sound counsel and used the existing infrastructure. The implied loss to the state of persisting in its course thus amounts to over $150 million.
Electoral Commission Spending on the Registration
The EC made two main claims to justify the need to spend all these millions that this COVID-hit, struggling, country can ill afford: that its equipment is obsolete (in its media campaign) and that the existing register is somehow not credible due to the presence of ineligible names (in the case it presented to the Supreme Court).
The first claim was thoroughly debunked by IMANI in several reports citing parliamentary records and auditor-general reports (the reader may review the evidence on the IMANI website: https://imaniafrica.org/2020/02/24/imani-the- dangerous-games-of-the-electoral-commission/ ).
The second claim was repeated in the EC's August 7th press briefing. Clinging to straws, the commission sought to give a modicum of respectability to the dangerous idea of “malicious bloating”, a wholly unsubstantiated view that somehow the register has been stashed with ineligible voters, especially foreigners, in a manner designed to benefit some specific political parties.
In this statement, IMANI will go over the claims and the evidence step by step to kill this pernicious idea once and for all and to establish firmly and clearly that absolutely NOTHING has been achieved by the wasteful decision to compile a new register. A dangerous precedent has been set and any future EC can use any reason to embark on a new mass re-registration exercise. The courts having vacated their responsibility to insist on administrative rationality would not be able to help either.
The Flawed Statistical Arguments of the EC
- The EC's argument in supported of the “deliberate bloating” theory has three legs:
a. There are 797,493 individuals on the disqualification lists (multiples/duplicates and exempted) of the 2019 Voters Roll. These people would have had to be added to the Voters Roll had a limited registration exercise been conducted.
b. Had a limited registration exercise been conducted in 2020, instead of the full re-registration, two million first-time voters would have registered.
c.The total number of persons on the register would have totalled 19.6 million, a full three million more than what has been captured in the 2020 Voters Roll.
All three strands of the argument are completely ridiculous.
- First, the only way an individual can be flagged for disqualification is through the “challenge” and “certification” processes. A person is only truly disqualified if they fail during the adjudication process to defend a charge of ineligibility after having been flagged for multiple registrations, or if their residency or citizenship is successfully challenged. For any of this to happen, the individual must first be on the main provisional register.
Therefore, it is practically impossible to separate the roughly 798,000 individuals from the 16,845,420 on the 2019 Voters Roll and then go on to claim that these people would have been automatically added to the register had a limited registration exercise been carried out this year. Why then are these mysterious people not being added to the 2020 provisional register just compiled? It is extremely strange that a constitutional body will publicly make such dubious arguments.
- Second, there is no way 2 million people would have been captured on the Roll had a limited registration exercise been conducted. This country has good enough data going back nearly three decades to project the likely turnout for limited registration. In 2006, roughly 630,000 people registered. 928,540 registered in 2014. 1,046,067 registered in 2016. And 1,211,395 registered in 2019. The only time approximately 1.8 million wrote down their names in a limited registration exercise was in 2008 due to the exceptional circumstances of the 2004 mass registration exercise. In 2004, photo IDs taken with digital cameras were first introduced. Despite the country having insufficient cameras, the exercise was conducted over just 2 weeks (compared to the 40 days used in 2012.) The situation was such that the total number of voters dropped compared to what was on the 2000 register. The high number recorded in 2008 was thus a correction, and never before or after has similar figures ever been recorded. To therefore expect 2 million people to turn up in the midst of a pandemic just 1 year after a registration exercise is quite shocking and completely bereft of statistical sense.
- At any rate, how many first-time registrants were recorded on the 2020 register just compiled? This is after all the most conclusive evidence of how many would have turned up had only a limited registration exercise been conducted. Our evidence suggest that first time registrants were less than 1.2 million. In any serious country, that should settle the debate.
- Third, the EC, very characteristic of its obfuscating style of recent years, deliberately chose to provide numbers for 35 instead of the 36 days of registration and chose to issue its press statement when two more mopping up days remained. Adding the figures from the omitted three days will certainly take the numbers past 17 million. The reader must bear in mind that this still fall short of the 40 days of official registration plus two days of mopping up witnessed in 2012, during the last mass registration exercise in this country. Nor are we accounting for the many Ghanaians temporarily overseas who would normally have returned in the leadup to the registration, and who would have taken the advantage to register had COVID-19 not compelled a shutdown of the country's borders. In these circumstances, whilst general fears that the pandemic will turn millions away did not pan out, some voter suppression was nevertheless present. Worst, we can only hope that the surge of infections owing to the overcrowded registration centers we saw shall not lead to more hospitalisations and deaths in coming days and weeks.
- The case can certainly be made that limited-registration figures of approximately 1 million in 2020 would have taken the total number of voters on the roll to 18 million thereabouts, which is 1 million higher than what would be recorded in this exercise. Does that not suggest that the register had been maliciously bloated for some malicious objective, such as electoral cheating by one or the other party? We devote the next section to this contentious subject.
No Malicious Bloating
- It is clear that had a limited registration exercise been conducted, a potential 1.2 million people would have got onto the register, roughly the same number it would have exceeded this newly compiled register by. The difference is easy to explain, however. Due to severe underinvestment in the Birth & Deaths Registry, the names of dead persons are not being removed from the register. In 2018, the Registry's allocated budget was $1 million in today's money, of which a mere 67% was released. If even a fraction of the $150 million given to the EC for this needless task had been instead passed on to the Registry to enable automatic electronic notifications of birthdates and deaths, something that would have cost less than $1.5 million, the $150 million wasted on this exercise would have been saved.
- Since 2012, projecting from the average death rate and adjusting for only those above 18 years of age and the proportion of the population on the voters register, 1.156 million names of dead persons must have been present on the register. It is clear that this accounts for the entire difference between the 18 million names that would today be on the register had a limited registration been conducted and the figures recorded in this exercise. In short, $150 million of our very scarce resources have been wasted to remove the names of dead people, who cannot be impersonated because of the biometric system, instead of the $1.5 million which would have more than doubled the capacity of the Births and Deaths Registry to improve records and automate notifications to the EC.
- What about minors? Nothing implemented in this mass registration exercise can be said to be more effective in removing minors than in previous registration exercises. About 40% of registrants used guarantors and virtually all of the remaining registrants used Ghana Cards. But recall that over 80% of Ghana Card registrants used guarantors too. In short, 90% or more of registrants are ultimately on this new Roll because of guarantors. A system whereby people merely state that in their belief a person is above 18 years old and is a Ghanaian cannot be an improvement in anyway. What is worse, current court decisions have merely served to wound the only institution whose true mandate is to report on birthdates and deaths, the two datapoints crucial to removing minors and dead people from the Roll.
- In an earlier contribution to the debate over the EC's adamant commitment to a new register, we outlined the motivations of the different parties to the controversy as follows:
The ruling party's only basis for supporting the mass re-registration was because of the NPP's long-held belief that through that process they could eliminate large numbers of ineligibles, especially foreigners, maliciously smuggled onto the register to distort electoral outcomes. Senior members of the party expressed an expectation to see about 14 million total registered voters by the end of the exercise, whilst the EC first set a target of 13 million, which it then revised to 15 million. One leading strategist cited the case of Ketu South's 140,000 registered voters, for example. Our view is that these fears were cynically exploited by the Electoral Commission to pursue a “procurement-focused” objective which had little to do with “cleaning up the register”. All the data does indeed point to such fears about malicious bloating being unfounded.
- Ketu South's registered voter numbers have actually increased at a pace slower than the overall national average. In 2000, the constituency had about 98,000 registered voters on its roll. Today it has a little over 140,000. This is a “mere” 40% increase compared to the nearly 70% increase over the same period nationally. Or, to compare like to like, take Fomena, Ashaiman and Bekwai, which have all seen a 60% increase over the same period. In fact, the entire Volta region (now plus Oti) has seen a less than 40% increase in their registered voter numbers compared to the 50% number seen in Ashanti or the even more spectacular 60% recorded in Central Region over the two-decade period.
- There are some simple truths governing the rate and manner of change of national electoral registers, especially in countries where one has to explicitly register aforehand. These include the rate of urbanisation, the competitiveness and degree of maturity of party politics, the use to which the Voter ID card can be put and the level of investment into the general electoral infrastructure. When these factors are taken into account, the proportion of the total population who are registered voters in Ghana, ~53% to 55%, can be seen to have been roughly consistent for many years and, furthermore, completely typical. In Cape Verde, it touches 66%, whilst in Tunisia it is over 60%. On the other hand, it is just over 40% in Botswana because of low political competitiveness. In Ghana, Greater Accra trounces Ashanti in number of registered voters, despite the latter having a bigger population, because registration centers are more cluttered/clustered (investment level), urbanisation rate is higher (95% to less than 70%), and competition for parliamentary seats is keener.
- Our own history as a country also bears this analysis out. As urbanisation, electoral investment and the level of political competition have climbed, so also have registration levels. These are basic empirical facts easily borne out by regression analysis. We hereby present a few charts and graphs to illustrate the points.
- We will urge the ruling party in particular to take a careful look at the outcomes of the registration and change its posture to data analysis presented by independent Civil Society Organisations (CSOs). It should discover that regions where it performed well in the 2016 elections such as Ahafo, Western North and Eastern have all seen noticeable declines. Western North for instance seems to have experienced the steepest fall. Meanwhile, the most competitive regions such as Bono and Bono East, Greater Accra, and the regions in the northern and upper political zones of the country have seen perceptible increases. Strongholds of the two major parties, Volta and Ashanti appear to have maintained roughly the same numbers. All of these facts are consistent with an analysis that discounts the conspiracy theories of malicious bloating and affirm the typicality and ordinariness of the current situation.
So where does this leave the country? First and foremost, the country should reject the triumphalism of the EC, and deny it the approbation it is so desperately seeking. Throwing $150 million at a non-existent problem and ending up at the same place you started from is no achievement.
In a desperate bid to prove to critics that it could compress the registration into 36 days instead of the more than 40 days warranted, it recklessly refused to enforce the social distancing and continuous sanitisation of scanners it had promised (see this video: https://youtu.be/fChfLAi7r6g); mobilised the old BVRs it had cynically claimed were obsolete, then promptly and shamelessly denied using any HP equipment when we had clear video evidence that they did (see here: https://youtu.be/5XIz4_e5MQ0); and disgracefully rigged the guarantor system by condoning the use of pre-filled guarantee forms, whilst blatantly turning a blind eye to collusion among electoral officers and party agents (see here: https://imaniafrica.org/2020/08/07/electoral-commission-of-ghana-ec-registration-recordings/)
This is a totally discredited Electoral Commission that needs to embark on serious soul-searching.
The rest of the country should not feel secure about the electoral process until the following has been done:
- A detailed independent asset audit of the Electoral Commission to account for the 60 million dollars of equipment procured between 2016 and 2019 that it claims have suddenly gone obsolete.
- Deep and complete reconciliation with the Opposition Parties and CSOs whose calls for accountability it has shunned so far. Failure to patch up and open up to scrutiny by all stakeholders will only deepen the rancour, including the ethnocentric tensions, that marred the integrity of the electoral process in many places.
- A serious national dialogue about strategies to fix the identification and register cleaning issues that remain wholly unresolved due to continued neglect despite the expenditure of tens of millions of dollars of this country's hard earned resources over the last decade on the electoral system.
- Tackle the “Procurement Raj” that has taken hold of the EC and is hell bent on milking the country at all cost through scheme after scheme.
For the sake of our democracy, we hope this country shall take these matters seriously.