The Question
This question was posted on behalf of an ACE user working in Nepal

The Election Commission of Nepal has created a new voter list (biometric data included) and the current registration rate is about 70% of eligible, in-country voters.  On the question of what is an acceptable registration rate, it seems to me like you have to measure acceptance based on the commission's efforts to register all voters wishing to vote - i.e. how service-minded the commission is - rather than any specific rate.  Nonetheless, 70% seems low to me. Is there some kind of consensus of a minimum standard that would make the process "acceptable" in a country context similar to Nepal?

Summary of responses
In summary, there is no consensus on a minimum rate of voter registration and what is deemed acceptable largely depends on the particular context in question. When attempting to judge if the proportion of eligible voters included on the list is acceptable or not, a number of factors should be taken into consideration.

Several replies ask what constitutes 100 per cent of voter registration. How has this figure been calculated and is it credible?  This is of course fundamental when it then comes to calculating an acceptable proportion.

Of great significance is whether voter registration is voluntary or compulsory. If the latter, then a high registration rate would normally be expected: Although there is no standard, several replies state figures in the nineties. While there is no agreed upon threshold in a voluntary system either, one expert sees that “a figure below 60% would be considered low in most jurisdictions.” Another believes a rate of 75-85 per cent to be “an acceptable standard.”

The nature of the registration awareness programme, e.g. its intensity, effectiveness and inclusivity, as well as the amount of money spent, must also be considered.

The registration process should have the support of all political parties, civil society and the wider electorate. Whether a registration rate is acceptable or not ultimately depends on whether there is widespread buy-in from stakeholders.

Expectations should also be measured against who is responsible for the registration process.

A country’s previous voter registration figures may be a useful benchmark, although if new voter registration technology has been introduced, it is noted that allowances must be made for variations in data between voter lists.

Allowance must be made for those eligible voters in a voluntary system who have had the opportunity to register but, for whatever reason, have chosen not to.

It is important to note that registering to vote may not equate to an intention to vote. In some counties, e.g. Mexico and Antigua and Barbuda, there is an incentive for voters to register in order to receive a voter identification card, which can then be used in other contexts for identification purposes. In such cases, one would therefore expect an even higher registration rate.  

All of the above factors taken together should influence any judgement on the validity of a rate of voter registration.

Examples of related ACE Articles and Resources
Encyclopaedia:
• Guiding principles of voter registration
• Encouraging registration

Consolidated replies:
• Inaccuracy of voter rolls

 

Names of contributors
1. Ronan McDermott
2. Fredrik M Sjoberg
3. Carl Dundas
4. Horacio Boneo
5. Idi Boina
6. John Arran
7. David Arabambi
8. Chris Kyriakides
9. Bucumi Julius
10. Juma Thomas Otieno
11. Charles Benjamin Nsimbi-Kabugujjo
12. Pamela Mapondera
13. Atem Oben Henry Ekpeni
14. Kisimba Albert 

An excellent, and frequently-asked question!

First up, some clarity is needed.

1. What constitutes 100% Where does this figure come from? Is this figure credible?
2. Is voter registration mandatory or voluntary?
3. Is the voter registration process state (or EMB) initiated or citizen-initiated?
4. Are all political parties on board - i.e. is there any significant opposition to the voter registration process? Are all sectors of the population on board?
5. How does the 70% compare with previous census exercises? Previous voter registration exercises?

 

Depending on the answers to these questions, 70% might be way below, or way above "acceptable".  If ECN has taken all reasonable steps to register citizens and has done so in an even manner (geographically, politically etc), then 70% may be acceptable. 

In most cases, the acceptablility of the level of voter registration is a political one. Huge shifts upwards or downwards in participation may trigger questions about the process.

If I appear to be dodging the question let me finish as follows. To my knowledge, there is no consensus.  80% might be a success in one context, and a failure in another.

I look forward to other contributions.

This is a frequently asked question by EMBs' officials. This question seems to be put in the context of a voluntary registration system, for in compulsory registration systems one expects the percentage to be in the high 90s of the total estimated number of qualified persons. In the case of voluntary systems, the level of registration may depend on the nature and intensity of the registration awareness programme carried out. An old example may serve to illustrate the point: in 1994 in Malawi, a non-traditional methodology was used to inform persons about the registration of voters exercise and the upshot was the during scheduled six week of registration the percentage lagged in the mid-forty percent. The period was extended bysome two weeks and the traditional method of spreading information among the rural population used and the percentage of the voting population recorded soon increased to about 77%. A few years ago Trinidad & Tobago achieved low 90s percent in a voluntary system and the EMB was uneasy that they had not met international best practices. The truth is that in a volutary system of registration, after discounting those qualified persons who are not interested, for divers reasons ranging from fear of the taxman or jury service or for security reasons, international practice has not yet recognized a maximum benchmark, although a figure below 60% would be considered low in most jurisdictions.

Hi,

Good question. There's no simple answer, but examining the IDEA database we get the following regional distribution of registered voters as a proportion of the population:

 

 

Region	mean	sd	N	min	max
1. Eastern Europ	0.8575064	0.147421	81	0.270453	0.9996893
2. Latin America	0.7611107	0.1983082	215	0.1786904	0.999464
3. North Africa	0.6767776	0.2554067	72	0.0917528	0.9920424
4. Sub-Saharan A	0.7914392	0.1573003	148	0.2461689	0.99594
5. Western Europ	0.8753463	0.1639193	368	0.1763324	1
6. East Asia	0.9444623	0.0574689	28	0.7931521	0.9950247
7. South-East As	0.7907855	0.2224905	53	0.0877347	0.985083
8. South Asia	0.8823655	0.0971752	39	0.6482567	0.9962505
9. The Pacific	0.8288176	0.1437089	34	0.4902381	1
10. The Caribbea	0.8055457	0.1502355	62	0.2446623	0.9955559
Total	0.8199798	0.1849563	1100	0.0877347	1

 

 

--
Fredrik M. Sjoberg
Postdoctoral Visiting Scholar
New York University
E-mail: [email protected]
https:///fredrik-m-sjoberg
Mobile (USA): +1 347 843 3248
https://sites.google.com/a/nyu.edu/fredrik-m-sjoberg/

I would add one question to Ronan's quite adequate list, and then make a comment. The additional question is: Is registration an action that it is only useful for voting, or has it some additional benefits? In the footnote included in a paper discussing, among other things, benchmarks, I argued: "Let’s for instance consider the case of a benchmark related to the proportion of the population of voting age inscribed in a Register. The benchmark in a country where a document universally used as ID is provided when registering – as it is the case of Mexico – would be much higher than that of other country where registration is somewhat difficult, and has no other utility than allowing the citizens to vote, as is the case of Guatemala."

The comment is that there is a direct relation between the amount of money spent in the exercise and the number of people registered.  However, there will be decreasing returns, so after a point even the addition of further funds will not improve significantly the registration rates.  In any case, it is not usual to find such low rates of registration and there are many countries as difficult as Nepal.

 

70% Par rapport à la population en age de voter? ou à un recensement antérieur?

The main questions I would be asking are:

1) 80% of what? In some countries the actual population figure is an estimate and the number of those who are of voting age could be less accurate again. Could there be large numbers of people recorded in duplicate in the state registry? Is the figure derived by extrapolation from an ageing census? 

2) If the total eligible population figure is accurately known then have all reasonable avenues been explored to work with other state, social and political bodies to make the list more inclusive?

The threshold of acceptability will vary hugely due to the many factors pointed out above, such that any benchmark figure will have little use beyond comparison with previous registration exercises in the same country. Far more important is that the main political parties and agents are fully involved in the planning and to a degree also the execution of any registration exercise, so that opportunities for and likelihood of discrediting the resulting voter list are reduced.

The answer to this question may be approached in two ways. There are voluntary and compulsory registrations. Voluntaty registration is dependent on the level of preparation of EMB in respect of the quality and number personnel,registration materials and time available for the exercise. The percentage also depends on the credibility of the past elections and how free and fair the elections were. A credible election will reflect the choice of voters almost accurately,this in turn will influence the percentage turn-out of eligible registrants positively. Consequently, an acceptable percentage of registration will tend close to one-hundred percent but not one-hundred percent because of the unpredictability of natural and related forces. For compulsory registration,the acceptable percentage of registration should not be less than ninety-five percent if the EMB is fully prepared. This is because the decision of the eligible registrants to register is not dependent on the credibility, reliability and acceptability of the past elections since the registration is made Compulsory.

Within this question lies a very important subtlety… Did the eligible citizens that could have registered as voters fail to do so as a result of voter apathy or a poorly executed VR process?

•  Voter apathy is an expression of free will and has little bearing on how well the VR process was conducted by the EMB; in my opinion, it is primarily up to the Political Parties to rally their supporters into participating in Elections.
• By contrast, a poorly executed VR process results in omissions that in effect disenfranchise legitimate citizens from casting their vote and has a direct influence on the acceptance of the Electoral outcome.

 

We need to distinguish between the above two classifications in our deliberations here. As far as omissions (and inaccuracies) are concerned it is important to tightly limit this percentage. In essence, the more closely we anticipate a contested election to be, the more accurate the voters’ roll must be.

 

Many factors can contribute to low participation of eligible citizens to register as voters – to be fair not all factors leading to low voter registration rate can or should be blamed on the EMB. Indeed, the EMBs efforts to register all eligible citizens wishing to vote must be analysed in order to establish where the problems lie and (if possible) how to improve the registration rate.

 

Now, as far as registration percentages go there is no mutually accepted minimum standard – such a factor would realistically speaking vary from one country context to another. For example, consider a voluntary- vs. compulsory-registration system; countries with different electoral laws, voter registration procedures; even the political culture varies. In some countries political interest is high, while in others lukewarm at best -- in fact, within some countries the voter interest in elections varies significantly from one region to another.

 

No matter what, when it comes to any primary electoral activity such as VR it is imperative to ensure that the political parties are on board and any legitimate concerns they raise are promptly addressed by the EMB. Acceptance of the Voter Roll by all major political parties before proceeding to Elections is vital.

 

Simply put: if there is no buy-in to the VR process then one cannot reasonably anticipate a buy-in to the Electoral outcome!

 

It must be borne in mind that in many countries the quality of the voters roll is a “foundation” upon which credible Elections can be held. In the event where the Voters Roll is inaccurate, incomplete or geographically biased (such that it may favor one party over another) then this “foundation” is deemed flawed and bears the potential to severely undermine the credibility and acceptance of the Electoral outcome that follows.

 

In the event where low voter registration rates can be traced to a poorly executed (or incomplete) VR process then the consequences thereof disenfranchise legitimate citizens from participating in an Election. The consequences of a low rate of voter registration under this particular scenario are analogous to those mentioned in an earlier question “Inaccuracy of voter rolls” (see http://aceproject.org/electoral-advice/archive/questions/replies/826801063 ).

 

[In the event that the 70% figure arose out of some statistical comparison with the older VR System's data...] When new VR technology is introduced analysis of variations between old data and new data must be done with a pinch of salt.

• Often introduction of newer Voter Registration technology may give rise to questions regarding statistical variations when compared to the older system’s data. In fact, some statistical variations can easily be explained. For example,
• In a new biometric voter registration system where all voters are registered from scratch (if implemented correctly), duplicate voter registrations that the older system may have missed will be eliminated.
• Another often overlooked aspect of new VR Systems starting registration from scratch is that VR records of deceased citizens that may still exist n the old system's data will not exist on the brand new system. 
• On the other hand, even when there is no changes to the VR System used, statistical variations can be observed with subsequent VR exercises of each Electoral Cycle (say based on various factors such as population movement, deceased voters, population growth, etc.)

 

[ In closing: The question states a registration figure of 70% -- is it possible to get clarity on whether this figure was based on a recent voter registration sample survey or whether this arose out of a comparison between new-system and older-system VR data? ]

1. Thank you for the interesting question. To get 100% of eligible voters registered in an ideal target to reach by different means of communication and sensitisation, good will of all stakeholders to sensitise the population and participate to the registration process.

2. You have to know that no where in the world that target is reached at 100%. But, for the credibilty of the electoral process, you have to do your best to get as high as possible rate of registered voters.

3. For your case, congratulations on for the score you reached, using new tech.! Go on! It would have been dangerous for the process, if you have been under 50% of eligible voters. Even there, it depends upon  the legal framework of each country. For example in Burundi, it is stipulated by the law that elections are free to mean that you can not force people to go to register or vote. In this case, the rate of registered people comparing to the number of eligible ones is not very important! Any rate is good to continue legally the electoral process. But, i emphase on this: for the crediblity, acceptence of election results, interest of peace and developpement, as EMBs, you must get as high as possible the rate of registered voters.

4.If in Nepal, the registration or vote is an obligation , you still have a lot to do in terms of civic and voter eduction!

Thank you

Julius Bucumi

Commisssionner in Charge of Electoral Operations, Logistics and Legal Affairs

Burundi EMB

 

 

 

 

 

 

Well, the question at hand seems quite important because it seems to be targeting a consensus Acceptable Percentile Voter Registration Turnout (APVRT). Knowing that voter registration is part of a political process stemming from the citizens themselves and dictated by a number of factors, the APVRT will to me be as good enough if the factors necessary towards its success have been exhausted. For example if proper voter information, civic education, and voter education among others have been done and the Voter Registration (VR) staff have been all over, over adequate time in a process that is politically acceptable, then, the best an EMB can expect in terms of APVRT is the outcome of its VR exercise.

In my understanding, this kind of outcome is a process-led and determined and people-supported. In such a sense, thresholds cannot be put on People-Supported-Process-Led activities and events. The level of the outcome is satisfactory in itself. I would have a problem with 80%, 90%, 70%, e.t.c. if it is an outcome of inhibitions from particular quarters/ or and players. Otherwise, any outcome represents the peoples will in participating in their democratic processes.

In normal politics of participation, the natural sociological tendencies of the practice of polity determine much of future behaviours and a change from one tendency requires time and radicalism to make its positive occurrence also a natural sociological tendency. Politics of participation is people based and directed hence you may find from time to time changes like in voter turnouts, VR. At times high turnout in VR ultimately lead to Low voting percentages. These dillemas are still issues in electoral management whose observation from time to time require political players and concerned EMBs to monitor and discuss in finding way-outs.

I agree with Ronan above.  The answer is very dependent on whether registration is voluntary (i.e willing only ) or is mandatory.  It also depends on whether the voter has to make an effort to register ( i.e go to a registration point) or the data is extracted from a central database.

It is also true that 'acceptable  %ge' depends on perception of the stakeholders in the EMB and in the process of registration.  Accuracy of the register will also affect the acceptable %ge ( e.g if the registered population is 105% of the total eligible population!)

Please find attached an extract from my Masters thesis for further information.

 

Charles

 I am also of the opinion that it depends on whether registration is voluntary or mandatory, as well as the level of voter education done by the EMB in advance of registration.  I have found that there is some level of voter apathy where registration is concerned, especially with younger voters - say 18 to 30 years.  In my country several business places (e.g. a commercial bank) require two photo IDs and if , for example, one does not possess a driver's licence and/or a passport, they will be inclined to register in an effort to obtain that second means of identification.  In addition, if it is a new voter registration system where biometrics is involved, some voters may be skeptical of the taking of fingerprints for fear that they could be shared with law enforcement agencies. This could be a deterrent, although not of a great magnitude, I would think, but it could have an effect on the turnout rate, especially if it is re-registration.  Another factor which would affect the rate could be the duration of the registration period  and the number of centres earmarked for the process.  If all the negatives are minimised, a rate of 75% - 80% for voluntary registration would be an acceptable standard in my opinion.

It would be difficult to give an appropriate answer without backgound information to the whole process. However, I would say that the best situation is when every eligible voter has been given the chance and access to the registration process.The other determining factor would be how widely the information on the registration process was spread or given to the electorate and whether there was wide consultation and awareness of this type of registration. Accessibility of the information also depends on the literacy rate of the population and whether this was taken into consideration at the time of voter and civic education. Therefore without background information it is difficult to judge.

Pamela Mapondera

Referring to the question, consensus and or acceptability of voter registration rates are contextual and greatly depend on various variables already mentioned. However, if one has define acceptability in this regard, one will have to assess the level of inclusive participation by key stakeholders and professionalism to reach consensus. The reverse may be true for some systems depending on factors that may trigger or attract eligible or potential voters to register. In Africa for-instance, voluntary registration attracts eligible people to register especially if the voter cards are permitted for use as a national ID cards that could eventually be used for various purposes. 

I want to think that the threshold for voter registration rates  should be determined following the system used and the method applicable to each system. Consensus should be reached by the stakeholders before the start of the exercise and based on principles that are acceptable and that gives room for all eligible to register with no constrains. Rate for voluntary registration ideally should be 70% and above but security issues may also play a significant part in preventing EMBs or Registration Commissions from attaining objectives.

Rates for Mandatory Voter Registration may depend on the method of application, using the civil registry or not. One will need to question if punishing eligible voters who refuse to register will increase turnout in a mandatory system. In any case, mandatory systems are ideal when applied using the civil registry and should be able to attain 90% and above.

 

70% c’est déjà un grand record dans la phase enrôlement des électeurs. Mais si le Népal est un pays avec infrastructure routière conséquent il est donc utile de faire l’optimisation des centres d’identification afin de joindre plus des gens possible. L’autre option serait de d’identifier les populations par groupement ou préfecture, en se rassurant de la sensibilisation de proximité par les chefs des entités concernées.

Ideally, the registration rate should be 100%. This has not been achieved even in countries like Australian where registration and voting are compulsory. High public confidence and expected benefits from an elected government as well as the efforts made by the EMB may or may not result from higher registration figures. It is always a combination of factors. In a country like Nepal where voter registration is voluntary, I would have taken 70% voter registration as reasonable.

Moreover other factors that could have affected voter turnout in Nepal included registration  fatigue, a lengthy registration process, and limited access to some of the registration centers. Daniel Steven Roberts in his dissertation "Why We Don't Vote: Low Voter Turnout in US", gives an interesting account despite the efficient availability of resources expected to result in much higher voter registration figures.