You have been staffed on a consulting team that has been selected to advise a southern state in the United States on the issue of electronic voting. The state government has been toying with the idea of using electronic voting to increase voter turnout. However, it also wants to do this only at a reasonable cost.
How would you frame this problem? What are the various issues you would like to analyze?
Candidate: Could you explain electronic voting to me?
Don’t hesitate to ask for clarifications or definitions once you’ve been presented the case, or at any time during the discussion. It’s essential that you share the interviewer’s understanding of the key terms.
Interviewer: E-voting, as defined by your client, is any system that captures votes electronically, rather than by a manual ballot.
The interviewer defines e-voting quite broadly; there’s no harm in asking a second clarifying question, this time using examples, so there’s absolutely no doubt that you’re both on the same page.
Candidate: So, e-voting might refer to a PC in a voting booth, or to a web site that can be accessed by a voter over the Internet?
After you’ve clarified the situation, it’s advisable to set the agenda for the discussion, as early on in the discussion as possible. This is what you should do next:
Candidate: I’d like to structure this analysis by first understanding the impediments to increasing voter turnout. I’d then like to analyze how e-voting might aid in this effort and the pros and cons of using e-voting, specifically the costs and benefits. This is how I would frame the problem.
Of course, simply having a computer in a regular voting booth most likely is not going to make much of an impact on voter turnout (though it might improve voting accuracy). If you’re going to assume that for the purposes of this study e-voting applies only to the ability to vote via computer at many locations, then say so.
Candidate: I am going to assume that we will study voting by computers outside the voting booth.
Interviewer: That’s fine.
Now go ahead and make specific inquiries. In this case, the first issue is: who’s not voting, and will e-voting rectify the situation?
Candidate: Has our client observed any patterns in voter turnout? Are specific populations less likely than others to vote? What, if any, are the demographic variables which have a bearing on voter turnout?
Interviewer: These are all good questions. There has been a preliminary analysis of citizens who do not participate in elections. But since all voting has been manual, there is a scarcity of data. Our client has made certain educated guesses on the groups of people who they would like to target, but is hoping for us to push the thinking forward in this area. How would you do this?
The interviewer doesn’t really give you any information to work with here, and instead asks you to speculate on the reasons why voters might not vote. This is easier than it seems. You can do this in several ways. You can list all the reasons why they might not vote; or you can classify non-voters into a couple of different groups. Let’s use the second approach. (If you used the first approach, you would need to list reasons such as lack of time, disinterest, lack of mobility, etc.)
Candidate: I would separate the chronic non-voters from the occasional non-voters. I would like to focus on the former group. The latter group might remain absent for unavoidable reasons such as illness or travel.
Interviewer: Sounds reasonable.
Another way in which you can classify non-voters is by demographic variables such as age, income or sex. Remember that you’re doing this so you can match the demographics of desired users with the demographics of non-voters. If these two groups match, then e-voting over the Internet might be a good way to increase turnout.
Candidate: I don’t have adequate information to identify specific demographic variables, but this is what I would like to identify: What are the demographics of chronic non-voters?
Interviewer: The client has identified two such groups of “chronic non-voters”: senior citizens, age 65 and over; and young professionals between the ages of 25 and 35.
Candidate: That is very helpful. I would now like to understand if e-voting will help to increase participation rates in these two groups. In order to do this, it would be helpful to understand why these two groups have lower participation rates to begin with. My hypothesis is that mobility may be a problem with the senior citizens, while busy schedules might avert the young professionals from voting.
If this were the case, e-voting would address both these problems. Participating from a remote location might be possible with Internet voting. This should increase senior citizen participation. In addition, remote e-voting is less time consuming than voting in physical booths. This should increase professional participation rates.
Interviewer: But e-voting is not without its disadvantages. What are some of these?
Candidate: There are at least three major disadvantages. First, security remains problematic. While electronic signatures are now legal, confirming identity remains a problem on the Internet. Second, there is the issue of connectivity and access. While the young professionals probably have a higher Internet penetration, the senior citizens may not all have access. Third, there remains the issue of benefit versus cost.
Interviewer: What can our client do to address these problems?
Internet security is a huge issue, and you know that governments and various private organizations are looking at it. You’re not sure what a state or city government can do to add to this discussion.
Candidate: There is little our client can do unilaterally to tackle the issue of security. This is a much larger problem that needs to be resolved. I would recommend that our client avoid taking a seat at this table, but follow the progress closely. I would expect that at some point reasonable and cost-effective solutions to this problem would be invented.
Interviewer: And access?
Access is a big problem. Not everybody has a computer. You realize that if non-voters have access to the Internet, then it’s not a problem. If they don’t, then it’s a problem, so try to make some suggestions of getting around it.
Candidate: Access is a problem. Our client has to determine if a significant proportion of non-voters has access to the Internet or not. If so, then it would make targeting them to participate in elections much easier. If not, then there might be ways in which increasing access might be possible – for instance, using school and library computers to allow individuals to vote.
This actually has to deal with another issue. I see e-voting as a solution to increasing turnout. If specific demographic groups have been identified, marketing and advertising campaigns can target these groups, urging them to participate.
My guess would be that voters in the 25 to 35 demographic would be more likely to vote as a result of this initiative, as the constraint on their voting is most likely time and not mobility.
Interviewer: And benefit versus cost?
Here’s where you have to consider if the additional cost of providing e-voting is justified by the increase in the number of voters the system attracts.
Candidate: The cost of an e-voting system should be relatively simple to estimate. The major components would be the cost of technology, and administration. A revealing cost metric might be the expense incurred per additional voter. The state also has to consider if it has the budget to fund this project, or if it can free the funds required and create a budget for it. There might also be some sort of federal initiative to increase voter turnout, which could help fund this project.
The benefits are harder to measure. I would expect that as participation rates increase, marginal benefits of inclusion fall. In other words, a state beginning with a lower participation rate would have a higher marginal benefit from this system than a state with a higher rate.