If You Can Vote for American Idol on Smartphones, Why Not Presidents?

Over the weekend, my 94 year-old father-in-law asked what I would do to assure that every American who could vote would do so. That was an unexpected question, but one I addressed gingerly. This post is my answer restated for a public venue.

Simple answer: Smartphone. According to PewResearchCenter, nearly 70 percent of Americans own one of the devices, but the number among voting age adults tops 80 percent, according to other estimates. Surely a program could be in place by the 2020 Presidential race, and if lawmakers were truly serious about universal suffrage, a Manhattan-like project could make it happen by the next Mid-terms. 

The program would demand commanding commitment from state and federal governments, starting with the development of voting apps for each and every state and building secure server infrastructure. Then there is the two-pronged problem of making sure everyone who is eligible to vote has a smartphone and every device owner is registered to vote. Authentication, to prevent fraud, is another concern.

The the progam would have three fundamental goals:

  • Achieve as close as possible universal suffrage
  • Provide real-time voting results, even for absentee voters
  • Increase voting accuracy, while eliminating need for absentee ballots

App Development
If built around an adopted standard accepted by all states, the smartphone app could resolve many of the problems with voting today. The process for voters would be similar, regardless of local variances, such as referendums in some locales. Registered voters could cast their ballots, captured in real time, right up until the polls close.

Of course, the result of that could send the news media and pollsters into a frenzy—candidates, too—if large pools of undecided voters watch and wait to cast their ballots based on early returns. But, hey, that’s democracy.

For critics who argue that smartphones would be too complex for older voters, come to my ophthalmologist’s office where many of the patients are elderly and tap on smartphones just like youngins.

The biggest problem I see is contractual, meaning who gets to make the app(s) and the RFP (request for proposal) process dictated by local governments. Fairness and cost should be lower considerations, and perhaps even ignored. Few mobile apps ever developed would need to be as reliable and absolutely secure; these priorities matter more. Quality should be priority over bidded price.

There would need to be safeguards that ensure votes are captured during times of network congestion, like before polls close, and clear mechanisms (such as text message receipt) that assure the citizen that his or her ballot was registered.

The apps could be built to decrease the likelihood a ballot would be invalidated. Mistakes like marking two candidates go away, for example.

Something else: The apps could enable voters to make more informed decisions. Each candidate’s name would be hot-linked to open the smartphone’s default search engine, so that citizens could do additional last-minute research on contenders. However, the app should not provide any additional information. Going that route would add unnecessary complexity, increasing the chance the ballot wouldn’t be completed, and lead to all kinds of messy ethical/influence problems. Smartphone in hand is a last-minute opportunity for citizens to do additional research—those who choose to do so.

Infrastructure
The more challenging part would be on the back-end. The Affordable Care Act shows that state and federal governments can build big back-end server systems suitable for the purpose.  Likewise, problems encountered when launching sites like healthcare.gov spotlight how and where failures can occur.

I expect jurisdictional and states rights issues to complicate the creation of a single-connect server system or local ones that are interconnected.

Security is a worry. Consider the problem of a complex DDOS attack swamping servers during polling, interrupting voting and leading to some ballots being invalidated.

Smartphone Program
Not every voter owns a smartphone, and not every smartphone owner is registered to vote. Solution to both is intertwined. Using a federal subsidy program, perhaps, people who don’t own smartphones could get one for free. Flagship phones from the top-five vendors, based on sales, would be eligible for the program.

RFP would be based on lowest bids. Meaning: To be included in the program, manufacturers would not collect from the government, say, $649 for 16GB iPhone 6s. The cost of getting millions of new customers, and at the same time serving the public good, justifies less profit on each device. By limiting the program to flagship phones, government could deter, or even prevent, manufacturers from dumping older or less-desirable handsets.

To participate, citizens would either need to be registered to vote or do so when entering the program. Voters who, by cellular carrier check, do not own a smartphone or possess a device that falls below a minimum standard (such as manufactured more than years earlier) would be eligible for the hand-out.

Registered voters who do not participate in the smartphone program, and are verified to own a device, would receive an income tax credit equivalent to the value set by the aforementioned RFP. Government could encourage any laggards to register by making voter registration requirement to receive the tax credit. Additionally, government could promise, and deliver, a larger tax credit for previous smartphone owners who do vote and a smaller, but in total value equivalent, credit to program participants casting ballots.

Authentication
Voter fraud has to be a concern, and there will always be some regardless of the polling system used. Authorized apps would be distributed through official app stores, mainly (and perhaps only) Apple App Store and Google Play. The dilemma: How to authenticate voters without adding complexity that prevents them from voting.

One idea: Distribute four-digit codes by mail that the citizen combines with the last four digits of his or her social security number when voting. Simpler still: Social security number is authentication mechanism, which also should enable processing of real-time votes, assuming voter registration already is tied to the SS#..

Since the goal of the program is to achieve, as close as possible, universal suffrage, security must be balanced with simplicity.

That’s my basic proposal, understanding that bureaucracy, local and federal laws, logistics, special interests, and other factors likely would be wrenches in the works.

Photo Credit: Ed Yourdon

Editor’s Note: A version of this story appears on BetaNews.