Hacker News new | comments | show | ask | jobs | submit login
The Economics of Hacking an Election (columbia.edu)
75 points by sohkamyung 71 days ago | hide | past | web | favorite | 26 comments



This is seriously uninformed article with bad assumptions and assertions.

The first assertion author makes is that attacker is resource limited which is a pointless assertion for stack sponsored entities unless cost of attack is astronomical.

The second point author makes is that US elections are highly decentralize for attacks such as hacking electronic systems, making effective attacks very expensive. Again this is completely wrong assumption for two reasons:

(1) attacker doesn't need to hack all of US precincts. Thanks to thin win margins in elections, you just need to flip only few precincts that have been predicted as "middle". Typically these are just 1-5% of total so cost of attack is actually very low and could be targeted with precision for a desired result.

(2) There are only few brands and systems of voting machines and other electronic systems are actually in use. Again, you don't have to hack all of them. Just hack one or two and they might be good enough to slip 1-5% margin you desire. The connection to Internet is irrelevant. There are well known methods to hide malware in firmware drivers that could have been planted by a state sponsored hacker who obtained job at one of these companies making these machines. These kind of malware would be impossible to detect without very extensive reverse engineering. Think about this for a minute: For a cost of less than a million dollar (i.e. compensation for single hacker) you can flip election of entire country. It makes a huge economic sense to any adversarial government.

Overall, hacking political system has shown to have ROI of 10,000% or more and it should make a perfect economic sense to businesses and external governments. You can literally donate just $100K to political candidate and he would pass a law for you that makes your company 10s of millions for many many years. It never ends to surprise me that political donations by commercial entities and ultra-rich is not considered a crime in US. In most other countries they are known as "bribe". Even better, just flipping few of the precincts you can flip entire election in to your favor and you can do it all even legally by precision targeted social media. It's just amazing power to behold on one of the largest and most powerful country.


This is incorrect:

You can literally donate just $100K to political candidate and he would pass a law for you that makes your company 10s of millions for many many years. It never ends to surprise me that political donations by commercial entities and ultra-rich is not considered a crime in US. In most other countries they are known as "bribe".

Companies cannot give money directly to politicians and individuals are limited. https://www.opensecrets.org/overview/limits.php

Political Action Committees can accept money, but must be independent from a candidate.


  Companies cannot give money directly to politicians
According to [1] Paul Ryan accepted $121,700 from Blackrock Inc, and [2] says Charles Schumer accepted $102,950 from Lockheed Martin.

Sounds like companies giving money to politicians to me?

[1] https://www.opensecrets.org/members-of-congress/summary/paul... [2] https://www.opensecrets.org/members-of-congress/summary/char...


They are aggregating the contributions of people associated with the company. Look at https://www.opensecrets.org/members-of-congress/contributors...

These tables list the top donors to candidates in the 2017 - 2018 election cycle. The organizations themselves did not donate, rather the money came from the organizations' PACs, their individual members or employees or owners, and those individuals' immediate families. Organization totals include subsidiaries and affiliates.


> must be (nominally) independent from a candidate.

It sounds a lot more effective than it is. In practice it’s a non issue.


Companies also can lobby which is the more hedged way of influencing policy since an established industry think tank can have more staying power and influence than a single or handful representatives or senators.

Edit: what I mean to say is that legal lobbying lowers the incentive to use campaign donations as a form of influence.


> Overall, hacking political system has shown to have ROI of 10,000% or more and it should make a perfect economic sense to businesses and external governments.

This trope amazes me. If this were true, you’d see a lot more campaign donations. Instead, you see a few billion out of an economy of trillions. There isn’t an investment opportunity like that just sitting around without investors jumping on board and bidding up the price. It’s not logically possible for this to be true, without making a bunch of other hand waving and dubious assumptions.


I agree that this is a trope. It a factoid defined/estimated/invented by athink tank or somesuch, published somewhere and then republished lots of times citing earlier publications.

It's inevitaly made up, because no one (sane) publicly acknowledges that contracts won as a result of lobbying.

But, I disagree with this:

"An investment opportunity like that just sitting around without investors jumping on board and bidding up the price"

Economists theorize using "perfect" markets like a stock or bond markets. Real life includes clunkier scenarios. An investment might be worth X to you and Y to me. It may be available to me and not to you, or worthless to you. I might be able to evaluate the value, while you can't.

That doesn't mean that the roi of "investments" that aren't traded investment products will usually be in that 3%-30% range. If Apple decide to develop a new product, that is an "investment." The cost of that investment is not subject to those kinds of market forces, which ensure that high ROI investments don't exist. The ROI could be negative or it could be orders of magnitude above average.

With something like "buying influence," there is no economic theory reason to expect any ROI in particular. It would be unexpected if the value of lobying was less than the cost, but otherwise it's a pretty open field.


I don't think hacking and donating should be conflated this way. What is the ROI for a dictator who decides to count their own votes?


Your counterargument misses the point made explicitly within the article:

"They may have vast resources, and in particular they may have more resources than the defenders—but they're still limited. Why? They'll throw enough resources at the problem to solve it, i.e., to hack the election, and use anything left over for the next problem, e.g., hacking the Brexit II referendum… There's always another target."

Too: there may be multiple parties attempting hacks. Getting your opponent to overextend (a war of attrition) is a time-honoured strategy.


Those are weak arguments. Author basically says hacker doesn't want to hack country A because then he will run out resources to hack country B. Seriously?

As mentioned in comment, hacking election is very cheap. On back of the envelop, I think $1 million should buy you enough ads to precisely target handful of swing precincts that will give you just 1-2% of the boost needed to put you over the thin win margin. If US elections were won by say, 25% win margins then it would become more expensive because it would be hard to scale. But most elections have less than 5% win margins these days and there are very good predictions available on swing districts. If you have access to voter registration database, you can literally target your ad at individual household level. As shown by last 4 major elections, just even flipping less than ten thousand household can achieve victories.


This makes it sound like the only people running targeted advertising would be the state-sponsored "hackers", but that isn't the case. Campaigns spend hundreds of millions of dollars on advertising, including digital, and one side still loses, and not always the side who spends less on advertising.

If you think it's so easy to swing elections with a few hundred thousand dollars of ads, you should definitely go into political strategy and prove it. They'll rain money down on you.


Equivocation of ad placement and hacking is a disingenious tactic, especially here.


I think as well this misses the point that an election target can be life or death. If the US elects a Putin-unfriendly government in the near future, one would expect some serious fire-and-brimstones reprisals from the Americans (and European allies) for previous election interference. I’d be amazed if those don’t focus on bringing about regime change in Russia, and it seems unlikely that can happen with the current leadership getting away unscathed.

Whether or not this is a good thing is left as an exercise for the reader.


Good points. Your last paragraph really shows how easy it would be / is which makes me wonder why is the Russia story such a big deal? There very well could have been similar attacks pulled off in the past with no one noticing (or it being legal all together)

Sidenote: politics only seems to have morality as a focus when it is in their favour.


> which makes me wonder why is the Russia story such a big deal?

Because it allows the responsible people to shift all the blame to an external actor:

"Sure, we might have used crappy and insecure voting machines, everybody with a clue warned us about for years, but that wouldn't have mattered if the Russians didn't hack them!"


>Your last paragraph really shows how easy it would be / is

His last paragraph is the equivalent of 'I'll create a GUI interface using visual basic; see if I can track the killer's IP' it displays a complete lack of understanding or knowledge of campaign finance, how the electoral system works, and how it's manipulated.


The real election hacking come from the fact that people are easy to sway. We succumb to subtle nudges in ways that we aren’t always consciously aware of. We now live in an era where a given nudge can be directed and concentrated to the right group of people and doing this at scale is practical.

The adage in politics has been, “All politics is local.” This appears to be changing in the U.S. Money and influence comes from all over and the issues that rile up people enough to vote are less and less tied to local problems. At least this is how I see things. I don’t know a good solution to the problem. Perhaps Asimov had it right in his story “Franchise”.

https://en.m.wikipedia.org/wiki/Franchise_(short_story)


Are voting machines really connected to the Internet?


All democratic elections are hacked, not in a technical way but in a social way. Take my country India, for example. The winning party in almost all elections till 2014 got just under 40% of the total vote. Minorities form 15-20% of the vote bank and the minority vote is always consolidated, i.e. they vote enmass for a single party (usually for The Indian National Congress (INC)). In other words, anywhere between 15-20% of the 40% mentioned earlier is the minority vote. The majority vote is always divided between the many many warring parties and the minority vote becomes the king/queen maker. No wonder the INC has always danced to the minorities tune.

I guess the majority had had enough and came together for the first time in 2014, resulting in a win for Narendra Modi.

I also guess the same situation happened in USA in 2016. The usually apathic/uninterested majority were frustrated/angered with years of ’minoritism’ and bonded together into a single consolidate vote bank that resulted in Trumps victory.


I won’t speak to the specific examples in either India or the US, but I will say that when you have extreme support for one candidate by members of minority ethnicities, that typically suggests that the other party is either apathetic or hostile to their interests. If there are enough members of the ethnic majority to reach a plurality, then that party will, and should, win. The fact that they vote en mass for one party isn’t “hacking”; it’s exactly how elections are supposed to work.


In an ideal world where people are intelligent enough to vote for their long term benefit, then what you said is true.

But let's say a country have little care to minority ethnicities. I have the many resources to arrange short benefit to them, such as occasional free dinner for those ethnicities, arranging cheap or free traveling trip for them. Essentially bribe those minorities.

Now I am elected to become a president. How do I get more resource to bribe minorities? I just receive donation from the rich and just set law to benefit them. Those law will get me donation and let the rich suck more money from the minorities. Now I have those dirty money in hand. Since minorities are living in a more challenging environment by my policy, those minorities will even more welcome my bribe and continue to vote for me.

Now you may ask, why are those minorities can't read news? Why they can't listen to the other side that may benefit to them? Firstly, minorities have very tough environment and they don't have education and even time to think about their future. Therefore, they just vote for whoever is familiar or whoever give free gifts. Secondly, a government could control the propaganda and paint those opposition as destabilizing country and blaming them as enemy and reason of their miserable situation.

It's not a hack, but a flaw. Democracy assume all people care enough about politics to understand the implication of each policy. If people don't have time resource for this, then it is pretty easy to spread propaganda to paint some enemy. People in poor condition just want to find someone to blame. And people trust the one who could bribe them a little benefit.

Of course, there are no better alternative than democracy, but people just too focus on their living and care to their own short term benefit. Long term benefit is difficult to promote since no one knows when will this be effective.


Voting is merely one way to elect (any kind of) representatives and indeed, the problems mentioned in the gp are inherent to it. Consider how drawing a lot is probably more fair and less prone to manipulation [1]

[1] https://en.wikipedia.org/wiki/Sortition


That argument is certainly true, but it applies to nearly everyone. Almost no one is politically well-informed enough to be sure of what’s in their long-term interest, with most people not understanding for example economics or environmental issues enough to take an informed position. So politicians frequently resort to short-term fixes, such as tax cuts, to keep people happy and voting for them. If more of those short-term fixes benefit minorities, that still isn’t worse than if they benefit majorities.


> If more of those short-term fixes benefit minorities, that still isn’t worse than if they benefit majorities.

If minorities vote for receiving short term benefits then everyone including the minorities are knocked out of the long term benefits.

If you can just win by offering a vehicle like horse cart and keep upgrading it through research and development, adding 1km/hr speed increment every 5 year. Then you'll probably never be able to develop a car because its development requires long term research and development program, even if nothing, atleast a working infrastructure.


> Firstly, minorities have very tough environment and they don't have education and even time to think about their future

I'll add here that even if they've time and education, they might lack proper nutrition if the government deliberately keeps the watchdogs which regulate food supplies, drugs and environment underesorced. This will reduce their IQ and ability to take a better decision.

The end game is whoever manages to keep people desperate and dumb.

Simply letting bad actors contaminate the food with lead will do this job of keeping everyone dumb and winning election next year by throwing breadcrumbs.




Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact

Search: