We spend most of our time discussing the most helpful careers that you should take.

We just created a three minute career recommender to highlight some of the options with the largest positive social impact for you.

As most of the people we talk to are deciding between reasonable to excellent options, this seems like the right focus.

But which careers are the worst?

Here we try to guess which mainstream jobs are most likely to do significant harm. As almost no one we know is considering careers of this kind we have limited our investment in this research; it’s an initial exploration of the topic, based on general knowledge and a review of the key figures.

Here are the criteria:

  • The job has to be legal. Needless to say, organised crime is a harmful career!
  • More than one in a million people has to work in the job in the OECD, so it can’t be incredibly obscure or specific.
  • It can’t be harmful only if you’re particularly incompetent (for example, being a bad teacher), deliberately trying to do a bad job, or violating the profession’s code of ethics.

It’s easy to think of jobs that are useless and just transfer money from one person to another. But being unproductive alone isn’t enough to make a top ten list. There are also notable industries that cause harm, but are so large that each person involved is only responsible for a small share of the damage. We list a bunch of options like this, that were considered but rejected, at the bottom of the article.

We narrowed the range to jobs that: i) harm other people or animals a lot relative to the the number of staff involved, or how much they are paid; ii) or transfer vast resources from the poor to the rich. 1

Our guesses for the top ten most harmful careers

1. Marketing and R&D for compulsive behaviours such as smoking, alcoholism, gambling, and payday loans (and maybe computer games)

Drinking or gambling in moderation are fine. And sometimes you really do need a loan for 24 hours to deal with a cash-flow problem.

Unfortunately, people have cognitive biases, and a high level of impatience, that can make it easy to convince them to use certain products to excess. People will already use these ‘temptation’ products plenty without anyone designing clever ways to hook them into using more.

A large share of profits for tobacco companies, beverage manufacturers, casinos and short-term loans come from a small number of users who can’t control their use (alcoholics and smoking or gambling addicts), or are trapped in a cycle of use (people forced to take new high-interest loans to pay back old loans). Frequently that use ends up ruining their lives.

Cigarette advertising has shrunk in the developed world, but is a growing industry in the developing world, where an increasing number of people now have incomes that allow them to smoke, and are not as informed about the health risks of smoking.

We considered adding addictive computer game development to the list, but relegated that to the ‘unclear’ category below.

2. Factory farming

Animal agriculture, especially management or innovation in factory farming or slaughterhouse work, can be objected to on a number of grounds:

About half a million people work in the meat and poultry industry in the US, which collectively slaughtered 8.5 billion chickens in 2012. If half of them work in poultry farming, there were 34,000 chickens slaughtered for each person that year. The industry has roughly constant returns to scale, so an extra employee in poultry farming would be responsible for reasonable fraction of that number. Sounds horrific.

3. Homeopathy and other fraudulent medical technologies

There are a lot of useless products that don’t work. Sometimes the people selling them know they are useless. Mostly they just end up wasting people’s time and money, which wouldn’t be enough to earn them a place on this list.

What sets fraudulent medical technologies like homeopathy apart is that they often discourage misinformed people from using established medicines that would actually cure their conditions. Homeopathic medicines masquerade as real medicines in pharmacies, for those who aren’t aware that homeopathic medicines have no impact at all.

The same would apply to anti-vaccination activists, though that’s not really a career path per se.

4. Patent trolls

Patent trolls exploit the poor design of patent law to hold up useful innovations that we could all benefit from, in order to get a small gain for themselves.

The patent lawyers who work in patent trolling organisations, and the people who came up with this socially destructive business model are the key players here. The number of people involved is tiny, but the disruption they create for the legal system and actually useful businesses is huge.

It has been estimated that patent trolls extracted $29 billion from more productive companies in 2011:

“The direct costs of NPE patent assertions are substantial, totalling about $29 billion accrued in 2011. This figure does not include indirect costs to the defendant’s business such as diversion of resources, delays in new products, and loss of market share. Even so, the direct costs are large relative to total business spending on R&D, which totalled $247 billion in 2009, implying that NPE patent assertions effectively impose a significant tax on investment in innovation.”

These are small operations, so I expect fewer than 10,000 people to be employed in them (probably significantly fewer lawyers). This suggests they extracted at least $3 million from productive companies for each person employed.

That extraction should significantly discourage companies from bothering to design new products in the first place. One 2014 study found that firms that lose a lawsuit to a patent troll on average spend $200 million less on R&D each year as a result. It also found that patent trolls tend to sue firms with fewer lawyers on staff, prompting them to reorient themselves towards legal fees rather than R&D.

Learn more.

5. Lobbying for rent-seeking businesses or industries

Many businesses try to get the government to protect them from competition, both local and foreign. Classic tools for doing this are tariff, quotas, outright monopoly protections, regulations that small business can’t afford to deal with, and licensing restrictions.

This is bad for two reasons: i) it stifles innovation in new products or services that ordinary people would benefit from, ii) it raises prices for everyone, in order to enrich the (likely wealthy) owners of existing enterprises.

In some egregious cases these protections even prevent people in poverty from making a living. For example, a small number of relatively wealthy people are enriched by US sugar tariffs, which happen to lower incomes for much poorer Brazilian sugar farmers. The Word Bank estimated in 2004 that full liberalisation of sugar markets would save $4.7 billion a year, of which $1.6 billion would go to Brazilian sugar farmers.

A similar thing is true for cotton:

“… the United States spent about $24 billion on cotton subsidies over the past 10 years. … Oxfam International notes that the more than 10 million people in Central and Western Africa countries who rely on the production and sale of cotton lose up to $250 million every year due to subsidies … ”

The Common Agricultural Policy in the EU has a similar effect, of redistributing money from the very poor to the relatively rich, while raising food prices for everyone. One attempt to model the gains from removing distortionary agricultural policies in developed countries estimated this would raise the GDP of different developing countries by 0.3-1%.

A special mention goes to private prisons, which lobby the government to imprison more people for longer, even ones that pose little threat to the public.

6. Weapons research

Weapons researchers develop new ways of waging war. While in some cases new weapons will be more targeted and less harmful than the old ones, in general we expect this work to be very bad for the world. Over the medium term new weapons technologies becomes widely disseminated and so both sides end up being able to use them. Usually this makes any remaining wars more destructive.

Furthermore, weapons researchers are among the most likely to accidentally design new technologies that could be used to accidentally destroy humanity, such as nuclear or biological weapons in the past. We just don’t know what new opportunities for mass destruction are available, and we are probably better off not knowing.

7. Borderline fraudulent lending or otherwise making a financial firm highly risky

Unfortunately, our financial system is designed such that a relatively small number of bad loans can put a bank at risk of insolvency, and through its interconnectedness to other firms, ruin the entire financial system in the absence of government intervention. This dynamic was one of the root causes of the 2007-08 financial crisis, and the risk still exists today.

In some cases the lending involved in that financial crisis was actually fraudulent, and probably illegal. In addition, some lenders did not do sufficient due diligence, and ended up making large loans to people who clearly weren’t going to be able to pay them back. The incentive for intermediaries was just to make a lot of loan without being too careful about who they were going to.

In other cases bank employees would knowingly build up large ‘tail’ risks for their banks which could cause them to collapse, while getting short-term profits and taking home bonuses. For example, insurers would sell derivatives called ‘credit default swaps’, without setting aside enough capital to pay out these insurance policies if a financial crisis occurred.

How bad is this effect?

One estimate from the Federal Reserve is:

The 2007-09 financial crisis was associated with a huge loss … We conservatively estimate that 40 to 90 percent of one year’s output ($6 trillion to $14 trillion, the equivalent of $50,000 to $120,000 for every U.S. household) was foregone due to the 2007-09 recession. We also provide several alternative measures of lost consumption, national trauma, and other negative consequences of the worst recession since the 1930s. This more comprehensive evaluation of factors suggests that what the U.S. gave up as a result of the crisis is likely greater than the value of one year’s output.

The US financial sector employs some 6,000,000 people. I am going to guess that the share of people that are involved in activities that could predictably increase systemic financial risk is around 10%. (After the fact we will know that only a small share of these did most of the damage, but we can’t isolate exactly who they are ahead of time.)

It’s typical to have a financial crisis about once every 20 years. The average size of the damage is around half that of the 2007-09 financial crisis, so we’ll say in the USA the typical financial crisis causes $5 trillion in lost output alone.

That would suggest through financial instability each of these 600,000 people lowered other people’s income in the US by $42,000 on average, for each year they worked. Most of the welfare loss here will be born by less educated or experienced employees who lose their jobs in a recession and never get them back. There are other harms from financial crises as well, in particular

  • Lost income in other countries, in particular poorer ones;
  • Political instability (more visible in Europe than the US currently).

The paper from the Federal Reserve was reluctant to aggregate all of these costs into a single dollar figure, because the harms are not easily comparable – but it’s reasonable to think they increase the total damage several-fold.

8. Fundraising for a charity that achieves nothing, or does harm

The share of national income that a country gives to charity changes little over time, though it varies widely between countries.

As a result, if you’re fundraising for a charity that is much worse than average, you’re at least in part taking money out of the hands of other organisations that would produce much more social impact with it.

If on average charities do a lot of good, then that kind of job is indirectly very harmful.

This is one of the few jobs you could end up taking without realising the damage you’re doing because it’s a hidden effect.

Our favourite example of a project you could donate to that does harm? Scared Straight, which significantly increases criminality among young people.

9. Forest clearing

Land use change is a huge contributor to climate change, especially relative to the economic value it provides for the people involved. Forests also deliver a range of other ecological services, such as nutrient cycles, that humanity can’t easily get otherwise. Giving What We Can recommends preventing destruction of rainforest as one of the most cost-effective ways available to reduce climate change.

Of course humanity does need timber, but this is better gotten through managed plantations that minimise the amount of established forest that needs to be destroyed.

10. Tax minimisation for the super rich

Tax minimisation uses up some of the world’s smartest lawyers in order to avoid money being transferred from tax-payers to the government to spend. Assuming you’re working in a country with a well-run representative government, that’s unproductive, and makes it harder for the government to operate a simple and cheap tax system.

If you’re doing tax minimisation for someone with the median income, maybe we can chalk this up as merely neutral.

But the more you’re working for the very rich, the more likely that that money would be better used by governments, who would spend a reasonable fraction on transfers to people with lower incomes, education, healthcare and so on. Of course most tax minimisation is purchased by the wealthy, who have much more tax to minimise.

This kind of role is especially unhelpful if you’re stoking the tax minimisation arms race – for example, finding new, borderline-legal ways to channel money into overseas tax havens. By some estimates more then $20 trillion is hidden from tax agencies by the super rich in this way.

The New Economics Foundation rates this as one of the worst jobs, and while I don’t agree with how they did their analysis, I agree it’s a harmful option.

Edit: Some people are suggesting that lower effective tax rates on the very rich will raise revenues or economic growth. See this footnote for the evidence that convinced me otherwise.2

Do you work in one of these industries?

If you know someone who works in one of these industries, please forward this post to them!

If you yourself work in a job like one of these, note that these roles take a lot of skill, which means you’re in a great position to transition into a much more meaningful and valuable career.

Take our quiz to find out which jobs offer you personally the greatest potential for social impact.

Options considered but not included

People have strong views about this issue, and I’m sure there are plenty of candidates we might have missed.

Let us know in the comments what you think should have made the grade, or isn’t as harmful as we thought.

A lot of contributors suggested options for this list; below I describe some of the other contenders, and briefly explain why they didn’t make the cut.

Harmful on average, but not harmful ‘on the margin’

Academic publishers

Academic publishers use their established prestige to lock up and charge for research that is paid for by the public, while getting academics to conduct the peer review process for them for free. Here’s one polemic against them.

We are very glad to see a shift towards open access journals. They almost beat out tax minimisation for the tenth spot.

However, while academic publishers are horrible, it’s hard to see that they would be much worse with further employees. They are exploiting a reputation monopoly over prestigious publishing, but that advantage won’t become much bigger if they have more staff. What is needed to solve the problem are policies that force public research to be open access.

Working for other existing monopolies

A lot of incumbent, protected monopolies do a bad job and take advantage of their customers. There’s a reason Comcast is the least popular company in the USA.

But if it’s a strong established monopoly, having an extra staff member usually won’t make a difference, and might even make customers’ experiences better.

That doesn’t apply if the staff member is working on something specific, like government lobbying or anticompetitive practices, to entrench that monopoly.

Harmful but not concentrated enough to make the list

Working in the coal industry

Climate change and particulate pollution are very bad things. But the responsibility for it is spread over tens of millions of people working in the sector, and billions of people using the products. Furthermore, greater availability of energy through fossil fuels is also helping to lift billions out of poverty. This makes it unlikely to be one of the most harmful careers per employee.

Manufacturing conventional weapons

Again, presumably it’s worse to have more weapons in the world. But the industry is so large, and the existing stockpiles so extensive, that each additional employee working in weapons construction contributes very little to the expected use of conventional weapons. Hence we rate researchers developing new and more dangerous weapons as significantly more harmful.

An exception would be distributing small arms in unstable states.

Pundit who just makes up opinions, or journalists who write misleading clickbait

People who go on TV and make up controversial opinions to get attention are adding to the noise in public debate and making it harder for people to understand the world and make good choices. Likewise, journalists who post sensational but baseless claims. But I expect this is more driven by public demand than suppliers pushing their work – if you don’t write clickbait, someone else probably will.

People who sell unhealthy food

People probably consume more unhealthy foods than they should considering their long-term welfare. Presumably someone who spends their life selling sweets ends up causing extra cases of diabetes or obesity and perhaps even doing harm overall.

But this is a huge industry that is already close to saturating its market, so I’m not convinced the damage done by one extra person working in it is that large.

A more leveraged harmful position is as a food scientist who makes unhealthy foods more addictive, which is included in number 1.

Enormous potential for harm, but only if you’re messing up

Military decision-makers

Start a war on bad grounds, and you can do more damage than almost anyone else.

But who do you want making military decisions? Presumably very skilled people who have the broader interests of humanity at heart. So the military may be a very important place to work – if you think you can do a good job.

Misleading pharmaceutical or academic research

Discovering new drugs and technologies that help people is great. Fooling people into thinking a drug that doesn’t work does work is the exact opposite.

Unfortunately, the incentives for commercial researchers, and even academics, is to publish more flattering or interesting results than really exist.

This can end up wasting a lot of resources. However, someone who would be upstanding and simply refuse to produce misleading results could do a lot of good in a role like this, so I wouldn’t want to put people off the whole field.

Merely neutral

Fortune telling scams and other flim flam

Fortune telling doesn’t work, but the people who use it keep coming back, probably because it makes them feel better. It’s a misleading product, but probably only neutral or maybe even a bit positive.

Some forms of law

Many components of law have a competitive element, where both sides hiring more lawyers just maintains the status quo ante. This can be unproductive, but in the absence of further details probably isn’t destructive.

Financial arbitrage

Financial arbitrage can be valuable in non-obvious ways, but the first order effect is to win money from others and take it yourself. But as we explain here, this is usually a redistribution to less wealthy people, so puts it more in the neutral than harmful category. A good analogy would be winning a poker game against someone very rich – not exactly productive, but certainly not harmful either.


Making more addictive computer games

Playing games can be great fun and we would love to see more computer games in, for example, retirement homes. The concern comes from gaming companies that design products that seriously hook some users and mess up their lives in a way that looks a lot like gambling addition. Some people on the team here who will go unnamed will admit to feeling unable to pull themselves away from games, even when they wish they could.

50% of Zynga’s revenue comes from 0.15% of users who use their products an enormous amount. Some of these people pay vast sums for access to ‘items’ in the game. It’s plausible that their lives are being made worse.

Another angle that’s hard to judge is the loss of time that might be spent in a more productive or enjoyable way. Zynga has just 2,000 staff, but 20 million daily users. If each player plays for an average of 15 minutes, that’s 25,000 hours of game playing each day, for each staff member. That’s an enormous amount of leverage. We hope these games are making people happier rather than displacing activities they would enjoy more, or activities that would benefit others, but have our doubts.

Working in the army

There’s little worse in the world than war. But it’s hard to pass judgement on the effect of ordinary soldiers without knowing the goals or wisdom of the people who decide when they go to war. There are also opportunities for good people in the military to reduce the risk of unnecessary conflicts.

Public company mergers and acquisitions

One paper suggests that when one large public company acquires another, they tend to do significantly worse together than they did apart. One reason this may happen is that it’s in the interest of the managers of the acquiring company for the companies to merge, but not shareholders. Unfortunately, shareholders have limited influence over such decisions. If this diagnosis is correct, public company mergers and acquisitions are quite harmful, and would be caused by a small number of people who work promoting such deals.

However, hostile takeovers by privately held companies of poorly managed publicly held companies may be very value creating.

We haven’t had time to properly investigate the evidence.

Artificial intelligence research

Some experts fear that speeding up the development of machines that are more intelligent than humans could be a very dangerous activity. Before we do so, they think we should figure out how to control them to ensure they don’t do things we wouldn’t have intended. Others, while agreeing that AI is dangerous, are not convinced that speeding up AI development overall raises the risks. We didn’t feel we could adjudicate this issue in such a short time.

Multi-level marketing

Multi-level marketing programs are a variant on pyramid schemes in which people are tricked into selling a product because they don’t appreciate the speed of geometric growth. Despite looking very much like a scam, they remain legal in most countries. However, I am not sure how much actual suffering or waste are produced when people are encouraged to participate in such schemes, relative to just transferring money to those who start them.

Take our quiz to find out which jobs offer you personally the greatest potential for social impact.

Notes and references

  1. Though the list is numbered, we didn’t have enough information to order them from worst to least worse.

    1. A National Bureau of Economic Research working paper published in November by Thomas Piketty, Emmanuel Saez and Stefanie Stantcheva found virtually no supply-side effect from cuts in the top tax rate since 1975. That is, there was no significant increase in output resulting from them; hence there would be little negative output effect from raising the top rate. Consequently, the revenue-maximizing top rate may be as high as 83 percent, they estimate.

    2. A National Bureau of Economic Research working paper published last month by Christina Romer and David Romer estimates that the revenue-maximizing rate is 84 percent. Even assuming a higher response by the wealthy to earning income – something economists call elasticity – than the Romers believe is likely, the top rate could still rise to 73 percent before a Laffer curve effect set in.

    Would a higher top tax rate raise revenues?