Economists in the tech sector

…led by Pat Bajari, Amazon has hired more than 150 Ph.D. economists in the past five years, making them the largest employer of tech economists.  In fact, Amazon now has several times more full time economists than the largest academic economics department, and continues to grow at a rapid pace.

That is from the new Susan Athey and Michael Luca paper “Economists (and Economics) In Tech Companies.”


'Amazon has hired more than 150 Ph.D. economists in the past five years'

And many are undoubtedly doing useful work, such as pointing out how market power is a good thing, just a small step to a much better bottom line for everyone - well, at least everyone that matters.

"..well, at least everyone that matters."

Are we supposed to believe that you are downtrodden and put upon by Amazon? That Amazon hiring technical economists is somehow an immoral act?

Well your comment is passive-aggressive, but at least there wasn't a wall of wiki text. So, that's an improvement.

I'm amazed, I must say. Rarely do I come across a blog that's
equally educative and engaging, and let me tell you, you've hit the
nail on the head. The problem is something that too few people are speaking intelligently about.
I'm very happy that I came across this during my hunt for something regarding

Why? Does anybody know what an economist, a tech economist, does at Come up with optimal differential prices suitable for a 'monopolist'? Are they on the staff to counter FTC claims of monopoly (so they can come up with studies showing why Amazon is not a monopoly in relevant markets)? If so, I would imagine a one-time consultant would be cheaper. Are they just signaling they are cutting edge? Buying favors on Capital Hill with lucrative job offers? (but economists have no political clout so I think not). Are these 'PhD economists' really also MBAs? At a loss to understand why Amazon needs so many economists.

I don't see what is so inexplicable about this.

They have a vast collection of data about their customers, sales, and business processes, and someone has to analyse it. Economists would be an important contributor to this.

This would include understanding what has happened (who their customers are, what is most profitable, what processes are most efficient, etc.), as well as predictive modelling of what will happen if they make changes.

@dan1111 - what is inexplicable is why you need PhD economists to do this kind of work. Somebody with knowledge of how to use an Excel spreadsheet, and maybe with a degree in business administration, could do that. After all, in a spreadsheet you would have a range of parameters and somebody in management would make a business decision based on x,y,z assumptions about the parameters. That's how multi-billion dollar M&A's are done, by way of example, that I'm familiar with. So I think the 'real answer' is that this is signaling by Amazon to the world that they are so smart (hiring "PhD economists", wooo).

Well they'd have to pay the MBA's $100K per year and they probably pay the PhD's $200K per year. So, you're talking about an extra $15 million per year. Maybe Amazon believes the difference in quality is worth the money.

Ray Lopez from a comment written below: "Perhaps the price of PhD economists has come down."

"The National Association of Business Economics' 2010 salary survey reported that its members with doctoral degrees earned an average of $150,000 a year, "

I have a friend who is an economist at Amazon. You can't do that kind of work in Excel. The data literally doesn't fit in an excel spreadsheet (which hold up to ~1 million rows of data). Even if it did, any non-trivial analysis would be impractical, including simple matrix multiplication.

In my experience, MBAs are often quite hopeless when it comes to quantitative skills beyond basic arithmetic (which is all that's really needed in things like strategy or valuation). Good luck having them code up a simple machine learning method. Maybe they could run an OLS regression in Stata, if they remember it at all from their grad school days.

Engineers have the quantitative skills but they are often terrible at causal inference (based on my discussions with people in the software industry). As I understand it, the economists at tech firms work with the engineers to implement experiments and collect the data. The economists analyze it, summarize it and present it (and recommendations) to the decisionmakers (often with an MBA).

A big change over the decades in business is that corporations can now more easily implement the kind of clever pricing strategies dreamed up by economists than in the old days when retailers often just marked everything up the industry standard amount over their wholesale cost because it was too much trouble in terms of paperwork to follow seemingly more profit maximizing strategies.

I can recall getting my MBA in 1982 and quickly being disillusioned by how boringly supermarkets and CPG manufacturers decided upon their prices. To work at Amazon today would be a lot of fun for my 1983 self.

An example of what many of those PhD economists have been up to:

Transfer pricing is a high stakes game.

This is a routine tax avoidance suit. Tax lawyers are constantly fighting with the IRS on what is an "arms-length" fair value transaction. You don't need "PhD economists" for this work, just a good team of tax lawyers (who will hire maybe one or two of these economists as consultants).

You have no idea what you are talking about. These suits are all about economic studies to support the transfer price. Every transfer price established by a major multinational is supported by transfer pricing studies written by economic "experts"--not lawyers. If they haven't done a study they are almost per se not in compliance with transfer pricing rules (in the US, IRC section 482 and the regs, etc., thereunder. Economists are also always called as experts in litigation for this type of case. It is a major source of employment for PhD economists both within industry and for major law and accounting firms. Been there, Ray---you haven't (not a claim based on identity, but experience).

And the case cited was by no means "routine".

Let me fix that headline for you Tyler : "Economists in the retail sector . " A sector , by the way , that has long employed economists .

I was going to say the advertising sector, but retail makes the same point. Calling it "tech" makes it sound like something profound, when it's mostly data collection and data mining in the retail (advertising) sector. There was an article in the NYT this past weekend about how Apple has mostly avoided the problems confronting Facebook, Google, and the rest that rely on an advertising platform for revenues (Apple's revenues are derived mostly from producing and selling hardware and to a lesser (but increasing) degree on user fees).

And people wonder why Amazon's not profitable....

Amazon has been profitable for several years.

They are not as profitable as they could be, because they are pouring so much money into expanding their business.

Random thoughts: Moneyball was about an economist applying the economic math and principles of optimization to baseball. A lot of what economists do seems to involve some form of optimization. Econometrics, although in many ways a branch of applied statistics, seems to mostly revolve around finding the combinations of values of causal variables that optimally explain some set of government economic data. There was a boom in high quality MBA students around 1990. For awhile Wall Street firms were substituting the hiring of Phd economists by hiring instead combinations of statisticians and MBA's. Not sure how that worked out. Engineering companies, which is generally what high-tech firms have as their core skills, isn't focused on optimization the way economics is- well perhaps no other field is for that matter. Perhaps Google's hiring of the economist Hal Varian awhile back has set off a trend. Final Panopticon thought- Will economists in the tech sector increasingly be used to amplify the information advantage in managing and developing stock portfolio's? Will there be an academic discipline competition writ-large between economists at engineering-minded high-tech firms and statistician Quants mostly managed by MBA's on Wall Street?

Yes, thanks, this echoes my thoughts above, as to why you need "PhD economists" to do this kind of work, when an MBA or somebody versed in Excel from your local community college would do just a well. Perhaps the price of PhD economists has come down.

Lets let Amazon spend money how they see fit. To see people questioning their strategy is fine, but belittling their strategy is another thing.

They obviously have some knowledge and ideas of what they want to do. And they have a decent track record. Google hires a bunch of smart people that they might not necessarily "need" right now but they know they want to have a knowledge advantage.

Perhaps Amazon is not hiring run of the mill PhDs or plucking good ones from university teacher roles. It often takes a while to get full professor status so I am going to assume there are many smart PhDs earning "lecturer" salaries that would enjoy 150k/yr. Furthermore, maybe Amazon hires, cultivates and weeds out their "crop" of PhDs regularly.

Lastly, if you are talking about pure research and analysis, no MBA in a 2 year program spends as much time analyzing a specific concept as much as a PhD who spent 3-5 years on a thesis about a niche topic.

Comments for this post are closed