Firm-specific skills and working from home

by John Q on September 17, 2020

One of the central features of the debate about working from home is that it leads to the loss of random, but productive, encounters with colleagues. I’ve responded with the observation that some of my best research ideas have come from largely unplanned encounters on the Internet.

It’s just struck me that there is a conflict here between the interests of workers and those of firms and managers.

A lot of universities (or, more precisely university managers), think of themselves as developing and promoting a corporate brand. In this context, research collaboration within the university (particularly if it is trans-disciplinary) is viewed very positively, while collaboration with other universities is less well-regarded. But for individual academics, the big rewards come from high-profile work within tightly defined fields, which implies a desire for collaboration with other people in the same field who will, in general, be located elsewhere. While intra-university collaboration may be rewarded in internal promotion decisions, the outside opportunities are greatest for people with external collaborators. Those outside options are routinely used as a bargaining chip in negotiations over salary.

This issue isn’t specific to universities. Labor market theory distinguishes between firm-specific skills and general skills (which are of value to any employer). Back in 1964, Gary Becker made the argument that firms would be willing to pay the cost of firm-specific training for their workers, but not for general training which increases their outside opportunities. (This seems entirely convincing to me, although the empirical evidence I found on a quick search is both limited and inconclusive).

What applies to training also applies to serendipitous encounters. Collaboration with co-workers can enhance productivity within the firm, but doesn’t do much for your market value outside. Conversely, if workers enhance their productivity at home by making more use of industry discussion groups, Skype chats with people in other firms who are addressing similar problems, and so on, that enhances their bargaining power relative to their employers.

In this context, it’s striking that the hardest push for a return to the office is coming from the finance sector, led by JP Morgan. Even though textbook finance is all about hard numbers on earnings, risk and so on, the industry actually operates largely on personal contacts, networks and exchanges of favours, particularly information. That’s why it’s concentrated in a handful of global cities, and why so much attention is paid to issues like “poaching” of staff, no-compete clauses and the like. It’s obviously in the interests of employers to build up internal networks and control external interactions.

As with all these issues, my ideas here are provisional and almost certainly wrong in some respects. So, feel free to correct me.



Zamfir 09.17.20 at 8:02 am

In my own (limited) experience in technical fields, I very much recognize the tension between corporate-specific specialization, and more generic job skills. I am not so sure about the wage negotiation side though.

In my experience, companies push people into overly specialized roles simply because (in the short term) the company highly need those specialized roles. More than an attempt to drive down wage negotiations. If anything, companies tend to pay on the high side for “trapped” specializations. They are often called golden cages – good wage, with more autonomy and lower workloads than the more general jobs at the same level. I think economists call it an “efficiency wage” – the company wants to discourage such people from looking around too much, by paying somewhat over the market wage.

The tension point is then more about risk than wages- what if some years later, the specialist is not needed anymore, or they want to leave for other reasons? People usually want to build general skills as a safety net for the future, more than as a bargaining chip in the here and now.

That has an impact on how the company looks at external contacts. Companies tend to encourage people to build a network of external contacts – it’s useful to know outside practices. It creates the chance that the worker leaves, but also that they bring in a useful new people. The main problem with specialized roles, from a corporate POV, is that the job market is illiquid. That might help to keep your own specialists trapped, but it also makes it hard to find new ones, and the second effect weighs heavier.


Phil 09.17.20 at 9:05 am

It used to be understood in universities that faculty are a loose assemblage of lone scholars with their own careers to make, with no a priori commitment to your institution (or to whoever else it happens to have corraled) and every reason to network and collaborate externally – and that the job of the management, insofar as there is a management, is to give them reasons to stay. Kind of an anti-firm, or the limit case of a firm (the maximum independence for the maximum number of people consistent with a firm still having reason to exist).

I once heard the head of a Law School at a Russell Group institution regale the department with the story of the Conference of Heads of Law Schools he’d attended, where one speaker after another had delivered their more or less managerial thoughts on How to Lead a Law School. He’d then stood up and said “I think you just need to make sure that people are able to do interesting research, because if you haven’t got that, you haven’t got a Law School”, and sat down. This was about twelve years ago now, but seems longer.


Tm 09.17.20 at 9:21 am

“Conversely, if workers enhance their productivity at home by making more use of industry discussion groups, Skype chats with people in other firms who are addressing similar problems, and so on, that enhances their bargaining power relative to their employers.”

It is amazing how many IT people spend significant amounts of time (private time, or are they doing it on company time?) contributing to open surce projects, answering technical questions by strangers on platforms like stackoverflow, writing technical blog posts and so on. They do this in part because they are really nerds and love doing it, in part they speculate that it enhances their market value. Whether that is true however I am skeptical. Perhaps for some, the top “influencers”, but I doubt it matters for most employers. Anecdotally, I have made the experience that such colleagues are not good teamplayers.


Tm 09.17.20 at 11:33 am

“the job of the [University] management, insofar as there is a management, is to give them reasons to stay”

There once was a college Dean who told the faculty in one of her first public talks: “you are all replaceable – for each of you, there are ten people out there who would love to do your job”. Needdless to say, she didn’t stay Dean for long (two and a half years). Of course she had been granted tenure when she was appointed Dean and remained in a cushy faculty job. Oh these ironies. If only those who preach social darwinism would lead by example – most of them wouldn’t last long.


Trader Joe 09.17.20 at 12:06 pm

“the hardest push for a return to the office is coming from the finance sector, led by JP Morgan.”

JPM is seeking to return its traders ASAP for a very practical reason – they are frankly more worried about fraud than they are about health risks.

As we know from the newspapers of the last 100 years, traders manage to commit fraudulent acts of various kinds even when operating on a trading floor where all of the calls are monitored, most everyone is on camera, external devices are highly controlled and there are a dozen people sitting within 6 feet of you to overhear your every action.

When people are trading from their bedrooms and only a fraction of these controls exist the risk that someone can get up to wrong doing, over trade limits or otherwise be a bad actor is exponential. Any of these can cost $100s of millions or even billion.

That’s why JP (and others) are in a rush to return at least some of their people to monitored trading floors – there are literally billions of things that can go wrong and someone getting sick from riding the subway is only a small part*.

*some of the firms are talking hiring private cars to keep the ace traders off of public transportation….a small cost relative to the dollars at risk.

P.s. I’m not a JP trader


Robo Friend 09.17.20 at 12:06 pm

Having worked exclusively in research roles in industry, I would say this is a reasonably fair breakdown with a couple caveats…

1) As a researcher in industry, there are often strict intellectual property / trade secret concerns with collaboration outside the firm, which is another reason firms with those concerns tend to encourage internal collaboration (including with higher compensation for good collaborators).

2) Inside firms, there often tends to be a lot of “process knowledge” to do the job well. Knowing things such as who might have relevant data and the ability to release it, who might serve as a senior “sponsor” for a project, etc. These can be huge frictions if you have to chase them down formally, but for those who have a lot of light and informal connections from “so what are you working on?” types of conversations, these frictions tend to disappear and productivity gets higher.

3) Aside from the managerial incentive to give compensation for productive internal collaboration mentioned in 1, I think there are also very strong limits to how much managers really know about how much worker productivity comes from “firm-specific” vs “general” sources. While these are cleanly separable in an economic model predicated on that separation, in practice, managers just see an employee with higher productivity in terms of output, or an employee seen as valuable by other colleagues who collaborate with them.


Phil H 09.17.20 at 2:13 pm

Re: firm-specific training
This is only anecdotal, but I worked for a few years at Huawei, and the feeling there was that they put quite a bit of effort into making sure their tech guys’ skills weren’t too transferable. People worked in narrow specialisms, and the training focused on “how to use the Huawei tools” not “how this technology actually works.”
If a lot of remote working eventually leads to significant atomisation, it might be that those companies which keep their workers in one building end up winning, because keeping their workers together may keep them focused more on shared goals, and that may mean that those companies can do more effective long-range planning.


Ebenezer Scrooge 09.17.20 at 2:23 pm

It’s not all structural. Some of it is a matter of tradition. Let me tell an anecdote.

I remember dealing with an English colleague in my field–banking law. After exchanging the usual pleasant lies on how many drugs we had(n’t) done in the 1960’s–and embarrassing the junior in the room–the conversation drifted to a listserv I belonged to. On this listserv, U.S. lawyers exchanged technical legal information. (Most law may be priestcraft in the service of power, but there are some genuine technical issues.) People read the listserv to be informed. People posted to solicit an answer, or show their legal chops to the community. Everybody was happy.

My English colleague (who worked for a US law firm) was jealous. He thought that English lawyers were too competitive for such a listserv, and simply preferred to silo the information in their firm.


oldster 09.17.20 at 2:25 pm

“It used to be understood in universities that faculty are a loose assemblage of lone scholars with their own careers to make, with no a priori commitment to your institution”

Yes, that is not a new thought. But it is not a very old one, either.

I taught in higher ed from 1985 to 2013, at five different places. Your description applies to me and my experience. But my father was also a university teacher, from 1945 to 1990, and what you describe would have been unimaginable to him.

He was a consummate university man — his intellectual community was his department, and then his university as a whole, where he served his whole life. He taught chemistry, but he lunched with anthropologists, served with linguists, and dined with theologians. Every member of that university was his colleague, and he had closer bonds to people in the music department than he had to chemists across the nation. He was as big a booster of the university that employed him as were its graduates and alums (he was neither).

My career was entirely different from his, and more like what you describe as the “lone scholar” model. At each of the universities I worked at, I was siloed in a department and had no contact with people in other departments. It did not matter to me, because my colleagues were people in my own discipline, generally at other universities. I met them at conferences, gave papers in their departments, read their papers in draft, chatted over email. They were at dozens of universities on several continents.

My father’s career was marked by his exclusive life-long loyalty to one university. I never cared much about the institutions I worked for. Loyalty was not given, and not demanded. I got on with the people in my departments, but my real intellectual playmates were mostly in other time-zones.

So the “lone scholar” model is not really that old. I would say that it really took off with the advent of email. Not that there was no inter-university collaboration before that. But email made collaboration with remote scholars far faster than it had ever been before. Low-priced jet travel must have contributed as well.

Here’s another reason to think that the “lone scholar” model is not that old: schools of thought used to line up, literally, with institutional boundaries. There was a Chicago approach, shared by nearly everyone who worked at Chicago, and extending beyond it largely by the distribution of its graduates. There was Marburg school of Neo-Kantianism, taught at Marburg. There was a Frankfurt School — and it was so-called because it happened in Frankfurt. These were essentially local phenomena: if you wanted to be at the cutting edge of that movement, you had to move to that place.

I do not think the same thing would happen nowadays. Movements can be co-founded by scholars living on different continents. There are probably still “schools” and proprietary approaches. But they are not limited by geography or institution.


hix 09.17.20 at 6:34 pm

Investment banks are just crazy. Don’t think their behavior can be properly explained by any rational exploitation of workforce plan developed by some sociopath. Dysfunctional culture all the way if you ask me. Just remains stable because they got a solid oligopoly with no danger of competition from outsiders that act saner. Almost sectarian behavior. Oh whatever, why bother with the qualifications, at least one individual big player Goldman Sachs essentially is a sect. Other than that, reading the words “corporate” and university in one sentence still causes me nausea.


Priest 09.17.20 at 8:19 pm

I can confirm at least anecdotally about reluctance to provide training. At my workplace we had a major software change years ago, and my then direct manager wanted to send the three of us in a support role for training/certification on that platform (later just myself when he couldn’t get approval for all of us), and higher management refused to sign off on it. Having that credential would have, for a time at least before inevitable software changes rendered it obsolete, given me a potential “exit visa” should the job circumstances become less tolerable. Instead they followed the “he’s a bright guy, he’ll figure it out” path and I just had to learn on the job.


Jason Weidner 09.17.20 at 9:14 pm

“In this context, research collaboration within the university (particularly if it is trans-disciplinary) is viewed very positively, while collaboration with other universities is less well-regarded.”

In my university, this is definitely not the case. Research collaboration with academics from other universities, especially from other countries, is highly encouraged and even rewarded. Now this is probably due to it being a Mexican university that is trying to establish what would be a new reputation for research excellence. So I assume this is an exception that does not undermine the overall claim, but I just wanted to point out a contrary example.


Moz in Oz 09.17.20 at 11:11 pm

In my field of software there’s a lot of certification engineering going round bit corporations, and those certifications are deliberately short term but portable. Many larger employers will provide training to get the latest version of relevant certifications, even though they’re generally expensive and do allow staff to migrate. They’re effectively insurance when large projects inevitably fail “but all our staff were certified”.

I’ve been certified twice: once for a brand new language/platform, where the company I was contracting to required it; and once when working for a major bank who offered training even to contractors like me. That was entirely built around the need for automated assessment. Which convinced me that insofar as we can test for programming ability, those tests do not do so. In the former job my role was essentially cleaning up messes made by people who had also passed the test but could not practice even the basic tenets of software engineering.

On the hiring side, I’m biased against people with industry certifications because it likely indicates incompatibility with my approach. I would rather work with someone who thinks than someone who can memorise gibberish.


hix 09.18.20 at 1:23 am

Universities and investment banks both look like out liner cases. Regarding the typical big corporate employer one can expect careers in Japan to be structured around firm specific knowledge first if not entirely. In the US the normal way of things is to de-emphasize firm specific knowledge over knowledge relating to one’s specialty. People move accordingly – switching employers at a fast rate in the US while sticking quite narrow to their general specialty. In Japan the employer is rather switched sometime between once till never and the training is very generalist. One qualification: That idealized model going back to corporate culture work done during the big Japan scare in the 80ths/early 90ths was only ever correct for a relatively small elite workforce.
Ouchis back then (1984) tried to find “Z” firms – American ones with a corporate culture more similar to the Japanese standard than was typical. Looking back at those companies today, they tend to be bankrupt, or meaningless players with a standard US culture today.

The cynical take on those old Z firms can be found in Kuda’s Engineering culture. He convinced me that the typical Z corp is nothing to be desired. Albeit modern corporate US tends to expect similar or worse excessive long hours from its professional workforce without the job security. Not an improvement. Regarding the home office theme, it’s interesting to note that Digital equipment was fast to introduce free home work stations with a connection to the firm’s intranet.


J-D 09.18.20 at 1:31 am

… At my workplace we had a major software change years ago …

Would that be the famous/notorious ‘New Testament’ upgrade?


marcel proust 09.18.20 at 2:39 am

@oldster wrote:

Jason Stanley blogged about this evolution nearly 15 years ago, also comparing his and his father’s experience.

See also this comment on Stanley’s thoughts.


JimV 09.18.20 at 3:07 am

At General Electric in the 1970’s I learned heat transfer, vibration analysis, metallurgy, Theory of Elasticity, material failure analysis and the accompanying mathematics. Also Unix and ANSYS. I heard anecdotally that sometime after Welch took over as CEO in 1980 he wondered aloud why GE did all that training, “Why not outsource the training and just hire people with Master’s Degrees?”

Also pre-Welch, the company philosophy was to invest profits in R&D, and plant and equipment improvements. Welch’s philosophy was that profits should be used to buy back company stock.

I don’t know, but think much of the pre-Welch philosophy was a hold-over from lessons learned in WWII: loyalty up requires loyalty down, train your soldiers, train your sub-officers to replace you, the army with the best equipment and logistics has an advantage, and so on. In a necessary and difficult war, cheating and short-cuts probably don’t get you far. In peace, the sociopaths rise to the top. That’s probably simplistic, but all the managers I started under served in WWII, were the best managers I ever had, and hated Welch. (Most of them were gone by 1985.)


faustus 09.18.20 at 4:26 am

Maybe it’s my field, but this seems completely not true to me. In my field (Public health) you basically have zero chance of getting published in one of the top journals if you aren’t part of a major international collaboration, and if a university want to build a brand in this topic they need their academics to work in these collaborations. Furthermore the skills that are most important to good publications and a good publication rate are basic skills that work in any institution. The primary firm-specific skills you need are how to use your university’s online teaching platform.

I don’t think this is just for public health academia either. Conferences are a core part of most academic disciplines and are explicitly designed to foster outside collaborations, and in most fields the primary source of research funding is a national organization that has a standardized submission format and anonymized review, so that the process of getting money is the same in every institution, which means by design that the required skills for this task are not firm-specific – yet despite this universities keep demanding greater success in this area, not less.

In the non-academic areas most closely connected to my discipline – medicine and the pharmaceutical industry – we see the same relative weight of external vs internal skills. Almost all of medicine is governed by standards and accreditation systems managed nationally, and even GPs in Australia – the smallest “firm” in that field in the country – are required to maintain skills through an annual points-based system that is set nationally. In pharmaceuticals every firm is required to follow the same set of nationally-established rules in a highly regulated market place, using the same skills. Yes these companies obviously have a large amount of intellectual property and internal systems and knowledge, but ultimately they’re restricted in how much firm-specific skills matter by the regulatory environment.

I don’t know how it works in other parts of academia, but I would guess it’s similar in other aspects of the physical sciences, and possibly even worse in areas like e.g. nuclear or plasma physics or astronomy where everyone needs access to a limited amount of nationally funded massive basic research equipment.

Also, once you’re past a PhD in my experience universities leave it up to you how to do further training in outside skills, and they would never stop you or discourage you from going on a training course if you have the money. And I really don’t think universities measure your employability based on nebulous skills you got from skype chats with external collaborators: they measure it from how much you teach, how much you publish, and how much grant money you earn. All of these depend on you having skills that are not specific to your firm at all, since they depend on your ability to negotiate a complex web of international journals and national funding bodies.

Maybe economics is weird but in the physical sciences I think your university won’t care about you at all if you don’t collaborate, and universities want you to do this precisely so they can build their brand. So I think this post is simply wrong?


Phil 09.18.20 at 8:56 am

oldster – interesting angle.

I wonder if the difference between your experience and your father’s is actually one of degree (no pun intended). Cynically self-steering lone scholar though I am, I would be perfectly happy to call myself a Fellow of Winstanley College Oxbridge and even to take part in the Life of the College alongside other happy Fellows. The difference is that because an Oxbridge fellow is (a) something to be (as John Lennon didn’t say) and (b) well rewarded, formally and informally. (In my student days I once attended a dinner where I sat next to a fairly senior academic, who complained about the length of the menu – “not two meat courses again!”) Institutions like Oxbridge colleges know they’re herding cats, and they do it by keeping their cats well-fed. Which is, perhaps, the way that more institutions – or a larger proportion of what was a smaller number of institutions – used to conduct themselves.

Institutional schools of thought do still exist in a smaller and more formalised way, in the form of Research Centres and similar. Nobody’s got time to let them develop of their own accord, though. A professorial colleague told me once about the time he’d brought in a multi-million pound research bid – all very prestigious, of course, but it did mean that he had his hands full for the next couple of years getting the budget spent and the research done. He had a letter of congratulation from the VC himself, which concluded “I look forward to your next bid”.


reason 09.18.20 at 9:03 am

Priest @11
Yes, but in a way they are right. What is actually really valuable is having used something, not just having done the courses. And what is especially valuable is having made the transition from something else.

I’m a bit with Ivan Illich on this – testing and training don’t belong together they should be separate categories done by different organisations. We shouldn’t be concerned about how somebody learnt something – we should be worried about if they are competent.


nastywoman 09.18.20 at 10:48 am

Now isn’t it a completely NEW LIFE – with already – how many people working from home?

And do you guys know – that there are predictions that soon over 50% of ALL people who are working – will be working from their home office?
AND we finally will have all of these useless office spaces for payable housing.



Michael Cain 09.18.20 at 2:33 pm

oldster @ 9: Similar things happened in many large corporate research labs. It used to be the case that a lot of research engineers stayed at the same company for an entire career (eg, the Bell Telephone Laboratories of song and legend). You met specialists in all sorts of fields. Then came the age of the start-up with small groups of people doing one very narrow project, often in hopes of being bought out by one of the bigger players. For the big players, research became something that you purchased more than something that you did.


Dwight L. Cramer 09.19.20 at 4:47 am

I doubt that it is possible to meaningfully generalize across such a broad variety of social and cultural actors by aggregating them into the construct of the ‘firm’. But I’m inclined to agree that, along with a host of other changes, the pandemic (assuming that it continues to disrupt for another year or more) will accelerate, intensify, bring into focus, etc. workplace changes. This post and the comments responsive to it are interesting to consider in conjunction with your other post concerning the ‘credentialization’ of professional training.


Phil 09.19.20 at 11:32 am

faustus @18:
once you’re past a PhD in my experience universities leave it up to you how to do further training in outside skills

True, apart from the diversity training, the health and safety training, the data protection training and the counter-radicalisation training – all of which are not only mandatory but have to be refreshed on a three-yearly basis. (Meaning that I’m overdue for my training on counter-radicalisation, despite (a) having done the only programme we offer staff once already and (b) having spent the last few years teaching, writing and conducting externally-funded research on counter-radicalisation.) Oh, and the teaching training, which isn’t mandatory so much as strongly encouraged – and about which I have Mixty Motions, inasmuch as it left me simultaneously fired up about the possibilities of creative pedagogy and deeply depressed about the likelihood of ever being able to do any of it. But that’s for another thread.


hix 09.19.20 at 12:17 pm

Wondering how exactly the push started at JP Morgan or which departments they are particular eager to return. My reference to investment banks as cults does not really cover all that much of JP Morgans business. It is only particular true regarding the core investment banking business. Trading is already less cult like and credit banking is an entire different story.

German public administration is a sector where one would think many do one case file after another jobs are particular easy to do remotely. You got standardized procedures, a self-motivated workforce largely left on alone with wide control spans anyway and cultural socialization is largely already finished upon finishing specialized education programs. It is also rather explicit with all the legalese focus. Most people stay on the job their entire life anyway, so training newcomers is not that important. But in reality it’s a sector that is utterly resistant, based on concerns about data security. It seems to be the norm that people are required to have a separate room that cannot be accessed even by their married spouses, or at least to have a home large enough so one can pretend this is the case that is. Banking more broadly also seems to be in the we got legal concerns corner as a core motivation – at least according to some official announcements for motivations to abandon remote work.


faustusnotes 09.20.20 at 4:29 am

Phil the only training I have to do is annual training in research ethics, and again it’s not firm-specific – every university in Japan has to provide the same content according to a standard set by the ministry of education, and typically obtain it from a very small number of national organizations. I’ve certainly never heard of counter-radicalization training!


Dona Receita 09.22.20 at 12:52 am

due to the scenario we had to reinvent ourselves and learn to work at a distance.

Comments on this entry are closed.