I've seen this question pop up periodically, usually addressed in the context of a single, usually historical, example. Bell Labs seems to most often get the nod as the canonical "Blue Sky" corporate R&D lab, though Xerox, DuPont, and IBM are others. DuPont specifically is the subject of a book by David A. Hounshell, Corporate Strategy: DuPont R&D, 1902--1980.
I'd suggest a broader consideration:
- What are examples of extant or historical commercial R&D labs?
- What large monopolistic companies haven't developed a significant R&D reputation, either for failure to establish a significant such operation, or a failure to develop awareness of their operation?
- What business sectors generally have failed to establish a strong R&D tradition?
I'm not going to go into a full answer for obvious reasons (a full response would be an interesting PhD thesis or course of study, not an HN comment), but we can outline some interesting aspects...
The companies which did establish large R&D labs have tended to be in high-tech areas, largely information technology though also chemical and hardware fields (DuPont, arguably Xerox).
Specifically I'm aware of: Bell Labs (AT&T), PARC (Xerox), IBM Research, Kodak Research Laboratories, DuPont Central Resource, RCA Labs, General Electric Research Lab, Westinghouse Research Labs.
Notably absent from singificant US industrial history: Railroads, shipbuilding, Steelmaking (e.g., US Steel), oil (Standard Oil), electric power utilities generally, and the financial services industry.
The aerospace sector stands out for having far less major corporate R&D research (or prominence) than one might otherwise expect. That's not to say that there are no such operations: Boeing, Lockheed Martin, and Northrop Grumman all have R&D operations. A possibly explanation is that to a major extent Nasa and US military branches effectively provide R&D cover for the sector.
My general sense is that there's less such activity in the sourcing sectors --- agriculture, forestry, fisheries, and livestock --- than might be expected, though Bayer (including Monsanto), Syngenta, DuPont, BASF, Cargill, and John Deere do have some activity. Again government research is significant (e.g., USDA, Fisheries & Oceans Canada (DFO), and significiant academic research (land grant universities in the US).
Manufacturing, construction, and transportation generally seem to have a weak presence. To the extent that auto safety labs exist, many are strongly supported by the insurance sector. Device certification (e.g., Underwriters Laboratories) are strongly similar. Both also see government consumer / product safety operations.
Healthcare research tends to be organised around government units, academic research, and specifically-organised new-product (mostly pharmaceuticals, though also devices and processes) commercial operations, the latter often spin-offs of larger established pharmaceutical companies as a risk-isolation / diversification process. (Organisation of Novo Nordisk, recently notably for Ozempic, is its own fascinating story.)
My general sense is that wholesale, retail, and financial services have little R&D in in the spirit of Bell Labs and similar Blue Skies research. There's quite a bit of tightly-focused product development, applied, and targeted research, but in scope, spirit, and freedom that's quite a different approach. There's some R&D in academic settings (retail/shopping behaviours, e.g.), or by government (the US Federal Reserve and the monetary system, adjunct to the financial sector).
Teasing out some general trends, corporate Blue Sky research seems to be favoured by:
- High tech fields, notably information technology, electronics, pharmaceuticals, biotechnology, and to a lesser extent aerospace and energy research.
- Less favourable sectors: traditional manufacturing, utilities, wholesale & retail trade, heavy industry, extractive industries, and transportation.
- A strong and long-term stable corporate structure. Note that AT&T was initially founded in 1885, whilst Bell Labs wasn't established until 1925, 40 years later. DuPont preexisted its own labs by a century (founded in 1803 and 1903 respectively). RCA (1919) predated its labs (1942) by 23 years. Xerox (1906) predated PARC (1970) by 64 years. Labs don't spring up overnight, and even large and technical companies might not form labs if their tenure is short.
- Monopolistic tendencies. Labs tend to form around monopolies, though this may be derivative of longevity. Not all monopolies form labs.
- Payoffs to R&D. In Hounshell's book on DuPont, much is made of the comparative fecundity and sterility of early (1930s) and later (1960s, 1970s) research. Effectively, the viable product search space had much early low-hanging fruit, particularly in the 1930s (most plastics we know of today were discovered over roughly a decade), and the failure to find marketable products or enhancements by the 1960s (notably a synthetic "leather" product, "Corfam").
- Insufficient or ineffective government-lead or academic research. In numerous fields, the most significant R&D still occurs in government / academic settings.
- Corporate mentality and mindset, possibly. There seem to be fields, including those with significant research potential in which R&D and sharing cultures simply don't exist to any great extent. I'd put finance and insurance in this group. "Why" remains a good question.
I'd suggest a broader consideration:
- What are examples of extant or historical commercial R&D labs?
- What large monopolistic companies haven't developed a significant R&D reputation, either for failure to establish a significant such operation, or a failure to develop awareness of their operation?
- What business sectors generally have failed to establish a strong R&D tradition?
I'm not going to go into a full answer for obvious reasons (a full response would be an interesting PhD thesis or course of study, not an HN comment), but we can outline some interesting aspects...
The companies which did establish large R&D labs have tended to be in high-tech areas, largely information technology though also chemical and hardware fields (DuPont, arguably Xerox).
Specifically I'm aware of: Bell Labs (AT&T), PARC (Xerox), IBM Research, Kodak Research Laboratories, DuPont Central Resource, RCA Labs, General Electric Research Lab, Westinghouse Research Labs.
Notably absent from singificant US industrial history: Railroads, shipbuilding, Steelmaking (e.g., US Steel), oil (Standard Oil), electric power utilities generally, and the financial services industry.
The aerospace sector stands out for having far less major corporate R&D research (or prominence) than one might otherwise expect. That's not to say that there are no such operations: Boeing, Lockheed Martin, and Northrop Grumman all have R&D operations. A possibly explanation is that to a major extent Nasa and US military branches effectively provide R&D cover for the sector.
My general sense is that there's less such activity in the sourcing sectors --- agriculture, forestry, fisheries, and livestock --- than might be expected, though Bayer (including Monsanto), Syngenta, DuPont, BASF, Cargill, and John Deere do have some activity. Again government research is significant (e.g., USDA, Fisheries & Oceans Canada (DFO), and significiant academic research (land grant universities in the US).
Manufacturing, construction, and transportation generally seem to have a weak presence. To the extent that auto safety labs exist, many are strongly supported by the insurance sector. Device certification (e.g., Underwriters Laboratories) are strongly similar. Both also see government consumer / product safety operations.
Healthcare research tends to be organised around government units, academic research, and specifically-organised new-product (mostly pharmaceuticals, though also devices and processes) commercial operations, the latter often spin-offs of larger established pharmaceutical companies as a risk-isolation / diversification process. (Organisation of Novo Nordisk, recently notably for Ozempic, is its own fascinating story.)
My general sense is that wholesale, retail, and financial services have little R&D in in the spirit of Bell Labs and similar Blue Skies research. There's quite a bit of tightly-focused product development, applied, and targeted research, but in scope, spirit, and freedom that's quite a different approach. There's some R&D in academic settings (retail/shopping behaviours, e.g.), or by government (the US Federal Reserve and the monetary system, adjunct to the financial sector).
Teasing out some general trends, corporate Blue Sky research seems to be favoured by:
- High tech fields, notably information technology, electronics, pharmaceuticals, biotechnology, and to a lesser extent aerospace and energy research.
- Less favourable sectors: traditional manufacturing, utilities, wholesale & retail trade, heavy industry, extractive industries, and transportation.
- A strong and long-term stable corporate structure. Note that AT&T was initially founded in 1885, whilst Bell Labs wasn't established until 1925, 40 years later. DuPont preexisted its own labs by a century (founded in 1803 and 1903 respectively). RCA (1919) predated its labs (1942) by 23 years. Xerox (1906) predated PARC (1970) by 64 years. Labs don't spring up overnight, and even large and technical companies might not form labs if their tenure is short.
- Monopolistic tendencies. Labs tend to form around monopolies, though this may be derivative of longevity. Not all monopolies form labs.
- Payoffs to R&D. In Hounshell's book on DuPont, much is made of the comparative fecundity and sterility of early (1930s) and later (1960s, 1970s) research. Effectively, the viable product search space had much early low-hanging fruit, particularly in the 1930s (most plastics we know of today were discovered over roughly a decade), and the failure to find marketable products or enhancements by the 1960s (notably a synthetic "leather" product, "Corfam").
- Insufficient or ineffective government-lead or academic research. In numerous fields, the most significant R&D still occurs in government / academic settings.
- Corporate mentality and mindset, possibly. There seem to be fields, including those with significant research potential in which R&D and sharing cultures simply don't exist to any great extent. I'd put finance and insurance in this group. "Why" remains a good question.