The commodification of labour coincides with technological advancements in production: it is perhaps most noticeable in relation to factories. Factory processes replaced the labour once done by skilled tradespeople. It might not be obvious how this trend has continued to this day and is now affecting professionals in complex fields including those in the data sectors. I am talking about the "made to order" and "off the shelf" acquisition of labour commodities. What I describe as commodities enter the production system as contract and temporary employees hired to handle specific jobs and projects beyond which their services might no longer be needed. Although this is the general idea behind labour commodification, in practice there can be some important structural risks.
I will momentarily reflect on my father's periodically visceral stories of meat packing facilities. He was a mechanic for much of his life. One job that he used to do was dismantle, clean, and reassemble meat-cutting equipment. He said that as he was trained, he also trained others, and this was how people learned to perform the job properly; in effect, he was posing the job of cleaning the equipment as a skilled trade. Many years after he left the industry, I recall extensive coverage in the local news of listeria outbreaks in meat packing plants. Now, for a mechanic actually examining the equipment every day, the risks posed by improperly trained staff or perhaps no staff at all are fairly apparent. The same can't be said when human resource decisions are made in settings disassociated from production realities.
Fields involving extensive amounts of data - I include programming among them - can also suffer from being disassociated from decision-making within the context of commodified labour. On one hand, it might seem logical to sporadically bring in individuals only as required - or to have them do particular jobs only when needed - based on perceptions of efficient production. I have chosen actually not to dispute the argument of "efficiency" since the metrics are plainly evident; business decisions can and perhaps should be made to achieve specific measurable objectives. I will counter this argument by posing some of the problems faced by skilled tradespeople. Not everything can be purchased off the shelf. Useful activities such as training do not always register on the metrics except perhaps as labour costs.
The idea behind creating a profession is to bring together a body of people with certain shared competencies. This might not seem like the commodification of labour, but actually it is definitely. Once pooled with comparable abilities within a particular field, it is possible to replace one lawyer with another; go to a different family doctor to get the same or similar quality of care; hire a qualified accountant from a market containing many. So it can be tempting to say, if this model works in different fields, the results should be similar in relation to those in the data sectors. However, consider the structural implications of commodification: change would likely be driven by extraordinary events through the executive rather than by routine interaction through quotidian processes. Change would tend to occur after problems happen; also, the changes might only overcome the specific problems at hand. Perhaps the most critical structural impact would be the insulation of the organization from learning.
Particularly in relation to data - something that carries attributes of a commodity - I believe we have this focus on the data itself with little emphasis on the weight of acquisition and analysis. There is an enormous amount of learning or structural capital that goes towards the processes and systems in place to collect and make use of data. I want to distinguish between the "learning" and the "systems" associated with data management and analysis. It is possible to have a system that nobody knows how to use. Under normal conditions, the system extends from or emerges as a consequence of the learning. Under the commodification model, I suggest that the situation is rather reversed: there is systems acquisition followed by numerous glorious battles and hell-fire in the ensuing attempts to learn. This is an unstable model.
The commodification of labour can be easily detected in job listings that ask for a number of unusually specialized skillsets; all of these tend to suggest an inability or lost ability to provide training. When I say "inability," this does not mean the organization cannot do it but simply that its business model might not allow for it. Considering the situation in more structural terms, such an organization has placed its operations and perhaps its future on the commodification model. What risk does this pose? Well, if it is true that an organization can set up shop locally and acquire labour commodities to perform specific jobs for production, then it must also be true except in protected markets that any organization can set up shop anywhere to produce the same products or offer similar services. The winner will be the organization that can provide the cheapest prices. So if the worker can be easily replaced under a particular production model, so can the company.
We therefore have this situation where a company that is highly information-oriented might find itself in an inferior competitive position on one hand and also unable to naturally change its internal circumstances. There is a disconnect between what it does and what it has learned to do. I am describing termination by adaptive failure - a problem that I feel is often pronounced during periods of austerity. I suggest therefore that neither complex fields nor complex professions can be easily sustained in times of constrained resources; in fact, reductions in complexity will probably be a persistent aspect of our time. We will see a deterioration in benefits from complexity in the years and possibly decades ahead together with sharp increases in organizational risks. We confront this volatile situation applying a relic of production systems - labour commodification - to the data sectors. Since commodification is difficult apply, and has been an instrumental in the decline of manufacturing, it seems all but certain to fail our needs in the age of abundant data.