One of the things I love best about geography as a discipline is its emphasis on empirical field research. It's not the only such discipline, and if compared with some hard core anthropologists, for example, geographers might be said to spend too much time behind a desk. But for the most part, academic geographers (and certainly the ones I admire most) spend as much time in the field as their teaching schedules, research grant budgets and personal lives permit.
I have an inherent distrust of the opinions of social scientists who spend their lives behind a desk. You read them often in the newspapers, pontificating about the situation in Afghanistan, how some particular innovation or regulatory change will save the planet from global warming (or vice versa, will bankrupt the country), or what the compensation package should be for a worker in an auto factory. When I read such opinions, my first thought is "Really? And how would you know that to be the case? Have you done your field research?" If the writer does not offer so much as a sentence about their methods, or does not state explicitly something to the effect of "I believe this because I was there", my skepticism rises. If it's opinion I think I might want to rely on, I do a Google search for their bio. And if it looks like they spend all their time behind a desk, I discount their opinion accordingly.
For example, I am convinced that the majority of "informed commentary" offered by scholars in the press or on TV on the subject of auto industry restructuring is made by people who have spent little or no time inside an auto plant. In my teenage years and through undergrad, I worked each summer in auto parts factories in Waterloo Region. Some plants were unionized, some weren't. Some were well-run, some weren't. Each had a fairly complicated set of social relations among those employed on the shop floor, in the office, and between the shop floor workers, the office staff, the on-site managers and the invisible owners who were always elsewhere, issuing dictates about how the work should be done. Each plant had its own particular physical infrastructure that was the accumulation of historical modifications and production decisions, and which affected performance of the employees.
If you don't air condition them, auto plants get scaldingly hot in the summer, especially if they operate 24 hour shifts. In one plant where I worked, from June through the end of August the temperature inside was routinely near 90F (Fahrenheit's how we measured it then/there) during the day time. The temperature was written on a chalk board at the start of each shift, and if it went over 100F during the shift the unionized workers could go home without penalty (non-unionized students like me stayed - if we left, we wouldn't get paid). What often happened is that on days when the temperature at the start of the shift was in the mid-90s, productivity slowed not only beacuse of the physical discomfort, but because some workers were watching the thermometer, taking long breaks and/or going down the road to the pub at lunch to chug a couple beers. But not all the workers. Others were paid on a per-piece basis, and they worked like demons regardless of the temperature. Whenever I was assigned to a piece-work assembly line I would cut short my breaks, not stop to wash my hands before eating lunch, and take other small steps to keep up with the experienced workers. I would come home at the end of a shift sweat-soaked, black with grease, and with a pay stub saying I had earned 4 or 5 times the adult minimum wage. (You won't find those piece-work demons any more, they were later replaced with robots).
Another summer I worked in a different parts plant that was attempting to get its "Q1". Twenty years ago, Ford had this slogan that "Quality was Job 1". It instituted a system whereby all its suppliers had to meet particular standards with respect to quality, and began frequent inspections of their suppliers' facilities to enforce compliance. If a company routinely passed its inspections with flying colours, the Ford people would give the company Q1 certification, and Q1 suppliers had a better chance of getting their supply contracts renewed when the time came. The plant where I was working had its problems. It was not unionized, and wages were considerably lower than a unionized plant. It had a much more ethnically diverse workforce than the CAW plant I described above, and a higher percentage of women on the shop floor. One corner of the plant had furnaces where molten lead was used; three or four women who worked near that end of the plant had had miscarriages in recent years. While there was no "proof" the higher-than-average miscarriage rate was tied to the lead furnaces, it was viewed by workers as highly suspicious. The lead furnaces were operated by recent immigrants to Canada who had no previous skills (it was a hellish job no one else would take) and were brow-beaten by their foremen, and I suspect that may have contributed to the quality problems. The most experienced staff performed easier tasks situated well away from the furnaces.
The management at that plant wasn't interested in anything but getting the Q1. They insisted on having all workers attend weekly production meetings held in the airconditioned offices at the front of the plant. We were lectured by overweight white men in cheap neckties about the importance of quality, then given a pep talk and sent back to the shop floor. The badgering must have worked; the company eventually got its Q1 rating (or perhaps it was because I went back to school in September). About 18 months after that they shipped the machinery to a new plant in Mexico and closed the Canadian plant, laying off all the workers.
These are but two small examples that help me make my simple point: the auto industry, its companies and its employees are as heterogenous as the rest of society. And yet, rarely does a day go by when I read some economist, political scientist or other social scientist giving simplistic diagnoses of what ails the auto sector and prescribing one-size-fits-all remedies. If you've never been inside the walls of an auto plant for any length of time, how do you know what you're talking about?
Don't misundestand me: secondary research of information gained from others is of key importance in any research project, and just because you've been somewhere does not necessarily mean you will understand all there is to know about how things work there. But to my mind there is no substitute for shutting off your computer, putting on your boots and going to take a closehand look at any phenomenon you seek to understand before you describe it to others.