In the late 19th century, in the wake of the strikes of 1877 and the Haymarket riot, armories began to spring up in major cities in the US. They housed "National Guard units thought to be more reliable than local police when upholding urban order might involve firing at strikers. New York had twenty of them, Philadelphia six." Ah--but those were the bad old days, right? This New Gilded Age is of a kinder, gentler sort, isn't it?
The latest issue of Economists' Voice has an interesting article by Samuel Bowles and Arjun Jayadev titled "Garrison America." (It's available free for non-subscribers once a registration form is filled in for 'guest access.') Bowles and Jayadev attempt to measure the percent of the American workforce devoted to protecting stuff rather than making stuff. This "guard labor" force, consisting of military personnel, police and security guards, and a more amorphous category of work supervisors, has grown four-fold in the US since 1890, when their data series begins. Comparing US guard labor with other countries, Bowles and Jayadev find that guard labor fractions are strongly correlated with economic polarization and political conflict, and negatively correlated with measures of political legitimacy and social welfare spending.
They further find that recent international crises are not to blame for the large percentage of the American workforce employed in "the disciplinary apparatus of society." Indeed, "the international security contribution to the guard labor force is down sharply, and "Police now considerably outnumber those working directly or indirectly for the Pentagon--a first since our data series begins in 1890." In fact, "By 2012, the Department of Labor predicts, the United States will have more private security guards than high school teachers."
An obvious objection to the paper's conclusions is that the authors have included way too many people in the 'supervisor' category. I hope to post an update later looking at this question in more detail, but let me here summarize how Boyles and Jayadev deal with this problem in this paper. They used the Dictionary of Occupational Titles to create a list of jobs that consist primarily of supervising people (rather than supervising data, or machines). The estimates supervisory personnel derived from this method are comparable to (but less than) estimates from available surveys of the supervisory tasks of workers in the US and other nations. The authors do not count logocrats, like economists, for example, nor those "involved in the production of weapons for self protection, of locks and security cameras, and surveillance devices such as trackers attached to long haul trucks..." They conclude that "We do not think we have overestimated the number of supervisors; we have certainly missed some types of work that could be termed guard labor."
Nor, in my opinion, do the authors over-emphasize the connection between economic inequality and guard labor, despite the strong correlation between the two in their international and US state-level comparisons. They caution that the high US guard labor percentage may be at least partly a product of advanced specialization in the US workforce:
If such measures of national wealth were developed, it would help countries reallocate resources from guarding stuff to making it, and might boost economic development and "enhance the livelihoods of the least well off."
Bowles and Jayadev quote John Stuart Mill lamenting "how a great proportion of all efforts and talents in the world are employed in merely neutralizing one another. It is the proper end of government to reduce this wretched waste to the smallest possible amount, by taking such measures as shall cause the energies now spent by mankind in injuring one another, or in protecting themselves against injury, to be turned to the legitimate employment of the human faculties...."
The latest issue of Economists' Voice has an interesting article by Samuel Bowles and Arjun Jayadev titled "Garrison America." (It's available free for non-subscribers once a registration form is filled in for 'guest access.') Bowles and Jayadev attempt to measure the percent of the American workforce devoted to protecting stuff rather than making stuff. This "guard labor" force, consisting of military personnel, police and security guards, and a more amorphous category of work supervisors, has grown four-fold in the US since 1890, when their data series begins. Comparing US guard labor with other countries, Bowles and Jayadev find that guard labor fractions are strongly correlated with economic polarization and political conflict, and negatively correlated with measures of political legitimacy and social welfare spending.
They further find that recent international crises are not to blame for the large percentage of the American workforce employed in "the disciplinary apparatus of society." Indeed, "the international security contribution to the guard labor force is down sharply, and "Police now considerably outnumber those working directly or indirectly for the Pentagon--a first since our data series begins in 1890." In fact, "By 2012, the Department of Labor predicts, the United States will have more private security guards than high school teachers."
An obvious objection to the paper's conclusions is that the authors have included way too many people in the 'supervisor' category. I hope to post an update later looking at this question in more detail, but let me here summarize how Boyles and Jayadev deal with this problem in this paper. They used the Dictionary of Occupational Titles to create a list of jobs that consist primarily of supervising people (rather than supervising data, or machines). The estimates supervisory personnel derived from this method are comparable to (but less than) estimates from available surveys of the supervisory tasks of workers in the US and other nations. The authors do not count logocrats, like economists, for example, nor those "involved in the production of weapons for self protection, of locks and security cameras, and surveillance devices such as trackers attached to long haul trucks..." They conclude that "We do not think we have overestimated the number of supervisors; we have certainly missed some types of work that could be termed guard labor."
Nor, in my opinion, do the authors over-emphasize the connection between economic inequality and guard labor, despite the strong correlation between the two in their international and US state-level comparisons. They caution that the high US guard labor percentage may be at least partly a product of advanced specialization in the US workforce:
In many countries, the job of getting people to abide by the rules is not left up to the specialists that we have included in guard labor. Anyone who has tried jaywalking in Germany will known what we mean: it's not the police who you have to worry about, but your (equally formidable) fellow pedestrians.So, if one accepts the conclusions of Bowles and Jayadev as legitimate, what implications do they have for public policy? The authors call for new measures of national wealth that would take into account the investments made in guard labor. The role of that section of the workforce, they argue, is not to produce new goods, but rather to "maintain the economically relevant institutional stock."
If such measures of national wealth were developed, it would help countries reallocate resources from guarding stuff to making it, and might boost economic development and "enhance the livelihoods of the least well off."
Bowles and Jayadev quote John Stuart Mill lamenting "how a great proportion of all efforts and talents in the world are employed in merely neutralizing one another. It is the proper end of government to reduce this wretched waste to the smallest possible amount, by taking such measures as shall cause the energies now spent by mankind in injuring one another, or in protecting themselves against injury, to be turned to the legitimate employment of the human faculties...."