Are Specialized Ants More Productive?

Specialist ants are no better at their jobs than non-specialists:

Are ants that specialize better at their job?, EurekAlert: Adam Smith ... wrote in 1776 that specialized labor provides benefits to human industry, and similar benefits have been suggested to explain the world-wide success of ants, and other social insects which live in colonies. Ants are found on every continent besides Antarctica, and their success has been attributed to the evolution of specialization – it has been theorized that this increases the efficiency of individual workers - but has rarely been measured. However, a new paper by Anna Dornhaus, published in this week's issue of PLoS Biology, shows that individual rock ants (Temnothorax albipennis) specializing on one task are no more efficient than those that perform multiple tasks.

Rock ants are not physically specialized for any particular task. Dornhaus ... measured how often and how readily individual ants performed certain tasks and considered an ant more specialized the more it concentrated its work on one particular task. She expected rock ants that specialized to work more efficiently, but that's not what she found. Dornhaus explains, "It turns out that the ones that are specialized on a particular job are not particularly good at doing that job."...

Although the specialists were not more efficient, they did put in more hours of work. Even though putting in longer hours might seem like the way to success, it wastes colony resources. Dornhaus found that specialists and generalists work equally fast. "Speed does matter because every minute they spend outside is dangerous and energy costly," she said. "They burn fuel, and they risk dying."

It's not known why ants choose the jobs they do, or why some are slow to begin work. She said it might be explained by how quickly an individual detects work to be done, like noticing dirty dishes in the sink. A person with a lower threshold will notice and wash the dishes as soon as there are one or two in the sink. However, a person with a higher threshold doesn't notice the dishes until there are at least 10 piled up. The dishes will still be washed, just not as frequently.

Dornhaus concludes that, at least in this species, a task is not primarily performed by individuals that are especially adapted to it. How much these results apply to the other tasks performed by these ants, or other ant species, or even social insects more generally, remains to be verified. Dorrnhaus's next step is investigating "switching costs," such as the time it takes to walk from one side of the nest to the other or the break in concentration when switching between tasks. Dornhaus suggests specialization might minimize such costs.

Adam Smith cited three benefits from specialization:

1. The worker would become more adept at the task.
2. The time saved from not changing tasks.
3. With specialization, tasks can be isolated and identified, and machinery can be built to do the job in place of labor.

Thought there are plans to look at Smith's second reason for an increase in productivity due to specialization, time saved from staying at one job all day, so far this research only looks at the first reason, dexterity. However, dexterity may not be the most important of the three reasons, and I think that Smith was thinking more about an individual's ability to learn by doing rather than inherent differences across workers. But to the extent that there are differences across workers, and to the extent that it matters for the job, humans may use a different selection mechanism than ants. As to the third reason, I'm not expecting ants to build much capital, so their long-run productivity - the wealth of the ant nation - may be quite limited.

I should note that Smith also thought there was a downside to specialization, i.e. that doing the same task over and over day in and day out dulls the mind.

Comments

[ECONOMISTA NON GRATA]

Adam Smith....? It would serve the author of this article to reference Charles Darwin for matters relating to ants. Ants as in every other species, are genetically predisposed to act and work in accordance with the survival of the species.

I know for a fact that I can wash my car much better than George W. Bush can, however, I pay him to wash my car because I don't have the time. I have other priorities to apply my skills to. The question is, does he do a good job...? The answer is no, but it's worth the fifteen dollars I pay him.

Best regards,

Econolicious

[Bruce Wilder]

Better that he should be washing a car than deciding stuff, or giving a speech, or starting wars, or blogging.

Still, that's comparative advantage -- not increasing returns.

[john Mondragon]

yeah...definitely don't think that ants learn.

[Meh]

There's an issue of how you classify specialisation.

Experiments (mostly in car factories rather than pin factories) suggest that giving a worker a range of tasks and some autonomy creates more efficiency than turning him into "the man who puts the head on the pin" - mainly because people just aren't that good at total repetition.

(Of course, if the job can be turned into a robot job, that's a different matter - but some aren't so easy to do that with.)

On the other hand, I don't think anyone would argue that Mark Thoma and a machine operator from the local factory should do one day econoblog, one day lathe turning to improve efficiency.

The short way of putting it is, specialisation between a bunch of essentially similar tasks (e.g. manual dexterity tasks, or blogging tasks) doesn't seem to get you much, except where it leads to automation down the line.

However, more than ants, humans have developed tasks which are so different that specialisation really is unavoidable.

[Dunc]

In my experience, time saved by not switching tasks is generally wasted in some other frivolity in the interests of maintaining worker sanity.

[ken melvin]

All 'twas discussed and changes made in the sixties and seventies. As someone who's automated too many processes where almost always staffing was reduced by 3/4, quality was improved by orders of magnitude, and productive by 40%, we all understood what was happening. These were mind numbing, boring jobs - monsters of man's creation - bad dreams brought to life - that provided livelihood. Other the blood on the floor, think “Jungle”. After the Great One, people came to believe any job was a good job – certainly better than no job. The jobs fed this monster created by humans, the monster ate the workers. Why man’s fascination with ants. Was such stolen from insects? I see new cars obviously designed by pubescent boys and wonder if it is they and theirs who design and build our factories? At the time, ‘twas exciting solving these puzzle, we oft celebrated our success, but I did notice that outsiders, e.g., my children, saw it all as very strange and somewhat comical.

[Robinia]

A balance of challenging tasks in one's life-- both physical and mental challenges, a variety of problems to solve over time-- creates health, inspiration and productivity in the individual worker. I suspect for ants as well as humans; "specialists" are those who have lost the wonder of tackling new challenges and simply grind on for sustenance, as per ken melvin's Fordist creatures out of the bad dreams made visible in the movie "Modern Times."

Any calculation that shows the US's specialization in white collar/services work as highly productive is probably not factoring in the costs to health and happiness of the epidemic of obesity/diabetes/heart disease, and the disconnection among the populace from the satisfaction and instant gratification brought by physical achievements.

[winstongator]

I think this is one of Smith's strongest contributions. I look at specialization in engineering and medicine. There are tasks where the difference between the best worker and someone just starting are pretty small - most landscaping, cleaning - and other tasks where experts can be almost an order of magnitude more productive. The bigger the productivity function of dexterity, the more important specialization: P(dex) = 1*effort + 0.1*dexterity vs. 1*effort + 100*dexterity. I guess you could calculate the effort required to gain an amount of dexterity and solve for whether that is an efficient use of effort.

[reason]

Isn't there are lack of awareness of the concept of "comparitive advantage" shown in this article. He can't measure comparitive advantage by measuring absolute advantage. It could be the specialists are weaker at everything and choose to specialise to minimise their disadvantage.

[ccm]

Your description of Smith sent me back to the original. I've never understood how people can read the first chapter of WoN and conclude that Smith promotes specialization because of dexterity and time saving. His third point always struck me as most important -- and your desciption of it completely misses the point. The passive construct you use fails utterly to capture Smith's concept -- the whole idea is that the worker becomes an agent of creative change.

Anybody with the mind-numbing job of repeating the same task will think about better ways to do it. It is because specialization promotes micro-invention, that it is such an important source of growth. Here's a link to the original.

[Bruce Wilder]

I think an individual ant has about 250,000 brain cells -- not much of a neural net, really. There may be some learning going on, but I suspect reason's insight would be more useful to detecting it: some of the adaptability of the ant colony is built into variation among individual ants.

The gains from "specialization" are about the gains from cybernetic control of the production process, gains, which are -- imho -- best conceived of, as a reduction in error and waste. A worker learning a skill, the design of machinery, the application of scientific theory to the invention of new techniques, entrepreneurial innovation in organization -- it is all part of the same opportunity to better control production processes, and entails sunk cost investments and technology-embodying dedicated capital investment. It is an absolute, not comparative advantage, and technical, not allocative efficiency.

While manual dexterity plays a role, I suspect that's ultimately trivial. The essential human roles in production are as goal-setters and drivers: choosing the destination, and driving the vehicle, with driving-the-vehicle as increasingly a matter of meta-monitoring of automated processes, if not optional, in an age of increasingly sophisticated computers.

In human social evolution, the struggle is always over whether there will be social specialization that separates goal-setting from driver. Will we have an aristocracy of deciders, supported by an increasingly redundant mass of chauffeurs? Will we be a society of a small number of self-appointed heroic mountain climbers, supported by a larger number of easily forgotten sherpas? The choice will be made by a middle cadre of geeks, designers and professionals, who make the meta-choices, in inventing and designing the components of our evolving systems of control.

Anthill Earth.

[OhNoNotAgain]

"I should note that Smith also thought there was a downside to specialization, i.e. that doing the same task over and over day in and day out dulls the mind."

And the body. Anyone that I've ever known that has done the same repetitive labor for more than a decade has severe nerve and muscular issues from that labor once they hit their 40's. There's an social externality there that may not be accounted for completely. I know that OSHA has some rules in place in an attempt to reduce repetitive stress injuries, but I'm not sure how much they are enforced and whether the companies actually take any of the information to heart, especially if they are off-loading the health care costs on to the general public.

[Meh]

Bruce - the decision has already been made - the econogeeks have spoken long since now - we will be a society of the superrich few and the redundant mass of chauffeurs. All our economic theories point to it, we have no conception of how else to do it. All our socio-political responses are band-aids - they work, but they don't have that quasi-mystical "rationale" and so we keep thinking we can do without them and stick with our religiously inspired notions of incentives and equilibria.

[Patricia Shannon]

With so much automation, why do people have to work 40 hours a week?

[Alex Tolley]

I think trying to equate ant specialization with humans is a bad analogy.

As to ant specialization, colonies do have specialization built in by genetics. The queen produces eggs, the males fertilize her, the soldiers fight intruders and the workers collect the food, tend the farms etc. Would anyone suggest that this type of specialization is not beneficial and that soldier should help with the work - no.

Workers don't change jobs in the middle of another, e.g. changing their minds and retracing their steps to help out the farmers before reaching the food source. So it is not clear that the switching that is going on is other than marginal in the colony's economics.

[ken melvin]

Patricia Shannon says...

With so much automation, why do people have to work 40 hours a week?

I suspect a fear that society might collapse due previous conditioning.

[Bruce Wilder]

PS: "With so much automation, why do people have to work 40 hours a week?"

That much stuff requires selling overtime.

[calmo]

So many glancing blows...remember calmo the last time you counted, 'twas way less than 250,000...neva mind the borderline cases...who could count that high, people?
Ok, maybe hari...
ok, Barkley maybe, if anne took a vacation for a few days...

OMG, it was Bruce! Ok, aside from the very specialized Bruce Wilder, people, who could count that high?

Ok, I think the (somewhat less productive, and coincidently somewhat less specialized) pack of hackers here don't want to talk about the army ants, nor their restricted, constricted, and philosophically stunted specialization, nor their productivity (howeva arbitrarily that might be defined).
[I know. Hardly 250,000 brain cells at work here...and failing in their valiant effort to redirect this thread.]

Ok, due to Brain Cell Tank bleeping 'empty', I need to bail on ccm's note:Anybody with the mind-numbing job of repeating the same task will think about better ways to do it. And just specially know then, that we are fundamentally different from ants who got stuck at the ant level...or we got stuck at the specialized human level and fail to see just how much better the ants are doing it.