- Nation-Building, Nationalism and Wars [pdf] | Would A Libertarian Military Be More Lethal?
- Formal or Informal, Legal or Illegal: The Ambiguous Nature of Cross-border Livestock Trade in the Horn of Africa [pdf] | A note on the police or – “Why I don’t trust the police.”
- Political Cleavages and Changing Exposure to Trade [pdf] | Slowly debunking the trade leads to peace fallacy
- Realism, Liberalism and the Iraq War [pdf] | When Should Intellectuals be held Accountable for Popular Misrepresentations of their Theories?
- High-Value Work and the Rise of Women: The Cotton Revolution and Gender Equality in China [pdf] | When (Where and Why) Women Were More Literate than Men
As an economic historian, I’ve always had a hard time with the idea of secular stagnation. After all, one decade of slow growth is merely a blip on the twelve millenniums of economic history (I am not that interested with the pre-Neolithic history, but there is some great work to be found in archaeology journals). Hence, Robert Gordon’s arguments fall short on me.
That was until I was sparked to react to a comment by Emily Skarbek at Econlib. Overall, she is skeptical of Gordon’s claims of secular stagnation. But not for the same reasons. She claims that there are many improvements in welfare that we are not capturing through national income accounts. This is basically the same point as the one made by the great Joel Mokyr (the gold standard of economic historians).
It is true that national accounts have some large conceptual problems regarding measuring output when there are massive technological changes. Yet, all these problems don’t go in the same direction. More precisely, they don’t all lead to underestimation of growth.
My favorite example of one that leads us to overestimate growth is the one I keep giving my macroeconomics students at HEC Montreal. Assume an economy with a labor-force participation rate of 50%. Basically, only males work. All women stay at home for household chores and childcare. In that case, all measured output is male-produced output. Since national accounts don’t consider household production, all the output of women in the households of this scenario is non-existent.
Now assume a technological change causing a shift of 10% of women to the workforce at the same wage rate as men. That boosts labor participation rate to 55% and output by 5%. However, that would largely overestimate growth caused by this shift. After all, when my grandmothers were raising my parents, they were producing something. It was not worthless output. Obviously, if my grandmothers went to work, there was some net added value, but not as much as 5%. However, according to national account, the net increase in GDP is … 5%.
Obviously wrong right? Now, think of the economic history of the last 100 years. Progressively, female labor-force participation increased as marriages were delayed and family sizes were reduced. Unmarried women stayed on the market longer. Then, the introduction of new household technologies allowed some married women to join the labor force more actively. Progressively, women accumulated more human capital and became more active in the labor force. So much that in many western countries, both genders have equal labor-force participation rates.
As they shifted from household production to market production, we considered that everything they did was a net added value. We never subtracted the value of what was produced before. Don’t get me wrong, I am happy that women work instead of toiling inside a household to handwash dirty clothes. Yet, it would be both statistically incorrect and morally insulting to say that what women did in the household had no value whatsoever.
The role of household production in reducing the quality of growth estimates goes back to the 1870s! A 1996 article in Feminist Economics (which I use a lot in my own national account sections of macroeconomics classes) shows the following changes in growth rates when we account for the value of household production. Instead of increasing to 1910 and then falling to 1930, growth in the United States falls to 1930. While the growth rates remain appreciable, they nonetheless indicate a massively different interpretation of American economic history.
Sadly, I do not possess a continuation of such estimates to later points in time for the United States. I know there is an article by the brilliant Valerie Ramey in the Journal of Economic History, but I am not sure how to compute this to reflect changes in overall output. I intend to try to find them for a short piece I want to submit later in 2016. Yet, I do have estimates for my home country of Canada. Combining a 1979 paper in the Review of Income and Wealth with a working paper from Statistics Canada, it seems that the value of household production falls from 45% of GNP in 1961 to 33% in 1998. When we adjust GDP per capita to consider the changes in household work in Canada, the growth path remains positive, but it is less impressive.
I am not saying that Gordon is right to say that growth is over. I am saying that the accounting problems don’t all go in the direction of invalidating him. In fact, if my point is correct, proper corrections would reduce growth rates dramatically for the period of 1945 to 1975 and less so for the period that followed. This may indicate that “slow growth” was with us for most of the post-war era. That’s why I reacted to the blog post of Skarbek.
It also allows me to say the thing that is the best buzz-kill for economics students: national accounting matters!
That is the title of this piece in the Left-wing British zine spiked online by Joanna Williams, a lecturer in higher education at the University of Kent. Here is the money shot:
A gender pay gap, albeit one that is rapidly decreasing, still exists; but the good news is that when occupation, contracted hours and most significantly age are taken into account, it all but disappears. In fact, the youngest women today, even those working part-time, are already earning more each hour than men. We need to ask why this is not more widely known and question the motives of those who seem so desperate to cling to a last-ditch attempt to prove that women remain disadvantaged. We should be telling today’s girls that the potential to do whatever job they want and earn as much money as they please is theirs for the taking, rather than burdening them with the mantle of victimhood.
The emphasis is mine. I know Jacques has dealt with the pay gap canard many times on this blog before (“Yes, women earn less than men but it’s not a case of unequal pay for equal work. It’s a case of unequal pay for unequal work.“), but it is still worth asking why politicians and so-called feminists are still beating such an obviously dead horse.
Politicians, especially anti-market ones, can use the pay gap to gain votes and hurt their rivals. This is an easy one.
Feminists are a horse of a different color, though, largely because there are so many variants of feminism out there (I am feminist in the sense that I think women are people, just like the old bumper sticker says!). Again, some of the peddling of this myth in feminist quarters is due to Left-wing animosity against markets, and some of it is just women in their thirties trying to remember what it was like to be in college.
Another reason might simply be economic. If an individual can get away with playing the victim in a business setting, why would she not do so? That is to say, if the rules are set to reward “playing the victim,” or if the rules were made several decades ago in order to combat an injustice (whether real or perceived), the most logical thing to do would be to play along with such rules.
The pay gap is therefore a political problem, not an economic one, and political solutions tend to be ones gained from obfuscating or ignoring outright the relevant facts of the matter.
The political undertones of the pay gap are exemplified by this 1995 paper (h/t Dr A) by two academic sociologists whose empirical work justifies Dr Delacroix’s and Dr Williams’s arguments (“it’s not a case of unequal pay for equal work”). In the conclusion of the paper, though, the sociologists go on to suggest that more legislation is needed to account for the overall pay gap. Why? Because men tend to find work in fields that pay more than women, and men don’t have vaginas with which to push out babies. In the minds of the sociologists, then, the best thing to do to ameliorate a non-existent problem (the pay gap that does not account for occupation, age, or hours worked) is to pass legislation that will somehow create more female engineers out of thin air (hello double standards, or hello decline in quality education).
Editor’s note: this old essay by Dr Delacroix might be worth reading in tandem with this article. It’s titled “Sasquatch and Liberal Academe.”
I agreed in Part One of this essay that there may be a small average pay difference of five percentage points between employed American women and men. It’s possible that even after you take into account all the facts mentioned before, lower education, less seniority, lesser presence in well paid industries, women, on the average, earn 95 cents on men’s one dollar. (That’s also on the average.) I agreed that this may be evidence of discrimination against women in the work place. But is it the obvious explanation? Is it the only explanation? Is it even credible at all?
There is a reason this is an essay for women only. I want to help you evoke forbidden topics, topics never or rarely approached in the social sciences or in the newspapers. It’s time to muster everything you know about your sisters, ladies; we are going into taboo territory!
I am not a woman myself but I have had occasion in my 30-year teaching career to observe hundreds of women. I have also observed women in various workplaces in two countries. What follows is not (NOT) supported by any study.* It’s a thinking exercise about what I suspect you know. See if it jibes.
Female students obtain better grades than male students in college. However, they don’t get to good grades the same way. Males take more risks, on the average. In my observation, for example, they are many times more likely to wing it than their female fellow students. In general (on the average), males are more forceful, much more likely than women to demand their due (even when nothing is due to them at all).
If you don’t ask for a raise, often, you don’t get one. If you are timid about organizing, you will earn less than those who are well represented by unions. Is it possible that women, on the average…. (Finish the sentence) and that this would account for the remaining possible five percentage pay differential between women workers and their male counterparts. Is it possible?
I have personally seen a few cases of discrimination against women workers. They consisted in creating work conditions that would make it difficult for the targeted employee to produce a work performance that would make raises likely. (Poisoning the water.) As I said, I have only seen a few cases perhaps fewer than ten. Every single time, the discriminator was a female supervisor. Perhaps, my limited experience is atypical. Or is it? What do you think? Do women in positions of power ever persecute other women? Often? Seldom? Never?
Now, I want to talk briefly about productivity in a manner that is especially proscribed in polite society. I think I have observed hundreds of time that women in a group with an assigned task spend large amounts of time in activities that seem unrelated to the task. Women tend to socialize at length where it’s not called for. The socialization often includes plotting against one another and excluding some. Perhaps the socializing is a kind of disarmament conference without which work cannot even proceed. Again, those are a subjective observations possibly based on women who are somehow different, not representative, different from the average. What do you think?
Here is what I observed when teaching, specifically. Most of what follows is based on teaching the same, clearly elite class every winter for 22 years. The classes were small and they involved important group projects. The projects were done mostly in class. The sample these 22 classes give us is obviously biased, unrepresentative. It’s biased because it was composed of students markedly superior to the run-of-the-mill student. The women in that class were probably more productive than most female undergraduates (and than most male undergraduates as well).
Here is what I saw: Insofar as it’s observable, groups of males engaged in a collective task, achieved the same results as women with considerably less time expenditure than women. It’s as if a straight A cost guys three hours, and women nine hours. (Those numbers are subjective, of course. I am holding myself back; I wouldn’t be surprised if the gap in time investment were wider.)
Is it possible that my observation is not in some way nullified by bias? Is it possible that what I think I have observed in connection with the time cost of doing something is somehow related to the earlier observation that women tend to socialize when doing so does not appear to contribute to the accomplishment of the task ? You decide.
My informal observations are surely not (NOT) equivalent to a formal, rigorous study. This does not mean that I should keep quiet about them. If they don’t ring a bell with your experience, ignore them. If they do, maybe it’s worth thinking about alternative explanations to the widespread belief in the general existence of uncalled for, arbitrary discrimination against female workers. Keep in mind that, at this point, we are talking about a five percentage point differential; we are considering 95 cents on the dollar, not the president’s 77 cents.
Now, let me switch angles of vision a little bit. The actual, residual difference between women’s and men’s remunerations after you have accounted for the obvious factor, (Part One of this essay.) is so imprecise that it leaves open the possibility that women earn more than men in the same jobs.** Again, I am frankly wading into the subjective and I am inviting you to wade in with me.
During thirty years of teaching, there were about ten times when I fielded female students’ complaints that they were discriminated against in some fashion just for being female. That was always in my role as an adviser rather than as a classroom teacher.
Every time, I would ask the complaining student if it were not possible that she was angry because her male counterparts seemed to be getting grades they did not deserve according to her judgment. I will let you guess what answers I received to this particular query.
Then, I would look the student straight in the eyes and I would propose the following hypothetical:
Suppose you are completely right, suppose the instructor discriminated against you simply for being female, let me ask you: Has it ever happened in your whole life that you got an undeserved pass just because you were a girl?
I did not have the presence of mind to keep exact tabs, of course, but I can’t remember a single time when the conversation did not dissolve into a smile!
Adding subjectivity to subjectivity, doing it consciously, I would guess that male instructors by and large prefer female students. They are less likely than males to be grossly disruptive; they are nicer; they are more polite; they smell better. (It matters in a room of forty in June.) Female students also often flirt with the teacher in unconsciously charming ways. (Disclosure: I am married to a former student.)
Would these behaviors tend to cause male instructors to treat female students negatively? Women instructors?
Maybe this is all my own private self-delusion. Maybe this preference for women employees does not exist in the workplace, as a far as male supervisors and male decision makers are concerned. Or maybe, I am not that deluded, or maybe males in positions of power have a built-in preference for female subordinates as much as they do for female students. How surprising would that be?
Take my case, for example: My mother was a woman, my daughter is a woman, my granddaughter is female, my sisters are female. They all love me, without exception. I addition, I am married to a woman. She cooks for me, very well, almost every day; she does my laundry (even when I don’t want her to); I talk with her more than I do with anyone else on earth. When times are tough, she is always, every single time, in my corner. I don’t want to get X-rated but it’s also true that 100% of my sexual satisfaction, in my whole life, was somehow connected to women. (Like many California men, I also have an ex-wife somewhere but she does not cause me any grief. I am lucky; she just moved on instead of exacting a just revenge.)
Is my experience different from that of most or all men? Did they have male mothers? Were their sisters guys? Are they all married to other men who love them and spoil them? Is a very high percentage homoerotic? You see my point.
Is it possible also that a lifetime bath of estrogen predisposed me to a positive bias toward women or, is it more likely that all these good treatment left me prejudiced to the point where….
…to the point where I would take care- with someone else’s money, most of the time – to pay female employees only 95 cents when I pay their men colleagues a full dollar for the same work?
Or, alternatively, is it possible that my unchecked, unthoughtful, mindless tendency is to treat them better and to pay them more, say, $1.05 to males’ $1.00 ?
When you take a sketch and you take the trouble to draw in all the details, often suddenly, you come to see that you misunderstood or mis-perceived the meaning of the original sketch.
Three big questions to finish.
If the good research (that I did not do) confirmed my speculation that women may earn on the average one dollar and five cents ($1.05) to men’s one dollar would there be big demonstrations to protest the disparity? Would anyone have the nerve to initiate legislation to close the pay gap by force? What do you believe?
Discrimination against women in any way, shape or form has been illegal in this country since 1964. Is there any one who thinks that if legislation has been inoperative for fifty years, unable to correct (alleged) pay discrimination against women, more legislation under a weakened president, imposed on a deeply divided nation will now solve the problem? This is a real question. Please, answer it in your own mind.
And if it were possible to pay women less for equal work, with the same quality of workplace attention, the same performance, the same results, isn’t it true that a rational hiring policy would require discrimination against men? If this were true, all employers would try to limit labor costs by hiring only women until there were no more women available for hiring in the relevant pool. Or is there some magic male solidarity that trumps everything, including business competition and the search for profits?
By the way, the last time I looked, in the USA, women had most of the wealth and most of the votes. (If it’s not true anymore, I don’t worry; someone will surely correct me.) Why don’t women, on the average, use their votes and their financial resources to erase the pay differential, to turn the 77 cents into at least one dollar?
Here is my stake in all this. First, I detest government policy based on lies and on deviousness. Second, my daughter is rearing her own daughter alone. The last thing I want is that my adored granddaughter should be cheated because her mother is being short-changed on her paycheck merely for being a woman. Are other men different? Am I a brilliant and inimitable exception?
Some stories have been told so long and so often that people don’t think of subjecting them to even the slightest of reality checks. The 77 cents on the dollar story does not stand up to scrutiny. It’s a purely political story designed to keep alive an artificial sense of grievance in an important segment of the electorate.
The 77 cents on the dollar story exploits women. It’s insulting.
A good tale to finish, the second funny thing that any feminist ever said:
Ginger Roger was Fred Astaire’s dancing performer for twenty or thirty years. The couple dominated movie screens in the forties and fifties and beyond. Fred was more famous than his partner, Ginger. At one point, irritated by the fuss over her partner, Ginger is said to have declared,
“I did everything Fred did backward and in high heels!”
* Such studies are practically proscribed in academia. If they were not, their findings would be boycotted by major journals. Political correctness reigns where it shouldn’t exist at all. Nevertheless, there are a few brave academic researchers who venture where no one is supposed to tread. Most are women.
** In the early 2000s, black women earned more, on average than black men. The case I pose to you is thus not completely absurd or imaginary. I don’t know if this disparity continued after the crisis beginning in 2008. It does not matter. I just wanted to impress on you that women sometimes earn more than men.
American women who work for wages or a salary, on the average, earn 77 cents when American men earn one (1) dollar, also on the average.
You have to be careful of averages. They are not naturally vicious but they are often used to deceive. That is, people routinely overestimate themselves and don’t slow down enough to understand what they are seeing and hearing when an average is mentioned.
Here is a little practice exercise: Suppose all women who lack education beyond high school quit work completely. (They might go on welfare or they might find hard working husbands, maybe currently illegal immigrants – Not a bad idea actually, if I say so myself!) If this happened, what would become to the 77 cents on the dollar?
(The answer is several paragraphs below.)
Consider also that “on the average” means, of course, that there are many women who earn more money than many men, women in government, for example. Take the female toll-takers at the Golden Gate Bridge, for example. They earn about $100,000 a year for very low-skill work. They thus earn much more than male cable-television technicians who do things most of us don’t even know enough to think about. (There are female cable technicians of course, but that’s not the point, right now.) That’s compatible with the 77 cents on the dollar figure.
That women earn on the average 77 cents when men earn a full dollar speaks of rampant and rank discrimination against women where it matters most, the workplace. Or does it?
Below are some relevant facts that all of President Obama’s economic employers know. I mean that the facts are so well-known that it’s inconceivable that they don’t know them.
- Fact: On the average, working women have less education and less seniority than men. That’s on the average.
As it happens levels of pay, in many industries depend largely on seniority (rightly or wrongly). Access to the best paid jobs in a given industry also depends much on level of education. Access to superior and well-paid jobs also often depends on achieving seniority. That’s a double-whammy on low education!
Answer to the question near the beginning of this essay: If women who had no college dropped out of the workforce, female workers would, on the average, suddenly have higher educational achievement. Then, the average pay of women nation-wide would go up. If all the women with no college education dropped out of the workforce, the 77 cents on the dollar would immediately disappear. I don’t know what the resulting number would be; it might be 80 cents on the dollar, or 90 cents on the dollar. What is certain is that it would be a higher, better number.
Repeat: If all the low-skill jobs requiring a modest level of education disappeared all of a sudden, if all the women holding such jobs lost their jobs, the average pay of women, including as compared to me, would immediately go up.
This is not some sort of foggy speculation, it’s an arithmetic certainty.
Similarly, if more women in the workforce had high seniority, the average pay of women nation-wide would also be higher than 77 cents on men’s dollar. Here too, it’s a mathematical certainty although I don’t know by how much the figure would change. This is all by way of remembering what averages mean.
- Fact: Working women concentrate in economic sectors where wages are historically low.
That’s low wages for both men and women. Sometimes, there are no understandable reason why pay is low in such sectors. Often it’s a sort of historical accident connected with an early union activity in those sectors of the economy. Sometimes there are good direct reasons for the high pay in sectors where women are rare. Blue collar work on oil platforms and commercial fishing are both examples of activities where few women are found. They are also dangerous activities. They are also physically strenuous activities. In those two particular sectors, pay is much higher than it is say, in the health industries, or in retail where many women areemloyed. This means that both men and women employed in fishing and on oil, platforms earn more money than either men or women in many other industries.
The average lower pay of women nationally is at least in part the result of their low participation in these highly paid industries. If there were equal numbers of women in those high-pay sectors as there are men, the national average pay of women would be higher than 77 cents on men’s dollar.
- Fact: Among those who work forty hours a week or more (“full time”), men work much longer hours than women on the average.
It’s often the case that, other things being equal, those who work longer hours earn more money than those who work shorter hours. They earn more for the total number of hours they work. (They may also be promoted faster but that’s not my point here; one thing at a time.) Incidentally, this is true both for base workers, such as assembly line workers and sales associates, and for so-called “exempt personnel,” personnel in supervisory and management positions. The mechanisms are different, union rules, formal pay scales and government-mandated requirements (think overtime pay), in one case, alleged “merit pay,” on the other. The results are similar: Work more; earn more.
Women earn less money than men on the average than men because they spend less time at work than men do.
Now, close your eyes and let me describe two imaginary workers. One has 25 years of seniority and three years of post-high school education. The same worker is employed in mining. Over the course of a year, this worker puts in 46 hours a week on average.
The second worker has one year of junior college and has been on the same job for eight years. That worker’s occupation is in one of the health industries. Calculated over one year, this second worker puts in 40 hours plus twenty minutes a week on average.
Now, keep your eyes closed and forbid yourself from stereotyping. You don’t know the sex of either imaginary worker. Keep in mind that they may well be of the same sex, for example (for example). One or the other, or something else…
Which of the two fully employed workers do you think earns the most money in one year in actuality?
Which do you think should earn the most money according to your own standard of fairness?
You get my drift?
It turns out that when studies compensate for these important factors, American women’s remunerations are about the same as men’s. That’s still on the average. I wouldn’t be too surprised if you could find a female fisherman with 25 years seniority and a doctorate who earns less money than her husband, a high school dropout who works in a candle shop. The relevant numbers are simply too small to affect comparisons of national averages.
Yes, women earn less than men but it’s not a case of unequal pay for equal work. It’s a case of unequal pay for unequal work.
It’s worth asking why women would heap upon themselves so many of the factors that result in comparatively low pay? I mean low education, low seniority, and working in less generously paying sectors.
You probably have your own hypotheses (plural) about why this is. Let me help with an additional fact:
- Fact: Women who are not married, have never been married, and have no children earn as much as men. Are you really surprised?
Many other studies confirm what we all already know: Women are the primary caretakers of both home and children by a long, long shot.
The care of children interferes vigorously with women’s ability to reach for higher paid jobs, and with their attention to their paid work, and to their ability to work long hours. It’s that simple.
Women workers fail to accumulate seniority because they quit working earlier and more frequently than men. They tend to move in and out of the workforce; that’s inimical to the accumulation of seniority, of course.
Women workers have less education than men workers, on the average, for slightly (only slightly) more complex reasons. At the lower end of the pay-scale women who work outside the home are not equivalent to men workers in general. For one thing, many low-paid working women, and increasing numbers of them, are single women raising their children alone. But we know that women with lower educational status are more likely to find themselves in that situation than women with more formal education.
Married women with children have on the average, more education than single women with children. Such married women are less likely to be in the workforce at all . Instead, their husbands are. Their husbands’ higher education and seniority enter into the national statistics. Their non-working wives’ also high numbers don’t because they are not in the labor force, precisely.
If all married women joined the labor force, the gap in education between employed men and employed women on average would shrink. It might even vanish altogether.
That would raise women’s average pay nation-wide, although the fate of poor ly educated, low seniority, women employed in badly paying sectors would not improve one bit.
If all married women joined the labor force and stayed in it, employed women’s seniority would equal men’s after a while. That would raise women’s average pay nation-wide.
The pay of women with low seniority would….
(Complete the sentence; this is a test!)
Those who claim the 77 cents on the dollar figure are comparing apples and oranges.
Those in government who do this know the facts. Why are they doing it?
Now, once you have taken account all facts above, the things we already know about different ways in which women and men deal with work, women on the average still earn a little less than men. The difference is much smaller than the difference between 77 and 100 (77 cents and one dollar). Nevertheless, as I write, I think it’s possible to argue that this small difference – maybe something like 5 percentage points – proves some degree of pay discrimination against women.
By the way, I don’t play down at all this kind of pay differential. If you gross $30,000/year, 5% more would be $1,500. Even with standard deductions, that’s a round-trip ticket to someplace, even someplace interesting.
In Part Two of this essay, I will leave the domain of what’s well know, of what the president ought to know, and I will take you with me on a trip of honest, frank speculations about women’s work.
Don’t forget to come back. The best portion is yet to come!
Tech. note: Anyone is welcome to challenge any of the assertions above. Here are the rules I play by: I you give me a general reading assignment, I won’t do it. It’s too easy to waste someone’s time on a wild goose chase. If you don’t bother to say, “Read this because it shows ‘this assertion of yours…’ to be false ,” don’t expect me to make the effort either. Also, evidence that does not come from a respected refereed journal is unlikely to make much of an impression on me.