The Mary Tyler Moore Effect

In 1970, every Justice on the Supreme Court of Canada was male. Prime Minister Pierre Elliott Trudeau had an all-male cabinet. Canada had never had a female premier - or even a female television newsreader. Little girls across the country were starting to feel the faint stirrings of ambition, often encouraged by mothers who wanted their daughters to live life to the fullest. But they had no idea - how does one be a "career woman"? What does a professional woman even look like?

Then the Mary Tyler Moore Show came along. It showed a woman who was smart, and independent, and made mistakes - and won out in the end. She was successful. She was likeable. She opened up a world of possibilities.

The Mary Tyler Moore Show first aired in 1970, and went off the air in 1977. This is what happened to US female labour force participation rates during this period.

Now it is impossible to establish any kind of causal relationship between the Mary Tyler Moore show and female labour force participation rates. It's equally if not more plausible that the Mary Tyler Moore Show was commercially successful precisely because women were already entering the workforce - induced to change their behaviour by washing machines and microwaves that lessened the need for household work and increased the desire for cash, by birth control methods that allowed for greater predictability and control over the timing of births, by stagnating male wages and technological change, and so on, and so forth. 

And yet...

This graph compares trends in US and Canadian female labour force participation rates between 1970 and 2012. What's striking about this graph, and the one below based on World Bank data, is that it seems Canadian and the US female labor force participation rates crossed in around the year 2000, when both countries had a female labour force participation rate of around 59%. By 2014, however, Canada's female labour force participation rate was 5 percentage points higher than the US rate (61.4% v. 56.3%). 

This next graph shows changes female labour force participation rates for selected countries and selected years between 1990 and 2014. I downloaded the data from here: http://databank.worldbank.org/data/reports.aspx?source=2&series=SL.TLF.CACT.FE.ZS&country=#

 

Even in a globalized era there are cross-country differences in both the level of, and trend in, female labour force participation. Moreover, it is not obvious that we are seeing continual convergence or progress towards greater gender equality. 

 Female labour force participation? It's complicated. The economics literature has identified many factors that influenced women's work decisions: household technology, birth control, education levels and the country's industrial mix. Yet there is increasing evidence that culture matters  too -- things like attitudes, expectations, ideas about what is appropriate behaviour for men and women. There is even some fairly strong evidence that television has a causal impact on people's attitudes.

Mary Tyler Moore - you made it all seem worthwhile. Rest in Peace.

 

Comments

[rsj]

I would say there is no reason to believe -- not even squinting at the chart for hours -- that this situational comedy had any effect on labor force participation, nor would it be a good thing if it did.

This type of token-feminism is both infantilizing and ahistorical. Things really *don't* happen because there was some fabulous woman that everyone imitated, and we tend to confuse our hero worship for true influence.

Moreover as you get older, you tend to lionize and elevate what were fairly pedestrian events in the past. Fairly ordinary things, through the lens of history, are elevated to earth shaking events via nostalgia and the inevitable compression and simplification of the past. Why not say "I liked Mary Tyler Moore a lot and am sad that she died." You don't need to weave some story around her using FRED data.

[Frank Restly]

Frances,

Probably part of the broader trend:

https://fred.stlouisfed.org/series/SPPOP65UPTOZSUSA

The percentage of U. S. citizens over the age of 65 has jumped from 12.5% to 14.8% from 2008 to 2015. That coincides pretty well with the dip in U. S. female labor force participation on your graph.

That does not fully explain the crossover that began in the year 2000 and I have no idea what Canadian population trends look like.

[Frances Woolley]

rsj -

Fair enough. But look at the obits: "Mary Tyler Moore, who died Wednesday at 80, left behind a legacy on film and television, not least a sitcom character who managed to effect change in real life for millions of working women." (from this). Or "As the spunky Mary Richards on the hit comedy “The Mary Tyler Moore Show” during the 1970s, Moore, who died Wednesday, helped set into motion the professional dreams of many young career women." (from this one).

I wrote this post in part because I wanted to see whether or not i could see an effect in the data. I agree with you, it's pretty hard to spot a change in the trend in the 1970 to 77 period- Female labour force participation rates (FLFP rates) go up sharply both before and after the Mary Tyler Moore show era. So that's a good check on the stories coming from elsewhere.

And to be quite frank with you, I'm also "sub-blogging" my southern neighbours - that recent decline in US FLFP rates is quite striking.

[Frances Woolley]

Frank "That does not fully explain the crossover that began in the year 2000"

Parental leave in Canada was extended from 10 to 35 weeks in 2001. Women who are on leave are counted as in the labour force, which bumps up the numbers quite a bit. Parental + maternity leave in Canada is now about a year. So that makes it easier for women to combine work + kids.

[Frank Restly]

Frances,

As one of your southern neighbors, "We are aware of the situation and are monitoring it closely" - I can't remember what movie that line is from.

As you are probably aware, women tend to live longer than men and so there are likely to be more retired women than men - assuming equal opportunity to work. This becomes especially pronounced when a country goes from a relatively high birth / death rate (post World War II baby boom) to a relatively low birth / death rate (today).

[Frances Woolley]

Frank -

Retired folks are usually excluded from labour force participation rate calculations - see, e.g., this explainer from Statistics Canada:

The participation rate measures the total labour force (comprised of those who are employed and unemployed, combined) relative to the size of the working-age population. In other words, it is the share of the working-age population that is working or looking for work.

Though that doesn't mean it's totally unaffected by demographics, because lots of people in their 50s or early 60s, who are technically "working age" don't participation for health or other reasons.

[rsj]

Frances,

I wouldn't use citations from obituaries as evidence. By definition, every famous person "impacts" millions in some way.

IMO this "role model" view of civil rights is pernicious and really doesn't have any historical basis. The subtext is that the average person just needs to be "inspired" by a famous person who is just focusing on their own personal advancement, and no one needs to dedicate volunteer time or treasure to disrupting existing structures or effecting change. This turns the pursuit of personal success into a type of civil rights virtue.

Similarly the entry of women into the labor force after the 50s was driven by hundreds of thousands of lawyers litigating discrimination cases, politicians passing laws, business men and women demanding change, activists and organizers, protestors, and of course large socio economic forces that often forced women into the labor force. Joan Collins, Elisabeth Taylor, Audrey Hepburn, Katherine Hepburn -- none of these were women's rights leaders. They were famous people not working in women's rights that the real womens' right leaders admired. I admire them too, but we should give credit to the right people as celebrities already have enough.

If we are to have a blog about declines in Labor force participation, on the other hand, then that's a great discussion. I've seen good arguments that this is driven by demographics, but if people could collect finer grained data, it could take the discussion further.

[Frank Restly]

RSJ / Frances,

"...but if people could collect finer grained data, it could take the discussion further."

I would be curious to see if the decline is sector specific, or part of a larger overall trend. While the St. Louis FRED site does break down total employed by industry, it doesn't further subdivide by male / female. A lot was made in the U. S. about the decline of the construction industry (dominated by males) after the housing bust and so I would be curious to know whether finance (down from it's previous peak), government employment (down from it's previous peak), or some other sector is bearing the brunt of the decline.

Cheers

[mpledger]

What's really interesting to me is the downward trend in China. What's going on there? Is it the same downward trend for males? Or is the relaxation of the one child law having an effect?

[Frances Woolley]

mpledger: "What's really interesting to me is the downward trend in China" -

That's basically showing the shift towards a market economy, I would think. 20th century centrally planned economies (China, USSR etc) typically aimed for close to 100% employment for both men and women (in the USSR this was facilitated by having universal publicly provided childcare; I don't know about China).

Though the retirement age for women in China is much lower than in North America (55) so it's also a question of how "working age" is defined in international statistics.

[Frances Woolley]

rsj - "if people could collect more fine-grained data"

This Statistics Canada working paper by Drolet, Uppal, and LaRochelle-Cote breaks down the US/Cda participation gap by age and education: http://www.statcan.gc.ca/pub/75-006-x/2016001/article/14651-eng.htm. They look at the 1976 to 2015 period and consider women aged 25 to 54. The gap in 2015 is wider than that shown above at 7 percentage points. Demographics aren't a big part of the story - in Canada the participation rate for women 45 to 54 is slightly higher than that of younger women, whereas in the US the rates are about the same. Basically Cdn women are more likely to be in the labour force at any level of education, and are more highly educated.

[Majromax]

It's a shame that FRED (and probably the BLS as a whole, to be fair) doesn't have the full age/gender/education breakdown for labour force participation/employment going back to the 70s.

Women had participated in the workforce for some time, but what Mary Tyler Moore portrayed on TV was a truly professional woman, not one with a secretarial job as a holding position prior to marriage and domestic life. Hypothetically, a girl who took Moore as an example would be more likely to seek college/university education and full-time professional work.

The decade after the Mary Tyler Moore Show was associated with improving wages among women full-time workers against stagnant wages for full-time men, and this would be occurring contemporaneously with the workforce entry of girls who watched the show growing up. (Again, FRED doesn't have this data series going back to the show itself. And I'd love to have such an easy-to-use tool for StatsCan data.)

[Chris Auld]

Related: a current Toronto job market candidate has a paper which attempts to estimate the causal effect of the Cosby show on black people's educational attainment:

http://individual.utoronto.ca/cornelson/uploads/5/2/3/7/52377999/cosby.pdf

(Clever, plausibly exogenous?, IV: Cosby show ratings were lower if there happened to be more NBA games on Thursday nights in a given region.)

[Frances Woolley]

Chris - thanks for the link - that is neat. I was wondering if there was some way in which one could find exogenous variation in MTM exposure - did it come on at the same time as Hockey Night in Canada (I've read some stories that talk about it being on Saturday night - I don't remember, I think I watched it mostly on re-runs, and it might have aired at a different time in Canada. If it was on Saturdays, then MTM might have been coinciding with the NHL expansion era, also WHA. Unfortunately the timing might be a bit off - IIRC NHL expansion is late 60s, not early 70s).

[Chris Auld]

Frances, I am a bit too young to have watched MTM when it aired, and I don't, and didn't at the time, watch hockey, so I don't know. Might be a good project for an MA essay or somesuch.

I see the link I posted is corrupt. Those trying to find the paper should search for "media role models cosby cornelson".

[Blaise]

You should do a synthetic control approach (as described by Abadie http://economics.mit.edu/files/11859) to see if MTM might have caused higher female labor participation rate. It would work if there is no other major change that cold affect FLPR at the same time. As a nice placebo test, you could run the synth control on male labor participation (we expect no effect, right?). You could also define your outcome indicator as number of women worker per male worker to purge changes in overall labor participation.