Monday, February 22, 2016

Hillary 1993: Largest Drop in Girl Names EVER; Chelsea Distant Second

Recently, I wrote an little post that got a lot of attention and some criticism, As First Lady, Popularity of Babies Named "Hillary" Dropped by an Unprecedented 90%.

The attention was likely due to the large number of people who are attempting to evaluate Hillary Clinton as a viable general election candidate. These people might rightly or wrongly assume that measures of popularity from her as First Lady are at least informative in predicting how she will do in the general election.

The criticism were largely based on my lack of scientific methodology and largely completely supported. The article was more meant as a statistical note rather than as a serious discussion. Yes, among all presidents since Nixon, the popularity of First Lady names have dropped over the term as first lady but none so much as the name Hillary.

However, some good potential alternative explanations are possible:
1. Are such drops typical of female names that peak in popularity in general?
2. Could Hillary Clinton have briefly lead to the popularizing of the name "Hillary" which later fell off after she became first lady?

So, I decided to come back to the data in an attempt to understand if the dramatic drop in the popularity of the name Hillary was just bad luck or likely related to her stint as First Lady, 1992-2001.
Figure 1: The number of girls born each year which were named either Hillary, Hilary, Chelsea, Chelsey, or Kelsey. All of the names dropped in popularity following 1992.


And that is when I realized how truly unusual the drop in the name Hillary was. Looking at the top 1000 most popular female names that peaked in popularity between 1880 and 2014, the name Hillary peaking the year before becoming First Lady, experienced the single largest drop in popularity of any name during the first year being First Lady, 1993.

Not only that, but the popularity of the Chelsea, Hillary and Bill's daughter, also took a tremendous blow during entire stint of Hillary's term as First Lady.

Table 1: This table show how the name Hillary and Chelsea dropping in popularity from their all time peaks in 1992. Hillary proportion and Chelsea proportion are what percentage of girls were named those names relative to the peak year. The rank is how low that proportion compares with the top 1000 most popular names in that proportion.

#   Year               Hillary Proportion   Rank                  Chelsea Proportion   Rank  
11992100.0%100.0%
2199342.2%169.8%22
3199416.2%147.7%8
4199512.3%141.8%12
5199612.4%236.3%9
6199711.7%227.6%5
719989.6%121.8%4
8199910.1%216.8%5
9200010.0%214.7%5
10200110.3%413.1%5

From Table 1, we can see that the popularity of the name Hillary dropped dramatically from the 1992 peak to 42.2% of that peak the next year to 16.2% of that peak the following year to 12.3% of that peak the following year. When compared with the entire top 1000 most popular female names which peaked, the name "Hillary" is ranked either 1 or 2 for the largest drop in popularity for the years between 1992 and 2001. For the 5th, 6th, 8th, 9th, and 10th years other female names have happened to drop temporarily in popularity below Hillary but did not stay that low for long.

It is for this reason that the popularity of the name Chelsea can be ranked 8 on average yet still be considered the second largest drop of any female name in the decade after the name has peaked. Looking at Table 2 we can see that the average 10 year ranking for Hillary is by far the lowest compared with any other of the 1000 names. The name Chelsea on the other hand is ranked second but it is really tied with Latoya and these names are not far ahead than the names Aisha, Mindy, Ciara, and Jaime.

Table 2: This table presents a ranking of the 20 largest drops in popularity of baby names for the last 130 years for the top 1000 most popular female girl names. The Ave 10 Year Rank is the rank of name in terms of largest popularity drop over the ten years following the peak. This is the average for each rank from the rank column such as that found in Table 1. Peak is the year at name peaked in popularity. Proportion 10yr is what proportion of children after ten years relative to the peak year are named that name. We can see that Hillary and Chelsea are the second lowest, but for this indicator they happen to be higher than Latoya which peaked in 1984 and Sheena which peaked in 1984.
#      Ave 10 Yr Rank      Name      Peak    Proportion 10yr  
12Hillary199210%
28Chelsea199213%
38Latoya19847%
410Aisha197724%
511Mindy197922%
612Ciara200518%
712Jaime197620%
816Jeannine192928%
917Chelsey199217%
1017Rosalie193829%
1118Sheena19847%
1220Tracey197021%
1328Arielle199129%
1428Peggy195822%
1537Ariel199132%
1638Gale195719%
1738Stefanie198329%
1839Deana197031%
1939Tracy197029%
2040Christie197537%

These tables indicate that the names Hillary and Chelsea hit their peaks simultaneously in 1992 and then dropped dramatically in the decade that followed, more dramatically than any other of the top 1000 female names that peaked in the last 130 years (at least when using the 10 year worse drop ranking average). Surprisingly, the alternative spelling of the name Chelsea as Chelsey and  also hit its peak in 1992 and is ranked as the 9th fastest falling name in popularity on Table 2.

So it looks pretty bad right?

Well, maybe not. Perhaps the name Hillary and Chelsea just became temporarily very popular in 1992 because of the popularity of the First Family, then dropped off in popularity after the family lost its novelty. If this is the case then we should see a tepid or non-existent growth in the popularity of the names Hillary and Chelsea in the 10 years proceeding 1992, perhaps a rapid peak in popularity in 1991 and 1992 followed by a modest decline over time.
Figure 2: This figure shows how the popularity of the names "Hillary" and "Chelsea" grew and fell relative to other female names that have peaked. The x-axis has been normalized so that 0 indicates the peak year for all female names while negative x represents years before peaking and positive, years after peaking. ALL is the average proportion to peak for ALL of the top 1000 female names. Top 100 is the average proportions for the Top 100 fastest falling names while Top 20 is the average proportions for the Top 20 fastest falling names.

From Figure 2, we can see that the Top 100 and Top 20 names (in terms of those that fell the fastest Table 2) did demonstrate more short term steep rising and falling than that of the average for all names and more so for the Top 20 than the Top 100. This may suggest the names Hillary and Chelsea coincidentally rose and fell in a slightly more extreme version than that of the top falling names.

However, looking at the years preceding the peak year we can see that it is unlikely that the drop in popularity of the names Hillary and Chelsea is due to a temporary fascination with these two names. This is because there was a decade long trend suggesting rising popularity of the names Hillary and Chelsea years before they would have been known as public figures. In addition to that, for the name Hillary in particular the rise in popularity was less dramatic than that of the average for the Top 20 suggesting leading into the peak year that it should lose popularity at a rate between that of the Top 20 and the Top 100.

This is not what happened! Within two year the name Hillary had fallen significantly lower in popularity than it was 10 year prior. The name Chelsea also fell dramatically, such that by the end of the 90s it was far less popular than it was during the beginning of the 80s.

Not only that, but the alternative spelling of the name Hillary as Hilary plateaued in 1990 but holding popularity at 96% of that of the peak until 1992 after which time it dramatically fell to 28% relative to peak levels in 1993. Likewise even the close name Kelsey also peaked in 1992 before falling to 25% of its 1992 levels within a decade.

Table 3: This table list the names which peaked for the years 1990-1994 in order of their total popularity. The column Proportion 10yr is what percent of babies were named this name after 10 years from the peak of name popularity.

#PeakNameProportion 10yr                     #PeakNameProportion 10yr
11990Alyson86%11992Carissa61%
21990Blanca55%21992Chelsea13%
31990Cassandra53%31992Chelsey17%
41990Courtney43%41992Christian41%
51990Cristina50%51992Hillary10%
61990Elizabeth74%61992Kasey44%
71990Erika51%71992Kelsey28%
81990Hilary7%81992Silvia70%
91990Katherine68%
101990Leanna55%11993Alexandra65%
111990Mara82%21993Alexandria63%
121990Meagan41%31993Hayley64%
131990Megan60%41993Jasmine73%
141990Rachael55%51993Kassandra43%
151990Samantha74%61993Katelyn88%
161990Stephanie31%71993Kelsie42%
81993Raven59%
11991Ana91%91993Susana59%
21991Ariel32%101993Tania83%
31991Arielle29%111993Taylor54%
41991Ashleigh53%121993Victoria76%
51991Bianca51%
61991Devon31%11994Alejandra76%
71991Kara50%21994Allison72%
81991Karissa72%31994Briana61%
91991Kayla72%41994Larissa77%
101991Kirsten66%51994Marina58%
111991Mercedes63%61994Marissa56%
121991Molly73%71994Tori47%
131991Shelby45%

But what if there was just an unusual number of babies born in 1992 and perhaps that was a peak year for many baby names?

From Table 3 we can see that the exact opposite seems to be the case. The year 1992 did not have an unusually large number of peaking names but rather only 8 names that peaked which was lower than the average of 11.2 for the five years between 1990 and 1994. By looking within the names that peaked for each of the years, we can also see that across all of the names that peaked that year, names that were variants of Hillary or Chelsea were the names were proportionally the least popular 10 years later. Overall, Table 3 supports the assertion that the popularity of the names Hillary and Chelsea and their variants were likely negatively affected by the Clintons's time in the White House.

Conclusions
From looking more closely at the names data, it seems pretty clear that the popularity of the name Hillary as well as Chelsea and their variants were powerfully damaged by the family's term in the highest office. Not only did the names Hillary and Chelsea decrease in popularity rapidly but this decrease was unparalleled among the top 1000 most popular female names recorded for the last 130 years.

How much this finding should be taken into consideration when choosing a presidential candidate, I don't know.

Finally, I would like to say something personal. Whatever happened with the Clinton family happened a long time ago. I was only 10 in 1992 and Chelsea was only 12. It is really hard to imagine anything that Chelsea could have done that would have warranted the kind of public disgust that would have driven the observed unpopularity of the name Chelsea leading it to drop at a such a rate only truly outmatched by her mother's unpopularity.

This strikes me as unfair. And truthfully, this whole analysis strikes me as distasteful.

I, of all people, am the last person in the world who should be criticizing people for their popularity. This is something I have struggled with as I have been the subject of organized negative attacks and public humiliation previously (see Turkopticon: Defender of Amazon's Anonymous Workforce).

Yet, these matters should be brought to the public light because they are of consideration as Hillary Clinton is seeking the nomination for the Democratic party. If there is some fundamental disgust in how a large portion of the American people see her and her daughter Chelsea then it should be brought forward sooner rather than later.

This is after all an election and good or bad an election is a popularity contest.

Source Code on GitHub

Related Articles:
As First Lady, Popularity of Babies Named "Hillary" Dropped by an Unprecedented 90%
Big Business Backs Hillary: Small Bernie
Hillary Clinton's Biggest 2016 Rival: Herself
Analysis: Clinton backed by Big Money: Sanders by Small
Legally Rig An Election: A Citizen's Guide to Gerrymandering 
Nevada:Sanders has 6x the Supporters as Clinton
The Simple Reason Sanders Is Winning
Cause of Death: Melanin | Evaluating Death-by-Police Data
Obama 2008 received 3x more media coverage than Sanders 2016
The Unreported War On America's Poor
What it means to be a US Veteran Today

3 comments:

  1. Wait, what? Why on earth would a drop in popularity necessarily be caused by "disgust", or be a sign of "profound damage"? There are many other more plausible hypotheses. The simplest is that people typically don't name daughters after political or historical figures, whether positive or not. There is a lot of research on the psychology of choosing baby names, and people steer away from "ambitious" or "heroic" names for girls. People want something safe, pleasant, adaptable. The name Hillary was safe and adaptable right up until it became synonymous with an extremely high profile world figure, at which point it would be perceived as an "ambitious" or "aspirational" name.

    I think this hypothesis will hold generally, and will be highly correlated with how synonymous that name is with the political or historical figure. Hillary is an extremely strong case of this, because she was always operating under the context where "Clinton" meant "Bill", so to this day she's typically identified only by first name. The name went overnight from being a safe and very common name to being something akin to "Oprah", "Beyonce", or "Nefertiti". I mean, sure there are parents who aspirationally name their kids "Hillary" or "Oprah", but they're vastly outnumbered by people who want a safe and unambitious name.

    ReplyDelete
    Replies
    1. Now that is not a hypothesis which occurred to me. Perhaps, Hillary and Chelsea are actually thought of as the most heroic women ever?

      Delete
    2. By the way. All of the women you mentioned had an increase in popularity of their name after they became famous.

      Delete