Are Hurricanes with Female Names More Severe or Costly than Male-Named Hurricanes? ()
1. Introduction
In 1953, the United States began using only female names for tropical storms. This practice ended in 1978 for tropical storms in the Eastern North Pacific and ended one year later for tropical storms in the Atlantic and Gulf of Mexico. Since 1979, the naming of all U.S. tropical storms has alternated between male and female names (National Hurricane Center, 2022a).
In 2014, Jung, Shavitt, Viswanathan, & Hilbe (2014) argued that hurricanes with female names were deadlier than male hurricanes. Later that same year, Malter (2014) using the same data, methodology, and variables reached the conclusion that female- and male-named hurricanes were equally deadly. One problem with the Jung et al. study was that the authors examined the death rates from U.S. hurricanes between 1950 and 2012 to show that feminine-named hurricanes cause significantly more deaths than male-named hurricanes. But, as previously mentioned, before 1978 only female names were used. Neither Jung et al. nor Malter used actual measures of hurricane severity.
In this paper, comparisons are made between hurricanes from 1979 to 2021 with female and male names based on 1) average maximum wind speed, that is, the estimated maximum sustained (1-minute) surface winds to occur along the U.S. coast, measured to the nearest 5 knots (where 1 knot, abbreviated kt, equals 1.15 miles per hour); 2) average central pressure of the hurricane at landfall, measured in millibars (mb); and 3) average cost (expressed in inflation-adjusted 2022 dollars) based on estimates from National Oceanic and Atmospheric Administration’s National Centers for Environmental Information that includes crop damage, individual payouts, and disaster money from the federal government to the affected states.
2. The Data
The analysis in this paper is based on 73 named hurricanes (excluding tropical storms) that made landfall along either the U.S. Atlantic Coast or the Gulf of Mexico between 1979 and 2021. All hurricanes produce sustained winds of greater than 74 mph (119 km/h). These data are shown in Table 1 and were revised in April 2022 to include the 2021 season (National Hurricane Center, 2022b). Table 1 includes the names; gender of the names; month and year the hurricane made landfall; the highest Saffir-Simpson U.S. category, a scale that ranges from 1 (weakest) to 5 (strongest) based on estimated maximum sustained (1-minute) surface winds produced at the coast; estimated maximum wind speed (in knots); and estimated central pressure of the hurricane at landfall (in millibars). Table 1 includes two hurricanes from the very active 2020 Atlantic hurricane season that were named Delta and Zeta, after gender-based names were exhausted. This practice of using Greek letters to name tropical cyclones was discontinued by the World Meteorological Organization after 2020. Moreover, neither Delta nor Zeta are included here in the comparisons of the remaining 71 hurricanes with feminine or masculine names.
Not all 71 hurricanes with feminine or masculine names (between 1979 and 2021) caused at least $3.2 billion (expressed in 2022 dollars) in damage. The 34 hurricanes that did, however, are listed in Table 2. These data are from National Center for Environmental Information (2022). For all U.S. hurricanes, Hurricane Katrina (2005, $186.3 billion) is the costliest storm on record. Hurricane Harvey (2017, $148.8 billion) ranks second. Hurricane Maria (2017, $107.1 billion) ranks third, but is not listed among “continental United States” hurricanes in National Center for Environmental Information (2022). Hurricane Maria (a category 5 hurricane) devastated the Caribbean islands of Dominica, Saint Croix, and Puerto Rico.
3. Methodology
Two-sample t-tests are used to compare hurricanes with female names against
![]()
Table 1. Hurricanes with female and male names, 1979-2021.
Source: National Hurricane Center, NOAA, “Continental United States Hurricane Impacts/Landfalls 1851-2021” at https://www.aoml.noaa.gov/hrd/hurdat/All_U.S._Hurricanes.html.
![]()
Table 2. Costliest continental United States hurricanes impacts/landfalls with female and male names, 1979-2021.
Source: National Hurricane Center, NOAA, “Costliest U.S. Tropical Cyclones” at https://www.ncei.noaa.gov/access/billions/dcmi.pdf.
male-named hurricanes. First, we compare average maximum wind speeds (in knots) by month (July, August, September, and October) and then by decade (1980s, 1990s, 2000s, and 2010s). Although a typical hurricane season begins in June and ends in November, only one of the named hurricanes in our sample of 71 occurred in June (Bonnie in June 1986) and one occurred in November (Kate in November 1985).
Next, we compare average maximum wind speeds across decades (the 1980s through the 2010s); four decades taken two at a time yield six comparisons. Similar comparisons are made of average barometric central pressure (in millibars). The lower the barometric pressure, the stronger the hurricane. Category 5 hurricanes, for example, have a central pressure of less than 920 millibars. Only two hurricanes in our sample, both male-named—Andrew in August 1982 and Michael in October 2018—reached Category 5 on the Saffir-Simpson scale.
We compare the average cost (in billions of 2022 dollars) of all female-named hurricanes that made landfall along the Atlantic Coast or in the Gulf of Mexico since 1979 against the average cost of all similarly defined male-named hurricanes.
All t-tests involve independent samples. All t-tests are two-tailed, that is, under the null hypothesis the two averages are equal and under the two-tailed alternative hypothesis the two averages are different. Under the two-tailed alternative, female-named hurricanes are not presumed to be stronger (or weaker) than their male counterparts. Finally, we run a chi-squared test relating the gender of the hurricane’s name to the hurricane’s highest Saffir-Simpson category.
4. The Results
Table 3 shows the results of all comparisons involving average maximum wind speeds and average central pressure across months and decades. Not surprisingly, maximum wind speed and minimum pressure are strongly correlated (r = −0.8469, p < 0.0001). By month, male-named hurricanes have the same or slightly lower average maximum wind speeds (slightly higher average central pressure) than their female-named counterparts. But, in no case are the differences statistically discernable (using an alpha-level of 0.05 or even 0.10). In no case are there discernable differences between female-named and masculine-named hurricanes by decade.
When average maximum wind speed (Table 4) and average central pressure (Table 5) are compared across decades for either female-named hurricanes (top half of the table) or male-named hurricanes (bottom half of the table), there are no discernable differences. For example, female-named hurricanes (as well as male-named hurricanes) were not stronger (either higher average maximum winds or lower average central pressure) in the 2010s than they were in the 1980s. If one compares female-named hurricanes in the 1980s to female-named hurricanes over the period 2010 through 2021, the difference in average maximum wind speed was still not significant (p = 0.4517), but the difference in
![]()
Table 3. Differences in average maximum wind and central pressure, hurricanes with female and male names, 1979-2021.
aThe p-value reported is for a two-tailed test. That is, under the alternative hypothesis, the average for hurricanes with female names is not equal to the average for hurricanes with male names. bMonthly comparisons include data from 1979 through the year 2021.
![]()
Table 4. Differences in average maximum wind, hurricanes with female and male names, by decade.
aThe p-value reported is for a two-tailed test. That is, under the alternative hypothesis, the average for hurricanes in Group 1 is not equal to the average for hurricanes in Group 2.
![]()
Table 5. Differences in average central pressure, hurricanes with female and male names, by decade.
aThe p-value reported is for a two-tailed test. That is, under the alternative hypothesis, the average for hurricanes in Group 1 is not equal to the average for hurricanes in Group 2.
average central pressure was slightly more noticeable (p = 0.1028). Adding data on male-named hurricanes in 2020 and 2021 made no difference in the comparison to average maximum wind speed in the 1980s (p = 0.6539) or in the comparison to average central pressure in the 1980s (p = 0.4422).
If there were data on male-named hurricanes before 1979, one could employ a difference-in-differences (DID) model. That is, one could compare the difference in intensity of female-named and male-named hurricanes before 1979 (when hurricanes were given alternating female and male names) to the difference in intensity of female-named and male-named hurricanes after 1979. But, all hurricanes before 1979 (and, in particular, between 1953 and 1979) were named after females. When male-named hurricanes from 1980 to 1989 were compared to male-named hurricanes from 2010 to 2019, there were no statistically discernable differences in either average maximum wind intensity or average barometric pressure; the same conclusion was reached when female-named hurricanes were compared. Moreover, when female-named hurricanes were compared to male-named hurricanes in the 1980s, there was no discernable difference; when female-named hurricanes were compared to their male counterparts in the 2010s, again, there was no difference. Based on the results presented in Table 3 and Table 4, a DID analysis would likely reveal that the change to giving hurricanes alternating female and male names would show that neither the hurricanes’ average maximum wind speed nor average barometric pressure changed in response to the exogenous event of changing names in 1979.
When the average cost (in 2022 dollars) of female-named hurricanes is compared to that of male-named hurricanes, there is no discernable difference between the two groups. Even when Katrina, the costliest storm on record, is included, the average cost of the seventeen female-named hurricanes in Table 5 is $35.68 billion and the average cost of the seventeen male-named hurricanes is $24.7 billion, but this difference is not statistically discernable (p = 0.4394). When Katrina is excluded, the p-value of the two-tailed test rises to 0.8847.
Finally, a chi-squared test on a contingency table relating the gender of the hurricane’s name to the highest Saffir-Simpson U.S. category (hurricanes “1” or “2” or “3, 4, 5” grouped together) yielded a test statistic of
and a p-value of 0.634. In other words, female-named hurricanes are not disproportionately severe (category 3 or higher).
5. The Number of Hurricane Deaths
The National Oceanic and Atmospheric Administration (NOAA) in 2011 compiled a list of the mainland U.S. tropical cyclones causing 25 or more deaths between 1851 and 2010 (Blake, Landsea, & Gibney, 2011). A look at the list reveals that between 1953 and 1978, there were eleven hurricanes, all with female names: Audrey (416 deaths, 1957); Camille (256, 1969); Diane (184, 1955); Agnes (122, 1972); Hazel (95, 1954); Betsy (75, 1965); Carol (60, 1954); Donna (50, 1960); Carla (46, 1961); Hilda (38, 1964); and Connie (25, 1955). Between 1979 and 2010, there were seven hurricanes including two tropical storms on the list, three with female names [Katrina (1200, 2005); Allison, only of tropical storm intensity (412, 5001); and Fran (26, 1996)] and four with male names [Floyd (56, 1999); Alberto, a tropical storm (30, 1994); Andrew (26, 1992); and Ivan (25, 2004)]. A comparison of the average number of deaths for the three female-named tropical cyclones (422 including Katrina, 33.5 without Katrina) to the average of the four male-named tropical cyclones (34.25) would not be very meaningful because Katrina was the third deadliest tropical cyclone on record, behind two unnamed hurricanes, one in 1928 (2500 - 3000 deaths) and the deadliest in 1900 (8000 - 12,000 deaths).
The number of deaths caused by typhoons depends not only on maximum wind speed, but a host of other factors. For examples, the number of deaths caused by typhoons will be influenced by the landfall’s geographical characteristics, population density, and the extent of vulnerable housing. The number of deaths will likely be lower in areas with evacuation shelters and disaster-proof homes to withstand hurricanes. But, deaths would be higher in areas where residents are reluctant to evacuate their homes. Less intense typhoons (category 1 or 2 hurricanes) may be accompanied by torrential rains and extensive flooding that result in a large number of deaths. Although there is no gender gap between male and female names for typhoon intensity, further analysis would be needed to determine the validity of the question whether hurricane preparedness is related to the gender of a hurricane’s name.
6. Concluding Remarks
A female-named hurricane is not a harbinger of doom. An analysis of the 71 named hurricanes that made landfall along the Atlantic Coast or the Gulf of Mexico between 1979 (when an alternating male-female naming system was adopted) and 2021 reveals that those with female names have neither stronger average maximum winds nor lower average central barometric pressure, by month or by decade. Female-named hurricanes are not, on average, more costly than their male counterparts. In short, female hurricanes are no more severe than male hurricanes. Hurricane severity does not depend on the gender of named hurricanes.
Acknowledgments
The author wishes to acknowledge the assistance of Maria M. Torres, Communications & Public Affairs Officer Meteorologist, NOAA Communications & External Affairs, at the National Hurricane Center in Miami, Florida who identified and provided the links to the latest available data used in this study.