TL;DR: The "20 times higher" claim likely stems from Dhejne et al. (2011), who found this was true in the 1970s and 1980s, but not after 1989 (with the authors suggesting improved health care and societal acceptance may have led to their observations).
Transgender (completed) suicide rates are poorly estimated for a slew of practical reasons, but are nevertheless generally accepted to be substantially higher than the general population, but probably not by a factor of 20. I selected 4 studies with reasonably large sample sizes which reported a factor of: (a) 3.6 [Amsterdam; completed suicide]; (b) 5.7 [USA veterans; completed suicide]; (c) 9 to 12 [USA survey; attempted suicide]; (d) 8 to 13.6 higher [Injustice at Every Turn (+deductions); attempted suicide]. We also should bear in mind that suicide rates fluctuate significantly based on a large number of factors (e.g., the USA suicide rate is around 14.4, and varies from 6.2 in the District of Columbia to 29.4 in Wyoming).
It seems the quote's implication is that transitioning leads to suicidality, whereas in reality there is a large volume of evidence that transitioning improves quality of life:
We identified 55 studies that consist of primary research on this topic, of which 51 (93%) found that gender transition improves the overall well-being of transgender people, while 4 (7%) report mixed or null findings. We found no studies concluding that gender transition causes overall harm.
What does the scholarly research say about the effect of gender transition on transgender well-being?, Cornell University. They website links to all 55 studies, along with 17 literature reviews and practitioner guidelines.
To begin, we need to understand the difficulty here. America's Health Ranking defines suicide rate as deaths due to intentional self-harm per 100,000 population. So we calculate it by
([total suicides in given year] / [population of interest in that year]) * 100,000.
In 2019, there was reportedly 47,511 suicides in the USA, so we calculate a USA 2019 suicide rate of 14.4. This is the so-called "crude suicide rate"; other sources give age-adjusted suicide rates, such as the CDC gives 13.9 for the USA in 2019 (see also Age-Adjusted Death Rates for an explanation as to how it's calculated).
Note there is significant fluctuation in suicide rates based on location from 29.4 in Wyoming to 6.2 in the District of Columbia (crude, USA, 2019; source) [aside: half of all successful suicides in the USA in 2019 were firearm suicides (source) which varies from state to state, and likely plays a significant role], and globally from 0.4 in Antigua and Barbuda and 72.4 in Lesotho (crude; global; 2019 source), and based on sex from 26.1 male to 7.0 female (crude; USA; 2019 source).
Now, if we were to perform the same calculation for the USA's transgender population, we'd run into two problems: (a) We don't know the number of "people who undergo gender transition"; a survey states 1.4 million adults identify as transgender, but that doesn't imply they transition, and it excludes non-adults. (b) We don't have accurate statistics of how many "people who undergo gender transition" die by suicide (often people don't disclose their or others' transgender status).
Thus, we necessarily need to work with subpopulations, ideally containing a large number of transgender samples.
Beyond this, there's a whole bunch of caveats:
However, rates of completed suicide are difficult to ascertain for the trans population. It would be highly unusual for a person’s trans status to be recorded on a death certificate; it is therefore not possible to ascertain trans suicide rates from coronial data. Conversely, because the trans population is ‘hard-to-reach’ published studies tend to comprise limited, and typically self-selecting, samples making it difficult to assess the extent of suicidality in the trans population as a whole. For example, the majority of studies adopt a gender binary approach to trans people, sampling only trans men and/or trans women; thus excluding non-binary trans people. In other cases, sample selection is restricted to highly specific subgroups of the trans population, such as ‘trans women with a history of sex work’ (e.g. Nemoto, Bodecker & Iwamoto, 2011) or solely those undergoing surgical intervention (e.g. Heylens et al., 2014). Together these factors contribute to partial information and potential underreporting of suicide rates in trans populations (Bauer, Scheim, Pyne, Travers, & Hammond, 2015; Haas et al., 2010).
McNeil et al., Suicide in Trans Populations: A Systematic Review of prevalence and correlates (pdf), Psychology of Sexual Orientation and Gender Diversity, 2017.
Other issues we encounter is that (a) medical definitions have changed over time, (b) medical treatments for transgenderism have improved significantly over time, (c) societal awareness and acceptance of transgenderism has changed over time. So we should also look to more recent studies.
Dhejne et al. (2011)
user141592's answer points to Dhejne et al. 2011 which utilizes Swedish Cause of Death Register data, and this seems to be the origin of the number "20": you can see this particular paper referenced by Mike Lee when discussing another bill Protecting Our Kids from Harmful Research Act. Indeed, Dhejne et al.'s Table 2 states that between 1973 and 2003 there were 10 suicides of their transsexual group (size 324), and 5 suicides of their control group (size 3240); hence the factor of 20. However, the paper highlights:
...the survival curve (Figure 1) suggests increased mortality from ten years after sex reassignment and onwards. In accordance, the overall mortality rate was only significantly increased for the group operated before 1989. However, the latter might also be explained by improved health care for transsexual persons during 1990s, along with altered societal attitudes towards persons with different gender expressions.
Dhejne et al., Long-Term Follow-Up of Transsexual Persons Undergoing Sex Reassignment Surgery: Cohort Study in Sweden, PLOS One, 2011. See also the literature review Dhejne et al., 2016.
There is more about this in an interview with Dhejne at The Trans Advocate who points to this Huffington Post piece about the way my research is misrepresented.
In other words, the "20 times higher" applies for the 1970s and 1980s, but no longer holds between 1989 and 2003.
Amsterdam Cohort (2020): 3.6 times higher.
A Dutch gender dysphoria clinic reports their long-term statistics with suicide rates. Their raw numbers are:
Forty-nine people died by suicide: 41 trans women (0.8%) and 8 trans men (0.3%), which is 64 per 100 000 person years in trans women and 29 per 100 000 person years in trans men.
Wiepjes, et al., Trends in suicide death risk in transgender people: results from the Amsterdam Cohort of Gender Dysphoria study (1972–2017), 2020. Sample size: 8,263.
The study emphasizes that in the Netherlands, between 1972 and 2017 suicide rates showed a fluctuating course, so they restrict the timeframe, and give the following statistics:
The mean number of suicides in the years 2013–2017 was higher in the trans population (40 per 100 000 person years; 43 per 100 000 trans women and 34 per 100 000 trans men) compared with the Dutch population in this time frame (11 per 100 000 person years; 15 per 100 00 registered men and 7 per 100 000 registered women)
In this sample, the transgender suicide rate is about 3.6 times higher than the Dutch population.
USA veterans (2014): 5.7 times higher.
The crude suicide rate among veterans with transgender-related ICD-9-CM diagnoses across the 10-year period was approximately 82/100,000 person-years, which approximated the crude suicide death rates for other serious mental illness in VHA (e.g., depression, schizophrenia).
Blosnich et al., Mortality Among Veterans with Transgender-Related Diagnoses in the Veterans Health Administration, FY2000-2009, LGBT Health, 2014. (paywalled). Sample size: 5,117.
In this context, from 2001 to 2019, the unadjusted suicide rate among non-Veteran U.S. adults rose 33.0%, from 12.6 per 100,000 in 2001 to 16.8 per 100,000 in 2019.
2021 National Veteran Suicide Prevention, 2021 (pdf).
So the transgender suicide rate is somewhere in the ballpark of 5.7 times higher.
U.S. Transgender Survey (2015); suicide attempts: 9 to 12 times higher.
Since we're dealing with ratios, "attempted suicide" may serve as a practicable proxy for "completed suicide", as has been done in some studies, e.g.:
While completed suicide rates among trans people are unknown, a history of attempted suicide is the strongest predictor of completed suicide across multiple populations.
Bauer, Intervenable factors associated with suicide risk in transgender persons: a respondent driven sampling study in Ontario, Canada, BMC Public Health, 2015.
To my knowledge at the time of writing, the largest study along these lines is the following:
- Forty percent (40%) have attempted suicide in their lifetime, nearly nine times the rate in the U.S. population (4.6%).
- Seven percent (7%) attempted suicide in the past year—nearly twelve times the rate in the U.S. population (0.6%).
2015 U.S. Transgender Survey p.10 (pdf) which includes 27,715 respondents of which 16,580 (60%) "currently live full-time in a gender that is different from the one assigned to you at birth" (p.244).
Injustice at Every Turn (2011): suicide attempts: 8 to 13.6 higher.
When asked “have you ever attempted suicide?” 41% of respondents answered yes. According to government health estimates, five million, or 1.6% of currently living Americans have attempted suicide in the course of their lives. Our study asked if respondents had ever attempted suicide while most federal studies refer to suicide attempts within the last year; accordingly it is difficult to compare our numbers with other studies. Regardless, our findings show a shockingly high rate of suicidality.
Grant et al., Injustice at Every Turn (pdf), 2011. Sample size: 6,450.
It is indeed hard to get an accurate estimate of the lifetime suicide attempt rate: I wasn't able to track down the source for the "1.6%" in the above quote, the closest I found was SAMHSA 2015 which are self-reported (and are necessarily incomplete) numbers, and this edition doesn't cover general population lifetime suicide attempts. I found a meta study Nock et al., 2008 writing: it is 1.9–8.7 percent (IQR, 3.0–5.1), so the factor is likely between 8 and 13.6.