Background: Though approximately 80% of teachers are women, they hold only about 50% of all principalships, and just a quarter of superintendent positions (Finnan and McCord, 2017; Robinson et al., 2017; Tienken, 2021; White, 2023). When women do take on leadership positions, evidence suggests they tend to work in districts serving students with more diverse needs (e.g., students with disabilities, English learners; White, 2023). These student and district demographic characteristics are also associated with lower academic achievement as measured by traditional metrics (e.g., standardized assessments, graduation rates), reflecting inequitable access to resources; these districts may be considered more challenging, and therefore more precarious (McGhee and Nelson, 2005). The tendency for women leaders to be more likely to work in more precarious positions has been described as a "glass cliff" in leadership, building on the glass ceiling metaphor. Defining precarity in education is a challenging endeavor appropriately up for much debate, and here we consider one possible approach. In this study, we examine how state and federal school accountability designations as a measure of precarity may be related to the likelihood of districts being led by women superintendents. Several studies have found evidence of similar phenomena to that first described by Ryan and Haslam (2005) regarding the differential circumstances to which men and women were appointed to the boards of large firms. These studies have found consistent evidence suggesting women are hired into more precarious positions than men. While this model has been studied in business, sports, and politics (Cook & Glass, 2013; Oelbaum, 2016; Ryan et al., 2010), few studies have explored the glass cliffs phenomenon in education (Smith, 2015; White, 2023). Smith's (2015) study examining glass cliffs among principals is important, though in many ways the context of the superintendency may mirror more the dynamics of leading in public firms; White (2023) examines the superintendency directly, but with a focus more on turnover rather than accountability as in this study. We examine the following primary research question: Is school accountability pressure predictive of a district having a woman superintendent? Data: New Jersey offers a unique research setting as a diverse state consisting of around 600 districts, employing a larger proportion of women superintendents than average across the US, about 34% in the state compared to around 25% nationally (Tienken, 2021; White, 2023). The New Jersey state department of education (NJDOE) provides publicly available data regarding the academic performance of school districts as well as student demographic information and personnel data. The Abbott districts were originally identified in 1985 and remained under heightened state control through out the period of study. We operationalize school accountability pressure using data on schools identified in need of improvement for each year from 2008-09 to 2018-19 and Abbott districts under state supervision. School accountability status data were collected from federal data for 2008-09 to 2014-15 (USED, 2012A; USED, 2012B) and from NJDOE (2023) for 2016-17 to 2018-19. We describe any school to have been designated for "low-performance" to be in improvement status (See Table 1). We also observe the criteria that New Jersey used to identify schools for improvement (i.e., percent proficient, growth, high school graduation, chronic absenteeism, English language proficiency). In addition to school accountability data, NJDOE provides data on staff that includes the names, experience, positions, and school placements. Notably, these data do not include the gender of personnel. Modeling our method for assigning gender from recent publications (White, 2023), we use the social security administration's lists of most popular first names by gender by decade from 1950 to 1980, and titles (Mr., Ms., Mrs.) provided for the superintendents of districts in the 2017-18 and 2019-20 school years, to assign gender to personnel. When we could not identify the gender of superintendents using these methods, we conducted a web search (e.g., school websites, social media, news articles) to identify gender. We merge in school, district, and county characteristics to use as controls. NJDOE data includes information on student race/ethnicity, free/reduced-price meal eligibility, limited English proficiency, and Title I status. The staff files include superintendent salary. Finally, we include data on millage rates from New Jersey Department of Community Affairs (NJDCA, 2023). Methods: We will begin by descriptively exploring the relationship between superintendent gender and school accountability pressure. We will then estimate a series of regressions following the general form: Woman[subscript dt]=[beta][subscript 1]SchAcctPressure[subscript dt]+X'+[alpha][subscript d]+[gamma][subscript t]+[mu][subscript dt] "Woman" is a binary indicator (1 if the superintendent is a woman, 0 if they are a man) in each district "d" and year "t." "SchAcctPressure" is binary indicator equal to 1 if greater than 51 percent of the schools in a districts were identified in need of improvement and zero otherwise. In subsequent models we exchange "SchAcctPressure" with a time-variant indicator for Abbott district. "X" is a vector of district characteristics including percent non White, percent Free and Reduced Prince Lunch, percent Limited English Proficiency, percent Title I schools, and tax effort (i.e., total tax revenues divided by total property value for each school district). [alpha] is a state fixed effect and [gamma] is a year fixed effect. [mu] is an idiosyncratic error term clustered by school. Results: The results imply that the position of women superintendents is more precarious than men and school accountability pressures exacerbate that inequality. Women superintendents are about 1 percent more likely to lead districts facing accountability pressures than men (See Figure 1, Panel C). Women are also more likely to be superintendents in Abbott Districts and more likely to be hired as new superintendents into districts facing accountability pressures than men (See Table 4).