本文選用1981年到2016年美國上市公司的資料檢測勞動者失業風險與公司債務期限結構的關係。提出假說公司會通過增加短期債務來降低勞動者對失業風險的感知繼而減少補償工資差額的要求。實證證據表明，上一年的失業救濟收益與三年內到期的短期債務比率呈顯著的負向關係。所有迴歸模型的結果與假設一致，並且在經濟和統計上都是顯著的。 本研究利用重新匹配後的樣本進行了雙重差分模型的驗證。使用二階段GMM迴歸解決內生性問題，並進一步發現實證結果在受融資約束的企業中更為明顯。作為穩健性檢驗，本研究使用每週的最大失業救濟收益替代原解釋變數，使用不同的短期債務比率替代原被解釋變數。此外，偽證檢驗表明債務期限的變化與前一年而不是當年或未來一年的失業救濟收益的變化有關。最後本研究還證實政治因素並不影響主要結論。 ;This paper selects data from American listed companies from 1981 to 2016 to test the relationship between the unemployment risk and the debt maturity structure. The hypothesis is that companies will reduce their perceived risk of unemployment by increasing short-term debt and then reduce the demand for compensating wage differentials. Empirical evidence shows that the unemployment insurance benefits last year has a significant negative relationship with the short-term debt ratio in three years. The results of all regression models are consistent with assumptions, and are statistically significant in terms of economics and statistics. The double difference model is verified by using the reselected samples. The two-stage GMM regression is used to solve the endogenous problem, and further finds that the empirical results are more obvious in enterprises with financing constraints. As a robustness test, this study uses the weekly maximum unemployment benefits to replace the original explanatory variables and use different short-term debt ratios to replace the original interpreted variables. In addition, the falsification test shows that the change of debt maturity is related to the change of unemployment benefits in the previous year rather than in the year or year ahead. Finally, this study also confirms that political factors do not affect the main conclusions.