Eficiência relativa dos gastos com terceirização de mão de obra das universidades federais brasileiras no contexto da pandemia

Abstract

The general objective of this research was to analyze the relative efficiency of spending on outsourcing cleaning and property surveillance services of Brazilian federal universities, in the context of the pandemic. Among the 69 universities, the sample included 30 of them, with data available for the feasibility of the study. The envelope analysis, DEA/BCC model, oriented to inputs, was carried out in the period from 2018 to 2022, as well as the DEA/IPM for the pre pandemic period (2018/2019), for the pandemic period (2020/2021) and for the post-pandemic period (2022). The results showed that the efficiency of cleaning and surveillance contract spending during the pandemic was negatively impacted, because the highest number of efficient DMUs was observed in 2018 and 2022, before and after the pandemic, and the lowest minimum efficiency values were also observed during the pandemic. In addition, with regard to the factors of the analysis with the greatest potential for improvement, the DEA modeling pointed to the need to reduce the number of surveillance posts in the pre-pandemic and pandemic period and, in the post-pandemic period, pointed to the need to focus on reducing the % of Surveillance Expenses in relation to the total current expenses settled. except for legal expenses, as well as for the reduction of the number of jobs in the cleaning contract. The correlation test revealed the existence of a weak and direct significant correlation between the variation in productive efficiency (IPM) of cleaning and surveillance contracts and the academic indicators of the TCU: Degree of Student Involvement with Graduate Studies (GEDPG), as well as the CAPES/MEC Concept of Graduate Studies, from 2018 to 2022, with emphasis on the fact that, when analyzed by period, it was observed that there was no significant correlation between the same IPM and CAPES/MEC only in the post-pandemic period.