Ottoman waqfs basically earned income in two ways, namely by renting their real estate and by operating their cash capital using the “muamele-i şer’iyye” methods. Waqfs that generated income by operating cash or by operating cash and renting real estate are called cash waqfs in the literature. The cash capital of cash waqfs could be increased in several different ways after the establishment of the waqf. One of the most commonly used ways was to add the remaining surplus to the cash capital after deducting the total expenses from the total income at the end of each hegira accounting year. Another issue examined in the studies on cash waqfs is the size of the cash capital of these waqfs and the way in which this increased and decreased.

Ottoman waqfs basically earned income in two ways, namely by renting their real estate and by operating their cash capital using the “muamele-i şer’iyye” methods. Waqfs that generated income by operating cash or by operating cash and renting real estate are called cash waqfs in the literature. The cash capital of cash waqfs could be increased in several different ways after the establishment of the waqf. One of the most commonly used ways was to add the remaining surplus to the cash capital after deducting the total expenses from the total income at the end of each hegira accounting year. Another issue examined in the studies on cash waqfs is the size of the cash capital of these waqfs and the way in which this increased and decreased.

This study is limited to 12 foundations established by ordinary people in Eyüp in the inspection zone of “Havass-ı Refia Kadılığı” between the years 1769- 1773. It aims to analyze the effect of waqf revenues on foundation cash capitals based on the data in the waqf accounting records within the framework of an econometric model. This modest study conducted within a very limited time period and area due to the difficulty of obtaining uninterrupted and regular data from archive documents, presents the findings obtained from the analysis and also offers a secondary contribution to the literature through data obtained from the accounting books used in cash-waqf foundation studies. The study shows that they are suitable for analysis by using econometric methods.

The accounting book used in the creation of the data set of this study is among the books recorded by the Ottoman kadi in the şer’iyye court records and is one of the accounting books held by the trustees of the foundation. The accounting book in question is registered with the number 246 in the records of the Havass-ı Refia Court in the Istanbul Mufti Office Sharia Registry Archives. The data set is based on the information in the income (revenue) sections of the accounting records. In the model, cash capital of the foundations was accepted as the dependent variable and interest and rental income of the foundations were accepted as independent variables. After providing logarithmic transformation for all variables in the model, the series were analyzed using panel data method.

Panel data analysis method was preferred in the study since it enabled the panel data analysis, cross-sectional data and time series data to be analyzed together. The study investigates whether there is an individual or time effect in each unit during the decision process in which panel data model is preferred. For this, F test and Likelihood Ratio (LR) test results were examined. According to these results, the presence of individual effect was determined. As a result of these findings , it was decided according to Hausman test which of the fixed or random effect models would be preferred. According to the test results, the H0 hypothesis, which says that the random effects model is appropriate, was rejected and the fixed effect model was chosen. In order to estimate and interpret the model, it is necessary to test the assumptions of the fixed effect model. For autocorrelation testing, Bhargava, Franzini and Narendranathan’s Durbin Watson Test and BaltagiWu’s local best invariant test, the modified Wald test for testing heteroskedacity, and Pesaran’s CD test for cross-sectional dependence testing were used. Test results for local best invariant tests of Bhargava, Franzini and Narendranathan’s Durbin Watson and Baltagi-Wu are 1.0743907 and 1.5401328, respectively. Considering that these values are less than 2, it is concluded that autocorrelation exists in the model. According to the modified Wald test results, the hypothesis H0 is rejected and it is understood that there is a heteroscedacity according to the units. Finally, cross-sectional dependence was also examined by Pesaran’s CD test, and according to the test results, the presence of the cross-sectional dependence was also determined, where hypothesis was rejected. In order to estimate and interpret the model, the problems determined as a result of the tests should be corrected. The presence of autocorrelation, heteroscedasticity and cross-sectional dependence were corrected by Driscoll and Kraay standard errors method.

According to the findings of the analysis, while interest revenues have a statistically significant effect on cash capital at a 90% confidence level, rental revenues do not have a statistically significant effect on the same . Accordingly, a 1% increase in interest revenues increases the cash capital of foundations by 0.43%. This result shows that the interest income of foundations plays a very important role in the accumulation of foundation capitals. In addition to the findings obtained by the econometric analysis conducted in the study, a secondary result is that the waqf data obtained from the foundation accounting books in the sharia court records are suitable to be evaluated by using econometric methods.