Effects of Digitalization on Agricultural Value Added in Türkiye: An ARDL Approach

Agricultural digitalization has become an important part of productivity growth and sustainable development of agriculture. However, empirical evidence remains mixed about how different dimensions of digitalization affect agricultural value added. This study looks at the short-run and long-run effects of digitalization on agricultural value added in Türkiye, by using annual data for the period 2001–2023. Agricultural value added is used as the dependent variable, while Internet use and fixed broadband subscriptions stand for two indicators of digitalization. Fertilizer consumption and rural population are added as control variables. The Autoregressive Distributed Lag (ARDL) bounds-testing approach is used to estimate both the long-run relationship and the short-run dynamics among the variables. The bounds test provides evidence of a long-run relationship at the 10% significance level, although the result is inconclusive at the 5% level. In the long-run estimates, Internet use has a positive but statistically insignificant coefficient, while fixed broadband subscriptions have a positive and statistically significant coefficient. Fertilizer consumption has a negative but statistically insignificant coefficient, whereas rural population has a negative and statistically significant coefficient. The Error Correction Model (ECM) shows that short-run deviations from the long-run relationship are corrected over time, with a negative and statistically significant error-correction coefficient. The diagnostic tests indicate no evidence of serial correlation or heteroskedasticity, while the CUSUM tests support parameter stability. An additional analysis using mobile cellular subscriptions extends the sample period and shows that the results are sensitive to the digitalization indicator used. In this alternative specification, the bounds test does not provide evidence of a long-run relationship. The findings therefore do not support a uniform relationship between digitalization and agricultural value added. Instead, the results vary across digitalization indicators and model specifications. Rather than combining digitalization measures into a single ICT indicator, the study examines alternative digitalization measures separately to assess whether the results depend on the indicator used.