The Asymmetric Effects of Agricultural Policy Support on Agricultural Employment in Türkiye: A NARDL Approach

Abstract

The agricultural sector is recognized as one of the primary economic pillars of Türkiye, contributing significantly to employment, GDP, and export revenues. However, the proportion of the workforce employed in agriculture has been diminishing in conjunction with the growth of the industrial and service sectors. Despite this long-term decline, annual fluctuations in the absolute value of agricultural employment suggest that the sector functions as a stabilizing factor, mitigating the impact of general labor market fluctuations and exhibiting high sensitivity to short-term changes in profitability and support policies. To support this sector, the government implements various policy instruments, including direct payments, input subsidies, and investment incentives. The extant literature suggests that the effects of these subsidies on employment exhibit a complex and paradoxical pattern. The promotion of capital-intensive technologies, the absence of direct employment effects associated with decoupled payments, and the consolidation of large-scale farms are the main mechanisms through which these interactions occur. It has been emphasized that reducing support could lead to more abrupt and disproportionate declines in employment, particularly by driving less efficient farms out of the sector. This reflects a downward asymmetry. In this context, the study analyzed the short- and long-term asymmetric effects of agricultural policy support, as measured by the OECD Producer Support Estimate (PSE) index, on agricultural employment levels in Türkiye. The dataset used for this study consists of annual time series data on agricultural employment, PSE, agricultural credit volume, agricultural product exports and imports, and the share of agriculture in gross domestic product (GDP) from 1986 to 2023. This data was obtained from the TurkStat, SBB, TBB, and OECD databases. The asymmetric relationships are tested using the NARDL cointegration methodology, which is a non-linear lag distributed autoregressive approach. According to the findings, positive shocks in PSE and agricultural credits significantly and positively affect employment in the long run, while positive shocks in agricultural exports negatively affect employment. Conversely, negative shocks to the share of agriculture in GDP increase employment, confirming the sector's \"employment buffer\" function. Long-run asymmetry is found in PSE and agricultural credits. In the short run, negative shocks to agricultural exports negatively impact employment, while positive shocks to agricultural credit have a negative lagged effect on employment. However, positive shocks to the share of agriculture in GDP increase employment in the short run. Short run asymmetry is observed for export and GDP share variables, but no significant asymmetry is found for PSE. Accordingly, the study suggests that policymakers adopt a multidimensional approach to agricultural subsidies that considers not only the amount, but also the design, orientation, and implementation context. The development of mechanisms that safeguard rural employment and promote alternative employment opportunities is of critical importance. It is recommended that flexible support and credit instruments be designed to target vulnerable groups in the face of short-term negative shocks, with priority given to labor-intensive subsectors and rural diversification initiatives. Moreover, policies aimed at facilitating transitions to non-agricultural employment—such as vocational training programs and support for local entrepreneurship—should be strengthened.