Do agricultural commodity prices asymmetrically affect the performance of value-added agriculture? Evidence from Pakistan using a NARDL model

Pakistan's agriculture sector is crucial to its economy, contributing significantly to GDP and employment. However, its growth has been hampered by various factors, including volatile commodity prices, insufficient credit, and limited skilled labor. This study addresses the research question of how positive and negative agricultural price shocks affect agricultural value-added (AVA) in Pakistan. The existing literature on the relationship between commodity prices and agricultural growth often overlooks the potential asymmetry in the responses to price shocks. Some studies suggest a short-term positive effect of rising commodity prices on economic development, while others demonstrate that commodity price fluctuations don't necessarily lead to long-term increases in per capita income. There's also a lack of research specifically examining the asymmetric impacts of positive and negative price shocks on agricultural growth in the Pakistani context. Therefore, this study uses a nonlinear approach to investigate the impact of price shocks, along with credit disbursement and labor force, on AVA in Pakistan.

Existing literature focuses on the agricultural sector's contribution to growth, the impact of government expenditure on GDP growth, and the effect of commodity price volatility on GDP. However, few studies analyze the differential effects of positive and negative commodity price shocks on agricultural value-added. Previous research has highlighted the potential asymmetry in responses to price shocks, suggesting that negative shocks might have more severe and lasting consequences than positive shocks. This study builds upon this literature by using a nonlinear ARDL model to investigate this asymmetry in the context of Pakistan's agricultural sector, considering the impact of credit disbursement and labor force alongside price shocks. The researchers discuss how various factors such as climate change, energy prices, exchange rates, and global market dynamics affect commodity pricing, influencing the agricultural sector's performance and economic growth in Pakistan.

The study employs a nonlinear autoregressive distributed lag (NARDL) model to analyze the long-run and short-run effects of agricultural commodity prices (PI), credit disbursement (CD), and labor force (LF) on agricultural value-added (AVA) in Pakistan from 1970 to 2018. The NARDL approach is chosen because it accounts for potential nonlinearities and asymmetries in the relationship between variables and does not require strict stationarity tests, making it suitable for smaller sample sizes. The authors use data from World Development Indicators, Pakistani statistical yearbooks, and economic surveys. The price index (PI) is constructed using wholesale prices of wheat and cotton. Before estimation, unit root tests (ADF, PP, KPSS) were conducted to ensure that no I(2) variables were present, a precondition for applying the NARDL model. The NARDL model is estimated using both a long-run equation and an error correction model (ECM) to capture both the short-run dynamics and long-run relationships. The positive and negative components of the price index were separated to capture the asymmetric effects of price shocks. The study uses a bounds testing procedure developed by Pesaran et al. (2001) to test for long-run cointegration among the variables. After estimating the NARDL model, the authors perform diagnostic tests (serial correlation, normality, heteroscedasticity) and stability tests (CUSUM, CUSUMSQ) to validate model specifications and assess parameter stability. Finally, dynamic multiplier graphs (DMG) illustrate the dynamic responses of AVA to positive and negative price shocks.

The empirical results reveal a long-run cointegrating relationship between AVA, PI, CD, and LF. Both positive (PI+) and negative (PI-) price shocks have positive and significant effects on AVA in both the short and long run; however, the effect of positive shocks is more substantial. In the short run, a 1% increase in PI+ increases AVA by 0.168%, while a 1% increase in PI- increases AVA by 0.507%. In the long run, the corresponding effects are 0.418% and 0.109%, respectively. Credit disbursement (CD) has a significant positive impact on AVA in both the short and long run. A 1% increase in CD results in a 0.379% short-run increase and a 0.055% long-run increase in AVA. Similarly, the labor force (LF) displays a significant positive impact on AVA in both time horizons with a short-run impact of 1.291% and a long-run impact of 1.138%. The model's diagnostic tests confirm the absence of serial correlation, heteroskedasticity, and other specification problems. The stability tests (CUSUM and CUSUMSQ) indicated that the model is structurally stable. The dynamic multiplier graphs visually demonstrate the asymmetric response of AVA to price shocks, confirming that positive shocks have a greater and more persistent impact on AVA compared to negative shocks.

The findings suggest that while both positive and negative price shocks stimulate agricultural growth in Pakistan, positive shocks have a stronger effect. This positive asymmetry might be because higher prices increase farmers' income, leading to higher investment and production. However, the substantial effect of negative shocks suggests the need for robust policies to mitigate the negative impacts of price decreases on farmer incomes. The significant positive impact of credit disbursement indicates the importance of supporting the agricultural sector with adequate credit facilities. The positive effect of labor force suggests that investing in education and skills development within the agricultural sector will be beneficial for overall agricultural growth. The results obtained are in line with several other studies that have also observed positive impacts of price changes in agricultural products. The relatively small coefficient of CD in the long run underscores the need for effective policy interventions to improve credit access and utilization by farmers. The study recognizes certain limitations inherent to the analysis, including the focus solely on Pakistan, the potential influence of unobserved factors, and the selection of specific agricultural commodities (wheat and cotton).

This study provides valuable insights into the asymmetric effects of agricultural commodity prices on the value-added agriculture in Pakistan. Both positive and negative price shocks have a positive impact on AVA, with positive shocks having a greater influence. Credit disbursement and labor force also play significant roles in stimulating agricultural growth. The findings highlight the necessity for targeted policies to improve credit access, enhance farmer income stability, and invest in agricultural human capital. Future research could extend the analysis to other agricultural commodities, incorporate other relevant factors like technological advancement, or compare the findings with other agricultural economies.

This study is limited to the analysis of Pakistan's agricultural sector and may not be generalizable to other countries with different agricultural structures and policy environments. The use of a specific price index based on wheat and cotton might not fully capture the variability of all agricultural commodity prices. Moreover, unobserved factors could potentially influence the relationship between the variables studied, and future research should consider incorporating additional relevant variables. The time frame of the study and data availability may also restrict the generalizability of results.