The Inflation Threshold and Economic Growth Dynamics in Algeria: A Smooth Transition Regression (STR) Model Study (1980–2024)

Abstract

This study aims to explore the complex and nonlinear relationship between inflation and economic growth in Algeria over the period 1980–2024, with a particular focus on identifying the “inflation turning point” at which inflation’s effect on growth shifts from positive to negative. The research is based on the hypothesis that the rentier nature of the Algerian economy coupled with its heavy dependence on oil price fluctuations and the central role of the state makes it difficult to analyze its behavior using traditional linear economic models. For this reason, the study employs the Smooth Transition Regression (STR) model, which is particularly suited to capturing gradual changes in variable effects at specific threshold levels. The analysis uses annual data spanning 45 years and includes five key variables: real GDP growth rate (GDPG), the log-transformed inflation rate (LNINF), credit to the private sector (FD), government spending (GOV), and the investment rate (INV). The data were sourced from reputable institutions such as the World Bank and the IMF. After ensuring the stationarity of the time series using ADF and PP tests, preliminary correlation and linearity tests were conducted before applying the STR model. The results revealed that the linear model fails to fully capture the dynamics of the Algerian economy. Investment had a statistically significant positive impact on growth (p < 0.05), while credit had only a weak influence (around 0.05). Inflation and government spending showed no clear effect unless a structural shift occurred. In the nonlinear segment of the model, inflation was found to become a constraint on growth after crossing a threshold of 0.708, corresponding to an annual inflation rate of approximately 5–6%. Beyond this point, inflation exerted a strongly negative and statistically significant effect on growth (p < 0.001), and the effectiveness of investment declined under high inflation conditions. Model diagnostics, including an R² of 0.9735, a strong F-statistic, and low Akaike and Schwarz information criteria, confirmed the adequacy of the STR model for this dataset. These findings underscore the importance of adopting more flexible economic policies, especially in maintaining inflation within safe bounds, improving the efficiency of spending and investment, and enhancing the role of financial intermediation in financing the private sector. The study also opens the door for further research using more advanced dynamic models such as LSTR or ESTAR, particularly in rent-based economies like Algeria.