Introduction

The Arab Spring started with the hope of achieving democratic and economic restoration but ended in a turmoil. The social, economic, and political impact of the Arab uprisings remains underexplored (Marchi, 2021). After decades of political hibernation, the Arab world witnessed mass protests and uprisings, with large-scale revolutions occurring in six countries: Tunisia, Syria, Egypt, Yemen, Libya, and Bahrain (Grinin & Korotayev, 2022). Political factors, such as internal conflicts, religious diversity, authoritarian government systems, and a lack of transparency, are considered major causes of the Arab uprisings (Korotayev et al. 2022; Barakat & Fakih, 2021). The revolutions were further driven by poverty, unemployment, and a shortage of economic opportunities (Malik & Awadallah, 2013), as well as economic factors, such as economic stagnation (Grinin & Korotayev, 2022) and food insecurity (Soffiantini, 2020). The impact of the Arab Spring on social, economic, and political stability was very much country specific and dependent on factors like regime type, economic infrastructures, demographic characteristics, and external interventions (Korotayev et al. 2014; Korotayev et al. 2022). The democratic process was thwarted in Egypt after a military coup in mid-2013, while Libya, Syria, and Yemen encountered civil war (Brownlee et al. 2015). Only the uprising paved the way for democracy in Tunisia. In most cases, the revolutions termed Arab uprisings failed to resolve internal conflicts and instead led to complex political situations (Haynes, 2020). Furthermore, the social media-led movement served as a catalyst for a worldwide surge in sociopolitical destabilization (Korotayev et al. 2022). The Arab Spring has been described as a contributing factor of terrorism, hampering peace and development (Fraihat & Yaseen, 2020; Schumacher & Schraeder, 2021; Issaev et al. 2021). In addition, the movement is associated with a reduction in economic growth, gross domestic product, foreign direct investment, and tourism income (Echevarría & García-Enríquez, 2020). Conflict-affected countries that seen the Arab uprisings have consistently experienced inflation of approximately 10% (AlShammari et al. 2023). The Arab Spring caused the food security in conflict-affected countries like Yemen to deteriorate, as evidenced by the fact that twenty million people in Yemen became food insecure and 10 million people experienced famine (Basha, 2023). However, no study has examined the causal impact of the Arab Spring on food prices. While acknowledging the impact of Arab uprisings to be country specific, we examine the impact of the uprisings on food prices, taking Yemen as a case study. In addition, we define Arab Spring in Yemen as a series of events resulting from the revolution after January 2011, not as a single event that occurred in January 2011. We include the January 2011 revolution, the resignation of President Saleh (November 2011), the rise of the Houthis (2014), and the Saudi-led intervention (2015) under the definition of Arab Spring-triggered events.

Food inflation can be triggered through diverse channels resulting from the Arab Spring. After the uprisings, Yemen experienced food supply shock which is considered as the main determining factor of food inflation (Kuma & Gata, 2023). Disruptions to agricultural supply chains resulting from political unrest can be termed as a determining factor of food supply shock. Moreover, low productivity is a key feature of Yemen agriculture. Arab uprisings caused unstable political setting which hampered agriculture extension. Moreover, Le Mouël (2023) suggests that the extreme dependence on imports for food supply accelerated the food inflation in Yemen. After Arab Spring, Yemen experienced Saudi-Led intervention which hampered food imports significantly. Moreover, Yemen experienced significant currency devaluation which was a major source of food inflation (Kadri, 2023). Currency devaluation influences food inflation through increasing import costs (Baek & Koo, 2010). Moreover, the Yemen government reduced subsidies on essential goods which can be another source of food price hike (Breisinger et al., 2011). Finally, trade disruptions were a major source of food inflation (Soffiantini, 2020). Yemen experienced trade disruptions as a result of widespread violence after Arab Spring.

The economic cost of the Arab Spring has yet to be estimated. This study aims to estimate the economic cost of the uprisings in terms of food inflation. We consider Yemen which experienced the Arab Spring since January 2011, and define the Arab Spring as the series of events that the country experienced after the movements starting in January 2011. The difference-in-differences (DiD) approach is applied as an identification strategy to address selection bias. In this study, 69 districts of Yemen are considered treated districts. The control districts are taken from 4 countries, Ethiopia, Chad, Uganda, and Zambia. The control countries are selected based on political stability in 2011 and data availability. Data availability refers to whether data are available in our dataset from the World Bank microdata. The countries for which data are available, and which were politically stable in 2011 include Chad, Ethiopia, Zambia, and Uganda. Table 1 shows how and why the countries are selected.

Table 1 Rationale for selecting control countries.
Full size table

The choice of these four countries as the control group, Ethiopia, Chad, Uganda, and Zambia, was based on the similarity of economic conditions of these countries with that of Yemen. All these countries are listed as Least Developed Countries (LDC) by the United Nations (Tekin, 2012). The economies of the four control countries are agriculture based, similar to Yemen. Around 60% of the population in Ethiopia, Chad, Uganda, and Zambia depend on subsistence farming while around 50% of Yemen’s population depends on farming (Varisco, 2018). The agricultures of these countries are characterized by low agricultural productivity (Ajl, 2018; Telleria & Aw-Hassan, 2011). Industries are under-developed in all these countries (Badeeb & Lean, 2017). Besides these countries are dependent on foreign aid and remittances. Moreover, these countries encounter unemployment and high poverty (Breisinger et al. 2011). Low export diversification is another common feature of all these countries (Matallah, 2020). In addition, infrastructure and institutions are weak in all these countries. Therefore, this study considers the four countries as control group to Yemen in economic terms.

Food prices are based on the Food Price Index, with January 2010 as the base month. The main research question of the study is as following:

RQ: What is the impact of the Yemeni Arab Spring on food prices in Yemen?

This research is grounded in Goldstone et al.‘s (2022) Phenomenon and Theories of Revolutions, which posits that a revolution represents a distinct mechanism for altering governmental or political (or even social and political) regimes. However, the revolutionary upheaval of political regimes typically results in a significant breakdown in the system’s functionality, often leading to unpredictable outcomes. Consequently, revolutions have been recognized as highly disruptive and potentially devastating approaches to social change and sometimes become counterproductive forces against progress (Goldstone et al. 2022). Goldstone et al. (2022) provided a framework that captures both the causes and effects of revolutions. The present study focuses on the effect side of revolutions. The theory suggests that any political change on a regional or national scale, all the more changes leading to a revolution, generally contributes to increasing crisis phenomena in the world. Goldstone et al. (2022) argue that even successful revolutions result in inflation due to turmoil, such as supply chain disruptions. The framework discusses the causes of inflation. Goldstone et al. (2022) suggest that revolutions evoke increased demand while economic production remains unchanged, resulting in high inflation. The framework underscores that revolutionary events can go on for a long time (years) in societies engulfed by the wave, through a succession of wars and coups. The framework argues that fragile states face economic crisis after revolution. In this study, we examine the empirical evidence from the perspective of the Phenomenon and Theories of Revolution and explore how the Arab Spring affected the economic factor of food inflation. By examining the associations between political conflicts and food security dynamics, we provide useful policy implications for relevant stakeholders.

Background

Historical background

Yemen experienced the Arab Spring as a form of pro-democratic protest against the three-decade rule of President Ali Abdullah Saleh (Issaev et al. 2022). The Yemeni people started the movement with the hope of political reformation, the end of corruption, and greater economic potential. The basic causes of the movement can be attributed to the high unemployment rate, food insecurity, and the scarcity of social services (Abdulkader, 2023). Youths were persistently at the center of the Yemeni Arab Spring, and social media was used as a tool to organize the movement and achieve mobilization. The Yemeni government dealt with the movement with an iron hand and suppression. As a result, the movement turned into long-term, violent local conflicts between protesters and security forces. The attempts of the government to suppress the movement aggravated the situation, and a new group of military and tribal actors emerged in opposition to the government. The fragmentation of the elite society resulted in the weakening of the Yemeni government (Juneau, 2013). In November 2011, President Saleh was forced to accept an agreement for a peaceful transition of power to Vice President Abd-Rabbu Mansour Hadi (Durac, 2012). Although the transition process was intended to be smooth, Yemen faced diverse challenges, including ongoing protests, the presence of Al-Qaeda in the Arabian Peninsula (AQAP), and internal conflicts. External actors like Iran and Saudi Arabia further complicated the situation in Yemen. The Houthi rebel group, supported by Iran, capitalized on the power vacuum, complicating the political landscape of Yemen. In 2014, Houthi rebels captured the capital, Sana’a, and other key areas, plunging Yemen into a full-blown civil war (Brandt, 2024). The conflict was further aggravated as a Saudi-led coalition intervened in 2015 to restore the internationally recognized government of President Hadi and counter the influence of the Houthi rebels. The Kingdom of Saudi Arabia intervened through a military operation known as ‘Operation Decisive Storm (Al-Hazm),’ triggering further regional and international involvement. The Saudi-led intervention resulted in the displacement of over one million people from Yemen (Al-Hinaiti, 2023). The conflict turned into a proxy war between Saudi Arabia and Iran. The Yemeni civil war resulted in one of the world’s most severe humanitarian crises, marked by widespread displacement, economic collapse, and acute shortages of food and medical supplies (Al Iriani et al. 2023).

Academic background

The causes of Arab Spring are widely studied. The main causes of the uprisings were the existence of internal conflicts within the elite, the diversity of religious and tribal groups, the lack of efficient mechanisms for transferring power, and the absence of safeguards against internal conflicts (Korotayev et al. 2022). Furthermore, Grinin and Korotayev (2022) attributed the Arab Spring to extremist concepts and ideologies in the community, accelerated urbanization, a rising proportion of youth in the population, and the juxtaposition of swiftly advancing education levels for a segment of the populace with inadequate education for others. Food insecurity was one of the key factors of the Arab Spring (Soffiantini, 2020). Political grievances were more related to the Arab uprisings than economic factors (Barakat & Fakih, 2021).

The Arab uprisings have transformed the social, economic, and political systems over the past decade (Haynes, 2020; Fraihat and Yaseen, 2020; Schumacher and Schraeder, 2021). The conflicts associated with the Arab uprisings have not abated in the last ten years, resulting in a complicated political situation (Haynes, 2020). The Arab Spring failed to achieve social change and instead destabilized the Arab world and prevented peace and development in the long run (Fraihat and Yaseen, 2020). Arab Spring is associated with a global surge of terrorism and ultimately destabilized the political conditions of Arab countries (Schumacher and Schraeder, 2021). Similarly, Issaev et al. (2021) argued that the Arab Spring weakened governments, creating an environment suitable for terrorism. Moreover, Aras and Oztig (2021) argued that the Arab movements from 2011 to 2014 failed to bring any political change but instead represented a learning opportunity for authoritarian governments and ultimately led to more repression of citizens and greater control over social media.

Previous studies estimated the economic costs of the Arab Spring. Echevarría and García-Enríquez (2020) argued that the Arab Spring reduced the GDP of Egypt by 12.04%. Kırşanlı (2023) found that the Arab Spring reduced economic growth by approximately 2.98% because of the increase in corruption. Moreover, Jha and Kırşanlı (2023) reported that the Arab Spring further democratized corruption, resulting in increased economic inequality. Ghazalian (2023) argued that the Arab Spring significantly reduced foreign direct investment, and Groizard et al. (2022) found that tourism decreased significantly in countries affected by the Arab Spring. Becheikh (2021) showed that the Arab movements failed to improve economic conditions.

Although previous studies have explored diverse and crucial aspects of the Arab Spring, some unsolved questions remain. We examine the causal impact of the Arab Spring on food inflation as an economic cost of the Arab Spring. We attempt to link the empirical evidence with the theoretical foundation of revolutions.

Materials and methods

Data source and summary

The dataset prepared by Andree et al. (2020) is used in the present study. We consider five countries: Yemen, Ethiopia, Chad, Uganda and Zambia. We use monthly panel data from January 2010 to December 2016. 69 districts in Yemen are included as treated districts, whereas 300 districts in Ethiopia, Mali, Chad, Uganda, and Zambia are included as control districts.

Table 2 presents summary statistics for Yemeni districts, categorized into periods before and after the Arab Spring. The data reveal that the food price index after the Arab Spring was significantly higher—by 0.73 points—than before the Arab Spring. After Arab Spring, Yemen experienced 1.76 numbers of violent events per month, while there were no violent events before Arab Spring. In addition, following the Arab Spring, Yemen experienced an average of 3.04 fatalities from violent events per month, compared to no conflicts before the Arab Spring.

Table 2 Summary statistics (before and after of Arab Spring only Yemen).
Full size table

Table 3 presents summary statistics categorized by whether districts experienced the Arab Spring, with Yemeni districts treated as affected and districts in Ethiopia, Chad, Uganda, and Zambia considered as control units. The data reveal that the treated districts have a significantly higher average food price index than the control districts. Compared to the control districts, the treated districts experienced ~1.33 more violent events on average. Additionally, the treated districts recorded approximately 2.22 more fatalities than the control districts. The average population of the treated districts is significantly lower than that of the control districts. Moreover, since missing variables can bias impact evaluations, it is important to note that the present dataset contains no missing values.

Table 3 Summary Statistics (Treatment vs Control Districts).
Full size table

Identification strategy

Difference in differences with fixed effect

A movement to restore democracy is not an arbitrary event; rather, it depends on the geopolitical context of a country. Therefore, estimating the impact of movements such as the Arab Spring is susceptible to selection bias. We employ the difference-in-differences (DiD) method as the primary identification strategy to address this bias, assuming that there is a parallel trend between treated and control districts (Goodman-Bacon, 2021). The average treatment effect on the treated (ATT) can be estimated by comparing the average change in the outcome of the treated group to that of the control group (Callaway & Sant’Anna, 2021). In essence, DiD compares the treated and control units on the basis of the assumption that the outcome variable should exhibit a parallel trend among them over the specified period; this method serves to mitigate endogeneity issues (Wooldridge, 2023; Ryan et al. 2019). The average treatment effect is estimated via a specified equation:

$$\begin{array}{ll}{{Outcome}}_{{it}}={\alpha }_{0}+{\alpha }_{1}{{Interaction}}_{{it}}+\\\qquad\qquad\qquad+{\beta {District\,Dummy}}_{t}+{{\rm{\mu }}{Month\,Dummy}}_{i}+{u}_{{it}}\end{array}$$

where Outcomeit refers to the log of the food price of district i in month t. Interactioni is the interaction of Post dummyt and Group dummyi. A Post dummyt variable is generated, which takes a value of 1 if the date is after January 2011, so Post=1 if Time ≥ January 2011. The Group dummyi takes a value of 1 if the district belongs to Yemen and 0 if the district belongs to Ethiopia, Chad, Uganda and Zambia. The Interactionit refers to the interaction variable of the Group dummyi and Post dummyt variable. In other words, Interactionit equals to 1 if district i experienced Arab Spring in time t and equals to 0 if district i did not experience Arab Spring in time t. The estimator α1 is the DiD estimation of the average treatment effect. The month fixed effects and district fixed effects are expressed by the coefficients µ and β, respectively, to address potential unobserved heterogeneity on the basis of month and district differences.

Synthetic difference in differences

As a robustness check, we apply synthetic DiD. Synthetic DiD combines the difference-in- differences and synthetic control methods. Synthetic control method creates a counterfactual control group when there is only one treated unit, using weights to pre-treatment period control units (Porreca, 2022). Arkhangelsky et al. (2021) introduced Synthetic DiD, in which treated and untreated units are matched before the treatment trend to reduce the dependence on the parallel trend assumption of DiD.

Synthetic DiD uses the weighting of both pre-treatment time periods and cross-sectional units in the generation of a synthetic counterfactual for causal estimation (Porreca, 2022). The unit-level weight and time weights are added to the synthetic DiD regression. The unit-level weight helps match the treated and untreated units according to the parallel trends before treatment, and the time weight ensures that the pretreatment periods receive greater weight (Pailañir & Clarke, 2023). Synthetic DiD is novel in that it incorporates both time and unit weights to create counterfactual units (Porreca, 2022). The synthetic DiD approach uses the following latent factor model following Porreca (2022):

$$Y=L+{\left(W{\rm{\pi }}\right)}_{i,t}+E$$

Here, Y refers to the outcome variable. Wi, t refers to a binary treatment variable. The treatment variable equals to 1 if unit i is treated in time t and equals to 0 if unit i is not treated in time t. Average Treatment effect on treated is measured as π. L captures both a vector of latent time factors and a vector of latent unit factors. E is the error term.

Two-way fixed effects model

We apply a two-way fixed effects model as a second robustness check. Both month and district fixed effects are included. The two-way fixed effects model accounts for both time and unit-level unobserved heterogeneity. The model used as a two-way fixed effects is given below:

$$\begin{array}{ll}{{Outcome}}_{{it}}={\alpha }_{0}+{\alpha }_{1}{{Treatment}}_{i}+\\\qquad\qquad\qquad+{\alpha }_{2}{{Post}}_{t}+{\beta {DistrictDummy}}_{t}+{\rm{\mu }}\,{{MonthDummy}}_{i}+{u}_{{it}}\end{array}$$

Treatmentit equals to 1 if district i belongs to Yemen and equals to 0 if district i does not belong to Yemen. Postt variable is generated, which takes a value of 1 if the date is after January 2011, so Post = 1 if Time ≥ January 2011.

Results and analysis

Main Results

The main results of the study, shown in Table 4, show that the Arab uprising in Yemen significantly increased food prices. The first column of Table 4, presenting the results of a basic DiD model, shows that the Arab uprising in Yemen increased food prices by 10%. The second column presents the results of the DiD estimation with monthly fixed effects, indicating that the Arab uprising in Yemen increased food prices by 10%. The third column presents the DiD results with both monthly and district-level fixed effects, indicating that the Arab uprising in Yemen increased the food price by 10%. The results consistently show that the Arab uprising in Yemen increased food prices.

Table 4 Average treatment effect of the Arab spring events on food prices.
Full size table

Figure 1 shows the time trend of the food price of both treated and untreated districts. The figure shows that the time trends for the treated and untreated districts suddenly diverged after the Arab Spring started in Yemen in January 2011. After 2012, the food price in Yemen increased radically, as evidenced by the sharp upward trend in the food prices of Yemen in comparison with those of control countries, which did not experience the Arab Spring. The trend of food inflation seems to be parallel before 2011, suggesting common trend assumption of DiD to be satisfied.

Fig. 1
figure 1

Time trend of food price.

Full size image

Furthermore, Fig. 2 shows the food price trend of Yemen over different political phases. The graph shows that the food price indexes of Yemen after January 2011 were in a rising trend. Yemeni Civil War in 2014 further triggered the food inflation. In 2015, Saudi Led intervention did not reduce food inflation trend. We argue that the political events, occurred after Arab Spring, were mediators so the upward trend of Yemen food price can be associated with Arab Spring experienced January 2011.

Fig. 2
figure 2

Food price trends over different political phases in Yemen.

Full size image

Robustness check

Table 5 reports the results of the synthetic DiD estimation, which reveal that the Arab Spring increased the food prices of Yemen by 8%. This estimate reports the average treatment effect on the treated. The results consistently show that the Arab Spring had an aggravating effect on food prices.

Table 5 Impact of the Arab spring on food prices in Yemen.
Full size table

Two-way fixed effect model has been estimated and reported in Table 6. The result suggests that the experience of Arab Spring is associated with food inflation. In other words, the Arab Spring experience is related to 14% increase in food inflation. The result shows that all the estimates of the impact of Arab Spring are consistent.

Table 6 Fixed Effect Estimation.
Full size table

Discussion

The study findings consistently reveal that the Arab Spring and the associated events significantly increased food prices in Yemen. This study provides empirical evidence of the devastating impact of the revolution, while acknowledging the impact is very much country specific. The Arab Spring was intended to restore democracy and economic stability in Yemen, but the revolution worsened the political and economic landscape of Yemen. Food prices were severely impacted in the aftermath of the series of conflicts triggered by the Arab Spring in Yemen.

Food inflation is believed to reduce significantly the welfare of citizens (Alem & Köhlin, 2014) and increase the poverty rate (Dessus et al. 2008); thus, the Arab Spring indirectly reduced welfare in Yemen. Food inflation can be attributed to the social, economic, and political changes resulting from the Arab Spring. The Arab Spring resulted in long-term conflict (Haynes, 2020) and destabilized the political situation in Yemen. Moreover, the Arab Spring enhanced terrorism by weakening governments (Schumacher & Schraeder, 2021; Issaev et al. 2021). No improvement in political change was made through the Arab Spring (Aras & Oztig, 2021). Increasing conflict, political instability, the emergence of terrorism and weakening government capacity characterized the political context and caused food prices to rise.

The channels through which the revolution like Yemeni Arab Spring can lead to food inflation can be multifarious. First, the revolution in Yemen disrupted agricultural supply chains, worsening food inflation. Low productivity of agriculture sector in Yemen is a major source of food inflation (Basha, 2023). Food inflation in Yemen can be attributed to local supply shocks from Arab Spring. (Kuma & Gata, 2023). Second, the revolution in Yemen led to currency devaluation, increasing import cost significantly. Le Mouël (2023) suggests that the extreme dependence on import for food supply accelerated the food inflation in Yemen as the global food price increased significantly. Moreover, currency devaluation resulting from the Saudi-led blockade following the Arab Spring is another key factor contributing to food inflation (Kadri, 2023). Kadri (2023) suggests that Yemen’s dependence on dollar-denominated transactions is a key factor of food inflation in Yemen. Moreover, Favari et al. (2021) argue that food inflation in Yemen was triggered by currency devaluation. Third, Yemen experienced continuous conflicts increasing government expenditure on military forces and reducing government subsidies on essential goods. The reduction of subsidies on essential goods may also contribute to rising food prices (Breisinger et al., 2011). Finally, the violence became widespread, having a significant negative impact on trade routes. The Arab Spring disrupted trade, which is considered a major cause of food inflation (Soffiantini, 2020). Food inflation can be attributed to the fact that the Arab Spring significantly reduced GDP, FDI and tourism (Echevarría & García-Enríquez, 2020; Ghazalian, 2023; Groizard et al. 2022), hampering growth and development (Fraihat & Yaseen, 2020). Furthermore, food inflation is said to be triggered by supply shocks resulting from the Arab uprisings in Yemen (Bandara, 2013).

The Arab Spring may have had heterogeneous impacts on food prices based on the context of a country (Inbar, 2012). This study reveals that Yemen experienced food inflation as a result of the Arab Spring. In other words, the Arab Spring led to events that triggered food inflation in Yemen. Our findings are consistent with those of Echevarría and García-Enríquez (2020), who argued that the Arab Spring affected economies negatively. The increase in food prices is expected to deteriorate food security in Yemen and reflects the devastating impact of the Arab Spring in the long run.

The present study aligns with the Goldstone et al. (2022) Revolutionary Theory Framework. The findings of the present study are consistent with the argument by Goldstone et al. (2022) that even successful revolutions create turmoil such as supply disruptions which result in inflation. The food inflation in Yemen, resulting from Arab Spring, can be attributed to the fact that revolution increases demands without increasing production. Furthermore, the impact of a revolutionary event is dependent on the context of the country facing revolution. Countries like Yemen which are fragile in terms of political and economic conditions face economic crisis after revolution (Goldstone et al. 2022).

Conclusion

This study provides policy implications for relevant stakeholders, suggesting that revolutions ultimately deteriorate food security conditions. An integrated approach needs to be adopted for a conflict-affected country to control food inflation. Developing countries should reduce import dependence in post conflict time. Improving agricultural production can help reduce dependence on imports. In addition, food supply chain needs to be improved to reduce food supply shock. Export diversification should be promoted in developing countries to control currency devaluation.

This study has several limitations that need to be considered when interpreting the findings. First, Yemen is taken as a case study, but the Arab Spring may have had diverse impacts across countries depending on the national context. Second, Ethiopia, Chad, Uganda, and Zambia are selected as control countries because they are comparable states and were not affected by the Arab Spring. To address this issue, we use both Synthetic DiD and a two-way fixed effects model, identification strategies that attenuate the requirement of parallel trends. Future research could explore the impacts of the Arab Spring on food inflation across different countries and on diverse economic outcomes.