## Introduction

Wheat plays an important role in the Egyptian national diet. It is one of the primary food sources in Egypt, next to maize and rice (FAO 2014). Egyptians get more than one-third of their daily calories and 45% of their daily protein from wheat, mainly in the form of bread (IFPRI 2000). Wheat is a major domestically-produced as well as traded food crop in Egypt. However, Egyptian wheat production is far from sufficient to satisfy the domestic demand for wheat. The most limiting factors for wheat production in Egypt are arable land and capital, and to a lesser extent, water and labour (Wichelns 2001). Its population was estimated to be 92 million people (World Bank 2016a). The country continues to have a high population growth rate of 2.7% annually (World Bank 2016b), so that it grows by about 2 million persons per year, which represents a large, sustained demand-driving factor. Additionally, Egypt is the most populated country of the Middle East and North Africa (MENA) region. Therefore, 45–55% of the wheat consumed has to be imported, which makes Egypt the world’s largest wheat importer (FAO 2014; Woertz and Keulertz 2015). The dependence on wheat imports makes Egypt vulnerable to increasing food prices or other market shocks to the world market which might impede its import needs.

Egypt is classified as a middle income country, but it still faces development challenges due to regional inequalities. The country can be divided into two parts: Lower Egypt and Upper Egypt, representing the North and the South, respectively. In Lower Egypt, the river Nile stretches out while Upper Egypt consists of mainly rural areas and desert. Upper Egypt is ranked lower socio-economically than the country average. Therefore, in these areas the World Food Programme provides assistance to improve food security, nutritional status and access to socio-economic opportunities for the most vulnerable parts of the population, mainly for women and children (WFP 2014). However, food insecurity also increased in Lower Egypt between 2009 and 2011, which shows that food insecurity is not only a rural issue anymore (WFP 2013). In Lower Egypt, Greater Cairo has the largest population with poor access to food (WFP 2013). From 2005 on, food security in Egypt became worse as a consequence of several crises (an avian influenza epidemic in 2006; food, fuel and financial crises in 2008 and rising global food prices in late 2010) and increasing poverty (IFPRI-WFP 2013). The Egyptian government provides a food subsidy programme for the most vulnerable part of the population to protect them from rising food prices.

On January 25, 2011, the major political instability in the recent history of the MENA, the ‘Arab Spring’, arrived in Egypt with uprisings spilling over from the neighbouring countries of Algeria and Tunisia (for economic details, see, for example, World Bank 2015). Consequently, Egypt entered into a politically unstable situation. Protests showed dissatisfaction about the regime of President Hosni Mubarak in different cities across Egypt. Demonstrations, protests and riots continued until February 11, when Mubarak decided to resign. The political situation of Egypt remained unstable throughout 2011 experiencing frequent cases of arrests, violent clashes, shootings and other violence from both the government and protesters. By the end of the year, the political instability had caused at least 850 fatalities in the country (BBC 2011). Such violent political instability can have manifold negative effects on the economic performance of a country and the economic activity of its citizens (Blattman and Miguel 2010). It is harmful for total productivity growth and discourages physical and human capital accumulation (Aisen and José Veiga 2013). As a consequence of the political unrest, consumers and institutions relevant to trade suddenly faced extremely high uncertainty. Trading partners who previously exported wheat to Egypt were less certain about receiving payments. This might have led to increasing prices of imports and a devaluation of the Egyptian Pound, which would have made imports more expensive. At the same time, a fall in consumer confidence might have occurred. People may have spent less money and saved a larger part of their income. As aggregate demand is determined by the money supply, the velocity of money, price level and output (Mankiw and Taylor 2008) and import demand might have decreased when prices increased and money supply decreased. This fall in aggregate demand might have led to a negative shock to wheat imports into Egypt.

The analysis of food security and vulnerability in the MENA has attracted considerable attention in recent years. However, scientific evidence on the economic and food security dimensions of the Arab Spring is scarce although this uprising had substantial and lasting effects for large parts of the population in the region. El-Dukheri et al. (2011) analyzed the farmers’ response to high food prices before the Arab Spring. They found several countries whose domestic markets were strongly affected by the global price rises but also identified several limiting factors. Wright and Cafiero (2011) analysed the response of governments in the MENA to the global food price crisis in 2007/2008. They called for an increase in stocks instead of pursuing higher degrees of food self-sufficiency. They stressed the disadvantage of substantial subsidies for food consumption, since these subsidies distort domestic price stabilization. Khouri et al. (2011) reviewed lessons learnt from this price crisis for the MENA and discuss several policy options. Lampietti et al. (2011) suggested three strategies that would be feasible and effective in the MENA for reducing the region’s vulnerability to price shocks which are the extension of domestic safety nets, the increase of domestic food production and the reduction of consumers’ exposure to food price volatility.

In this paper, we extend the analysis of vulnerability and food security in the MENA by assessing the effect of political instability on international trade. To our knowledge, it is the first paper which analyses trade-relevant externalities of the Arab Spring. This analysis is therefore a first step towards a comprehensive analysis of the economic and food security dimensions of the Arab Spring which has been the most comprehensive and most far-reaching wave of political destabilisation experienced by the MENA in recent decades. We analysed monthly wheat trade data from the Comtrade Database of the United Nations (UN Comtrade 2014) from 2010 to 2014. We statistically tested if the level of wheat imports in the months immediately after the outbreak of the Arab Spring in Egypt significantly differed from the seasonal import pattern in normal years.

The remainder of this article is structured as follows: section 2 describes the role of wheat in the Egyptian economy and policy making. The Arab spring in Egypt and the vulnerability of the Egyptian population are reported in sections 3 and 4, respectively. The next two sections give a theoretical background and describe data and methods. Results are given in section 7 and are discussed in section 8.

## The role of wheat in the Egyptian economy and policy making

Average wheat consumption per capita at the global level is about 100 kg per year, but in Egypt the amount is substantially higher at about 225 kg per capita per year (USDA 2014). Since baladi breadFootnote 1 is the cheapest food due to high subsidies, Egyptian consumers have high incentives to take advantage of consuming it (Mansour 2012). Baladi bread and wheat flour are available to all consumers in Egypt, without restrictions and regardless of income level (Ramadan and Thomas 2011). Figure 1 shows the difference between the production and consumption of wheat in Egypt from 2009 until 2015.

Agricultural reforms in 1987 encouraged Egyptian farmers to produce wheat. In the following years, crop yields improved due to several factors such as higher relative wheat prices, better rotation of crops, use of seed varieties that are more resistant to heat, drought and salinity, implementation of more productive cultural practices and a more liberal policy that encouraged producers to invest in modern technology (IFPRI 2000). During the period from 2004 to 2013, grain production went up from approximately 6.8 million t in 2004 to 8.5 million t in 2013 (International Grains Council 2014). However, the domestic production of wheat is, by far, insufficient to satisfy the demand, which is almost 20 million t (Fig. 1). The most limiting factors are scarcity of agricultural land and water (Solieman et al. 2015). Other factors that limit domestic production of wheat are high costs of seeds, pesticides and fertilizers (Mansour 2012). Fluctuating prices of these input factors discourage farmers from investing in new technology and in improving production practices. Egypt has a shortage of high yielding wheat seed because of a shortage in resources and land on which to produce seeds (Mansour 2012).

Importing wheat to meet the excess demand makes Egypt highly dependent on other countries and responsive to world market developments. Egypt imported 54% of its wheat in 2009 (Ramadan and Thomas 2011). The main wheat trading partners of Egypt are the Ukraine, France and Romania. Together they provide 52% of the total wheat imported by Egypt (UN Comtrade 2014). Greece is the source of 98% of all wheat flour imported by Egypt.

Egypt knows a long history of subsidized food. Currently, baladi bread, wheat flour, sugar and cooking oil are the principal and most highly subsidized food commodities. The availability of cheap bread and wheat flour for every household in combination with the high population growth rate creates a huge and continuously rising demand for wheat. Consequently, Egypt has the highest wheat consumption level per capita in the world. Since Egyptian consumption of wheat is about twice as much as its production, Egypt is highly dependent on the world market for wheat imports.

## The Arab spring in Egypt

Figure 2 shows that domestic food prices started to rise significantly after 2004. The food price volatility index is calculated as the standard deviation of the deviations from the trend over the previous 5 years (FAO 2013). As Egypt’s dependence on wheat imports makes the country vulnerable to increasing food prices in the world market, the global price rises of 2007/2008 are also visible in the domestic food price index. This led to minor protests and strikes in 2007/2008, which called for more than just a reduction of food prices. Bush (2010) stressed that the political movement united a number of previously divided forces, which jointly and publicly expressed their critique of the president, widespread corruption in the government and brutality of security forces. Between 2010 and 2012, the currency exchange rate gradually decreased (OANDA 2015), domestic food prices were additionally driven up and food price inflation accelerated.

According to Beissinger et al. (2012), “the Arab Spring revolutions were massive political upheavals, with citizens in multiple countries taking to the streets against their respective regimes”. The Arab Spring started in Tunisia and spread to almost all MENA countries. This political movement had initially no features of what is specifically characterized as ‘Arab’ or ‘Muslim’ (Roy 2012). The demonstrators were mainly concerned about their individual citizenship and not about religion or the geopolitical conflict in the Middle East between the Israelis and the Palestinians. They called for “dignity, elections, democracy, good governance and human rights” (Roy 2012).

Egypt has been ruled by former military officers since 1952 (Goldstone 2011). Before the revolution of 2011, three presidents ruled the country: Gamal Abdul Nasser (1952–1970), Anwar Sadat (1970–1981) and Hosni Mubarak (1981–2011). During the period of President Sadat, the foreign debt increased from US$1.3 billion to US$19.5 billion, mainly due to financing domestic consumption paired with its quickly growing population. Therefore, Mubarak was forced to start with improving the economic situation of Egypt when he came into power (Kandil 2012). At the beginning of his presidency, Mubarak also took some steps to make governance more democratic. In 1981, a state of emergency was proclaimed in Egypt. The Emergency Law was prolonged every 3 years by the parliament. In 1992, the Anti-Terror Law came into play as an addition to the Emergency Law. As a result, authorities acquired more power such that they could limit the basic freedoms of Egyptians. There was no (fair) trial for suspects. Strikes, demonstrations and public meetings were prohibited. Newspapers were censored or closed down. Civil society organisations were impeded from organising protests. In 2004 the start of the first wave of protests was marked by the establishment of the Egyptian Movement for Change (Kifaya). In the collective protest movements, three waves of protests can be distinguished. The first was mainly against corruption. And the second was focused on improving living conditions, such as higher wages and better working conditions. In 2008, the third wave of protests was launched in another form. Young people started a political ‘Facebook’ group and mobilized more than 70,000 members. Members debated mainly about concerns regarding free speech, nepotism in the government and the country’s economy. They also suggested new ideas for reform in Egypt (Hassan 2011).

The literature discusses several economic, social and political factors which decisively contributed to the outbreak of the Arab Spring. Between 2002 and 2008, the gross domestic product (GDP) grew each year by an increasing percentage. After 2008, GDP growth dropped from 7.2 to 4.7% in 2009 (Worldbank 2014). This economic stagnation could have led to dissatisfaction and possibly contributed to the revolution. It seems implausible that unemployment contributed to the revolution, since total unemployment as a percentage of the total labour force decreased in the years prior to the Arab Spring (Worldbank 2014). However, youth unemployment was high. In 2000 and 2011, youth unemployment was more than 25 and 35%, respectively (Worldbank 2014). Youssef (2011) argues that the revolution was primarily created and owned by the country‘s youth. Frustrations stemming from bad employment prospects might have been another important reason for the outbreak of the Arab Spring.

After 18 days of increasingly violent protests, President Hosni Mubarak resigned on February 11, 2011. From April to August, protests took place against the slow pace of political change. In November, people protested against the military, which was still in power. In that month the parliamentary elections started. The Muslim Brotherhood won the parliamentary elections and Mohammed Morsi was nominated for the presidential elections in May 2012. Also in May, the state of emergency was ended. Mohammed Morsi won the presidential election in June, and he became the new president of Egypt. The Morsi government quickly restricted the rule of law and political freedoms, yet they did nothing to tackle Egypt’s economic decline (Beissinger et al. 2012). Two years after the protests against Mubarak, Egyptians started to protest against Morsi. Finally, the military intervened and removed President Morsi from office on July 3, 2013. Interim President Mansour announced that new presidential elections would be held before the parliamentary elections. In May 2014, the former chief of the Egyptian army, Abdul Fattah al-Sisi, was elected as the new president.

## Vulnerability of the Egyptian population

The human population pyramid of Egypt in 2010 shows that the Egyptian population is very young: 51.2% of the population is under 24 years of age, 43.3% of the population is aged between 24 and 64 and only 5.5% is older than 65 (UNDES 2012). Over the past three decades, Egypt’s population has grown from 45 to 85 million, which explains the so-called demographic ‘youth bulge’. According to Khalil (2011), two phenomena are connected with this youth bulge. Since Mubarak came into power, both corruption and youth unemployment reached very high levels. This combination led to despair and frustration among young graduated Egyptians. Egyptian society was neither able to provide millions of young graduates job opportunities (except for a small elite) nor prospects for a future. Despite the repressive regime, young Egyptians took another course. They used modern communication technology such as cell phones and the internet in a peaceful way to express their frustrations (Khalil 2011). Although the revolution initially was started by the young generation, people of all ages and the rich, poor and middleclass, educated or not, were involved in the revolution. They were united in their desire for change and an end to oppression (Khalil 2011).

Gini coefficients of Egypt are rather similar to Gini coefficients of western countries (World Bank 2015). Gini indices are not available for each year. The most recent numbers are 30.77 in 2008, 32.14 in 2005 and 32.76 in 2000. However, poverty headcount ratios are 15.43 in 2008, 18.46 in 2005 and 19.37 in 2000. This shows that at least 15% of the Egyptian population is living on less than US$2 per day (World Bank 2015). According to Amnesty International (2011), around 32 million of the 80 million Egyptians were living on or near the poverty line of US$2 a day in 2009 (see also IFPRI 2014). The Human Development Index (HDI) is an average of three indices: a long and healthy life, access to knowledge and a good standard of living (UNDP 2013). Egypt’s HDI value increased from 0.452 in 1980 to 0.682 in 2013, placing Egypt within the medium human development category (UNDP 2014). The HDI does not take into account inequality in the distribution of human development across a population at the country level. Therefore, the Inequality-Adjusted HDI has been introduced. When all three dimensions are discounted for inequality in distribution, the HDI for 2013 drops from 0.682 to 0.518 (UNDP 2014). Between 2009 and 2011, the number of Egyptians that were food insecure rose by 21% (IFPRI-WFP 2013). Figure 3 shows that Eygpt faces a serious risk of food insecurity.

According to Transparency International (2014), Egypt is less corrupt than most other African countries. It seems that unique circumstances led to the outcome of the revolution in Egypt. The permanent state of emergency, population demography, youth unemployment and corruption together all played a role in the revolt.

## Theoretical framework

Violent political conflict is likely to create negative externalities both for demand and supply in a national economy. It induces high levels of uncertainty and impedes the proper functioning of production and trade-related institutions, which may lead to distrust or higher costs. The responses of institutions, the private sector and individual actors to this uncertainty may cause significant impediments to economic activity. This was also very likely to be the case in Egypt in 2011 due to severe political destabilization and the general institutional chaos the country experienced then.

Several studies show a link between violent political instability and negative effects on national savings, investments, income and growth (Blattman and Miguel 2010). According to Aisen and José Veiga (2013), political instability can affect economic growth in three ways. It can negatively affect the accumulation of physical capital, lower technological progress and have a negative effect on human capital accumulation. Growth in physical capital and human capital is dependent on investments. Uncertainty causes people to invest less, since future returns are expected to be lower or unpredictable. Physical capital accumulation and human capital accumulation may stagnate as a result of declining investments in capital and education. Uncertainty can also slow down technological progress when research and development projects supported by firms and the government are reduced or postponed (Campos and Nugent 2002; Aisen and José Veiga 2013). According to Leduc and Liu (2014), an increase in uncertainty acts like a negative aggregate demandFootnote 2 shock. It raises unemployment and lowers inflation at the same time. Uncertainty increases the real option value of waiting (Dixit and Pindyck 1994; Lambarraa et al. 2016). Firms will temporarily pause their investment and hiring. This decrease in hiring and investment causes a reduction in the reallocation of production factors from low to high productivity firms. As a result, productivity growth falls (Bloom 2009). The effects are likely to be visible in the performance of private companies as well as public institutions carrying out the imports of food into Egypt from the world market.

Trading countries that export wheat to Egypt may be less certain about receiving payments. Therefore, import prices increase and the Egyptian currency depreciates, which makes imports more expensive. At the same time, consumer confidence is likely to deteriorate. People start to spend less money and save a larger part of their income. According to Mankiw and Taylor (2008) aggregate demand within an economy is determined by the money supply M, the velocity of money V, price level P and total output Y. The following eq. M/P = kY with k = 1/V shows that money supply divided by price level equals the velocity of money times the output. This shows that a price increase and reduction in money supply lower aggregate demand. A fall in aggregate demand led to a negative shock for the wheat trade of Egypt.

Conflict can create unintended economic effects for food production and trade. Ihle and Rubin (2013) stress that military or security policies have economic side effects on food markets: including more volatile prices, commodity gluts or shortages, unemployment and higher transaction costs. According to Anderson and Marcouiller (2002), instability impedes trade by raising the prices of traded goods as a result of increased transaction costs. Transaction costs increase when contracts are poorly enforced by legal systems and if economic policy is unclear and biased. An increase in transaction costs could also have contributed to the fall in wheat demand in Egypt. General chaos in the country may have caused uncertainty among traders. Wheat sellers are less certain if wheat buyers will abide by the contract. In order to create more certainty, contracts should be adjusted. Adjustments of contracts involve higher costs, which can result in lower demand.

Public institutions can become temporarily dysfunctional because agricultural resources or storage facilities are destroyed, specialists of the national food system are removed from their positions or food distribution networks are disrupted causing individuals to be prevented from obtaining food (D’Souza and Joliffe 2013). Probably, the revolution in 2011 disrupted food markets and distribution channels in Egypt. Egypt has limited storage capacity, which could have been full due to lower sales. Difficulties in the storage and marketing of wheat could have contributed to a fall in wheat demand. Blattman and Miguel (2010) point out that conflict can also be a consequence of food insecurity. Terms of trade volatility can stimulate armed conflict in two directions. Increased import prices can trigger conflict by suppressing the real wage. Although consensus exists around the negative impacts of political instability in the short run, there is no consensus in the literature on how long the economic effects persist (Blattman and Miguel 2010). Therefore, temporary food insecurity and vulnerability caused by trade disruptions are likely to be one major socio-economic consequence of violent political conflict. Recent empirical evidence confirms this supposition. Koren and Bagozzi (2016) found a robust association between the degree of food insecurity and armed conflict. Natalini et al. (2017) develop a model for simulating this causal link based on observed data.

## Data and methods

The following empirical data analysis focuses on seasonal patterns of wheat grain and wheat flour imports into Egypt between 2010 and 2014 since no monthly data prior to 2010 are available. Data was obtained from the Comtrade Database of the United Nations (UN Comtrade 2014), which contains detailed merchandise data that is provided by countries and given to the United Nations Statistics Division (UN 2013). We focus on ‘wheat’ and ‘wheat flour’. The database provides detailed information about the trading partners, net weight of wheat grain or wheat flour traded in tonnes per year and its value.

The analysis aims at finding evidence for a significant deviation in wheat imports per half year between 2011 and the other years. For this aim we performed a t-test complemented by a dummy time-series regression. The data is split into the first and second half year of 2011 and the first and second half year of all other years. In this way, the period from July to December 2011 can be compared to the same period in the other years. An independent samples t-test has been carried out. This test assumes the null hypothesis of the average monthly imports during the second half of other years is equal to the average monthly imports of the second half of 2011. Under the null hypothesis, the test statistic is distributed with 22 degrees of freedom.

As a second approach we used a time-series dummy regression model in order to test for differences in the level of imports and exports in the second half of 2011 relative to all other observations. As the emphasis of the model lies in the level of the imports, we were not interested in the autoregressive dynamics, but only in the average trade level per half year. Therefore, the model takes the following form:

$${trade}_m={\beta}_0+{\beta}_1{D}_{2011}{D}_{H2}+{\varepsilon}_m.$$
(1)

The variable trade m is the observation of the m-th month of one of the four trade variables: imports of wheat, exports of wheat, imports of wheat flour or exports of wheat flour. The dummy variable D2011 takes the value 1 if the observation falls into the year 2011 and the dummy variable DH2 takes the value 1 if the observation was made in the second half (that is, in one of the months between July and December) of any of the 4 years. Therefore, the interaction between both dummies, i.e. D2011DH2, marks the observations which have been made in the second half of 2011, which is the period of interest for this analysis. Thus, the two parameters to be estimated have the following interpretation. The parameter β0 denotes the monthly average of the variable trade m of the observations from all months except the ones in the second half of 2011. The parameter β1 denotes the difference between this average and the monthly average of the second half of 2011. The sign of this parameter signals whether the average monthly quantity of trade m in the second half of 2011 was larger or smaller than the average of the other months. If this parameter is significant, then the result signals that there was indeed a statistically significant deviation in the pattern for the given trade variable in the second half of 2011 in comparison to the other months.

## Results

Figure 4 shows a remarkably stable seasonal pattern of wheat import quantities, which has its maximum in the second half of each year. Except for 2011, imports peak around October or November of each year since wheat is grown as a winter crop in Egypt. The crop is sown in November and harvested in May (FAO 2014). During the growing season, Egypt has to import wheat from abroad. In the months following May, when the domestic harvest is obtained, imports decrease. Wheat has to be imported partially because the quality of the wheat produced in Egypt is poorly suited for bread-making. Therefore, domestically produced wheat is mixed with imported wheat to improve the quality (Ahmed et al. 2013). After July, wheat imports increase and peak around October until November. The General Authority for Supplies and Commodities (GASC) is responsible for purchasing wheat on the international market and for its domestic distribution. ‘GASC tries to keep a 5-month supply of strategic wheat stocks’ (Mansour 2012). In recent years, the private sector has also kept at least 2-months of stock, or more (Mansour 2012). The smaller peak in March each year, almost 5 months after the peak in autumn, is likely due to the run-down of the strategic wheat stocks. The visual assessment of Fig. 4 indicates the substantial deviation of the import pattern of wheat in 2011 from the other years with no peak in wheat imports during the autumn of that year.

Imports of wheat flour are much smaller in magnitude and are spread far more throughout the year. Wheat flour is also partly exported to other African countries, such as Madagascar, Burundi and Ethiopia (UN Comtrade 2014). Fig. 5 shows the seasonal pattern of wheat flour imports in t per month. Except for some peaks in November, no distinctive pattern could be found. Remarkably, from February until the middle of May in 2011 no wheat flour was reported to be imported.

For the independent samples t-test, equal variances cannot be assumed since the Levene’s Test is significant. The t-test leads to a test statistic of 4.48, which corresponds to a p-value of less than 0.01. Since this p-value is less than 0.05, we rejected the null hypothesis and obtained strong evidence that the mean monthly import quantity of the second half of other years was not equal to the mean for the second half of 2011.

Table 1 presents the regression results of model (1) for all four trade variables of interest. A first look at the overall model significance shows that only the wheat import model is a meaningful statistical formulation (since it is the only model in which the p-value of the overall model significance test is less than 0.05). This goes along with the finding that in the remaining three regressions β1 is never significant, meaning that it could be dropped. A second look at the economic significance also clearly indicates that only the wheat import model is of an economically relevant magnitude as intercept and the dummy coefficient estimates amount to 486,000 t and −300,000 t, respectively. Wheat imports amounted on average to about 400,000 t per month in 2010 and to about 340,000 t per month in 2011. The models for the other three dependent variables lack statistical significance, and are economically not relevant as the two coefficient estimates of each of them are of negligible magnitude of a few thousand tons or less.

The results of the wheat import model provide the following insights. On average, Egypt imported 485,000 t of wheat per month except for the months in the second half of 2011; 95% of these observations fell into the range of about 385,000 to 585,000 t. The estimate is significant at the 5% level, which indicates that this monthly average is statistically significant. The large magnitude of about half a million t also stresses the economic significance of these imports for the country and confirms the trade analysis above. Of key interest is, however, the magnitude, sign and significance of the parameter β1. The point estimate illustrates that average monthly quantities strongly differed from the reference period during the second half of 2011; almost 300,000 t less than the average were imported in each of these 6 months. The sign is negative, stressing that the imported quantities indeed decreased on average. As the p-value amounts to 0.04, the analysis shows that this negative difference is not only large in economic magnitude, but also significant in statistical terms. This strong evidence for a substantial economic side-effect of the Arab Spring on wheat import quantities appears to be even stronger because the model is estimated by using only a small number of observations (49 observations), which stresses the relevance of the effect found.

## Summary and discussion

This analysis aimed at assessing food-trade related externalities of violent political conflict in Egypt. It examined whether the Arab Spring, the most important phase of major political instability that shook the MENA in recent years, affected imported quantities of a key staple food. The Arab Spring broke out in Egypt in early 2011. It led to the eventual replacement of the Mubarak government and was accompanied by a major shock to political and economic institutions and actors as well as roughly 1000 fatalities.

While imports of wheat grain are of substantial magnitude of several hundreds of thousands of tonnes and show a pronounced seasonal pattern, which is closely related to the domestic cropping season, wheat flour imports do not show such a pattern and are of negligible magnitude. Exports of both commodities are also negligible. The quantity of wheat imports shows a general upward trend between 2010 and 2014. The main trading partners of Egypt for wheat imports are Ukraine, France and Romania. As wheat is a winter crop in Egypt, which is sown in November and harvested in May, imports decrease after May. After July, imports increase and peak around October or November.

To our knowledge, this paper is the first analysis of trade-relevant externalities of the Arab Spring and can, thus, be a first step towards the comprehensive analysis of the economic and food security dimensions of the Arab Spring. The quantitative analysis is based on a dummy time series regression, complemented by a t-test for obtaining statistical evidence on whether the Arab Spring in Egypt indeed impeded wheat imports. Both approaches provide strong evidence that the outbreak of the Arab Spring in Egypt indeed yielded a negative shock to the country’s wheat imports. General chaos during the revolution, causing sudden high uncertainty for consumers and traders, which led to a fall in wheat demand, is likely to be the main cause. This is strong evidence that the Arab Spring led to temporarily impeded wheat imports, which are likely to have temporarily raised food insecurity and vulnerability in Egypt. This insight contributes a valuable case study evaluation of the economic effects of violent political conflict that Blattman and Miguel (2010) called for. Our results complement evidence from the MENA that economic relations through trade have an inherent pacifying potential leading to reconciliation across the hostile parties of political conflicts (see, e.g. Dobers et al. forthcoming).

Although the empirical approaches adopted are quite straightforward to cover such a phenomenon, which might seem very complex on first sight, we believe the empirical strategy serves the purposes of the analysis very well. The research question focussed on assessing whether or not wheat was imported in the second half of 2011 in quantities comparable to other years in which the Arab Spring played no role. Such an analysis can straightforwardly be based on trade flow data. As regards trade flows, the question reduces then to testing with a suitable model whether there was a statistically significant deviation of significant economic magnitude from the otherwise stable annual pattern of Egyptian wheat imports. Institutional reasons causing the identified significant drop in imports might be complex and lie beyond the interest of this analysis since they are also only hardly quantifiable. We hope that further qualitative research will clarify them.