2011 was not the end, but the beginning, and maybe it wasn’t even that. Perhaps the deep crisis of the traditional social contract that underpinned most of the MENA region’s ruling regimes since independence had begun even before. The first cracks could be already be heard by sensitive ears in the 2000s, when the liberal reforms introduced by several Arab governments failed to address the increasing needs (and new expectations) of fast-growing populations and the first, scattered protests began to pop-up. What changed in 2011 was simply the fact that this time the cracks became audible to everyone in the world.
Despite all the efforts of many Arab regimes to reestablish and stabilize the pre-2011 authoritarian status quo over the last decade, successive waves of protests have kept erupting in numerous countries, from Algeria to Lebanon, and Iraq. North Africa saw the end of two decades-long presidencies in 2019 alone – that of Omar Al-Bashir in Sudan and that of Bouteflika in Algeria. In the meantime, despite the attempts of other regimes to keep a façade of normality and stability, new demonstrations have been recorded also in Morocco and in Egypt, where the Al-Sisi government’s applies levels of repression that have even surpassed those of his authoritarian predecessor Hosni Mubarak. Even Tunisia, the only country that came out from the 2011 upheavals with a full-fledged, although fragile, democracy, has seen continuous high levels of political contention.
Such trends can be observed clearly by analyzing the data provided by the successive waves of the Arab Barometer surveys – Wave III (2012), Wave IV (2016-2017), and Wave V (2018-2019). In Figure 1 we can see the percentage of the people interviewed in four North African countries – Morocco, Algeria, Tunisia, and Egypt – that affirmed of having participated at least once in a demonstration over the previous three years (in Wave III, which was carried out in 2012, people were instead asked whether they had participated in the 2011 Arab Spring).
Figure 1 – Percentage of interviewees who affirmed of having taken part in at least one protest over the previous 3 years (in the case of 2011, the interviewees were specifically asked whether they took part in the Arab Spring protests)
Apart from Tunisia – where the level of participation, although still high, has been decreasing – all other countries have seen a dramatic increase in the number of people engaged in contentious actions since 2011. Even in Egypt, which recorded a large contraction in political demonstrators between the Arab spring, and the 2016/2017 wave, explained by the draconian and repressive measures employed by the Al-Sisi regime after its establishment, saw its participation increase afterwards. Several years later the percentage of people, which claim to have engaged in protest, had even surpassed the levels seen during the 2011 upheaval.
The increase in participation is not the only element that has been evolving after the so-called Arab Spring. By analyzing the data provided by the Arab Barometer from Morocco, Algeria, Tunisia, and Egypt, it is also possible to depict the socioeconomic profiles of protest participants and how they have evolved in the last decade.
For example, by running logistic regressions on the Arab Barometer data (using participation in protest as the dependent variable, and the main demographic, and economic indicators as independent variables), it is possible to compare the main determinants of participation in 2011 and 2019. Table 1 shows the relationship between participation and the urban/rural residency of the interviewees, their gender, age, education level, religiosity and income in 2011 using the aggregated data from all four countries. In Table 2 it is possible to see the data for the 2018-2019 wave.
Table 1 – Socioeconomic determinants of participation in 2011 (aggregated data from Morocco, Algeria, Tunisia, and Egypt)
Table 2 – Socioeconomic determinants of participation in 2018-2019 (aggregated data from Morocco, Algeria, Tunisia, and Egypt)
At the aggregated level, it is possible to draw a first profile of the typical North African protester in both waves. He/she is more likely to live in urban areas, be male, younger, more educated and less religious than the average, and to belong to the upper (richer) part of the distribution of income.
Some of these findings fit the general mainstream descriptions provided by international media: the data confirm that generally protesters are younger, prevalently from urban contexts, and more educated than the average. However, they also disconfirm in part some other established stereotypes: for instance, many have described different levels of religiosity among participants as one of the main differences between the protests in 2011 and those in 2018-2019. They based this consideration on the fact that in 2011 Islamist movements emerged as the (temporary) political winners of the upheavals in most countries (the two most notable examples being Tunisia and Egypt). However, this assumption was not borne out in the data, which indicate that participants in both waves tended to be less religious than the average. Moreover, despite mainstream narratives which depict protesters as mostly belonging to the lower socioeconomic strata of their countries’ populations, the data show instead that they are more likely to belong to the upper part of the income distribution.
The outcomes at the aggregated level are not always confirmed when data are disaggregated for each country. Tables 3, 4, 5, and 6 show the results of the regressions for each country using the data from the 2012 wave, while Tables 7, 8, 9, and 10 show the results using the data from the 2018-2019 one.
Table 3 – Protest participation determinants for Algeria in 2011
Table 4 - Protest participation determinants for Morocco in 2011
Table 5 - Protest participation determinants for Tunisia in 2011
Table 6 - Protest participation determinants for Egypt in 2011
Table 7 - Protest participation determinants for Algeria in 2018-2019
Table 8 - Protest participation determinants for Morocco in 2018-2019
Table 9 - Protest participation determinants for Tunisia in 2018-2019
Table 10 - Protest participation determinants for Egypt in 2018-2019
The analysis of each country’s results shows some notable differences with the aggregated outcome. For instance, Algeria shows a countertrend regarding the participants’ incomes in the 2018-2019 wave: contrary to what was observed at the aggregated level, in Algeria protesters seem to be poorer than the average. In Morocco, the most notable difference between 2011 and 2018-2019 is the fact that in the latter the urban-rural divide is no longer a determinant of participation (while in 2011 protesters were more likely to be from urban areas). Even in the case of Egypt we can observe some notable evolution: in 2011 income showed a positive relation with participation (meaning that richer people were more likely to protest) and age showed no significant influence. On the opposite, in the 2018-2019 age emerged as a negative determinant (meaning that younger people were more likely to protest) and income lost its significance. Only Tunisia shows rather similar results in both waves, probably signaling the persistence of the same set of social and economic problems.
In sum, by analyzing the data provided by the Arab Barometer over the last decade it is possible to affirm that protest participation is decisively on the rise in all the North African countries under consideration. It is also possible to observe some interesting evolutions occurred in the socioeconomic determinants of participation for each society.
The 2011 Arab Spring was not an isolated episode, nor the end of the story. It was the next evolution in a story we have been watching unfold for almost a decade, and one that seems likely to bring more political change and instability over the next years.
 For a more detailed description of the methodology employed see this similar analysis written by the author.
 The stars beside each value indicate the level if statistic significance of the results. No star means that they are not significant (beneath the 95% confidence interval). One * represent significance above the 95% interval, two ** mean significance above the 99% interval and three *** represent significance above the 99.9% interval.
 To measure the respondents’ level of religiosity I used, as proxy, the question “do you pray daily?”, whose possible answers are “Always”, “Most of the times”, “Sometimes”, “Rarely”, and “Never”.
 Income was normalized to be compared across the three waves. In the third wave respondents were asked to indicate their precise monthly income. However, in the fourth and fifth waves they were simply asked to indicate weather their income was above of beneath the population’s median. Hence, I transformed the income variable of the third wave accordingly into a binary variable indicating weather the respondent’s income is above of beneath the median calculated on the income figures indicated in the original variable.