Colonial legacies of uneven state development in MENA

Allison Spencer Hartnett, University of Oxford

*This memo is part of POMEPS Studies 31, Social Policy in the Middle East and North Africa. Download the full PDF here.

Harold Lasswell’s (1936) contention that politics is fundamentally a study of who gets what, when, and how neatly sums up current thinking on social policy in the MENA region. Scholars describe MENA states’ provision of welfare services like health and education primarily in terms of Lasswell’s “who gets what.” Scholars like Baylouny (2010) and Cammett (2014) have advanced our understanding of how welfare provision is targeted in weak or retreating states, but we know less about when and how states develop the capacity to deliver public services. A growing political economy literature suggests that historical legacies of state capacity are a productive point of departure for understanding downstream social policy outcomes.

Scholars of the region often recall the negative effects of European colonialism in the Middle East, but few have empirically studied colonialism’s institutional legacies on the long-run development of state capacity. Independent MENA states inherited bureaucratic and political institutions from their colonial predecessors, and although recent scholarship in other fields has shown that “history matters,” historical legacies and their ramifications for MENA’s present remain poorly understood. I argue that rigorous process tracing and serious study of history allows political scientists to make sense of the region’s variable social policy capabilities in the present.

A large literature in political science and development economics examines the historical origins of the state to explain variation in the generosity and capability of states’ provision of services.[1] In the Middle East context, the colonial period stands out as a critical juncture for institution and state building. Colonial rulers established new central governments that increased day-to-day state-society interactions through new public services like education and health care. Evidence from other colonial contexts shows, however, that colonial institutions were not uniformly distributed across territories. Iyer’s (2010) research on the long-term effects of colonial rule on India finds that areas administered directly by the British have lower access to schools, health centers and roads in the post-colonial period than indirectly ruled regions. This essay builds on Iyer’s work to examine how colonial policies impacted long-term social policy outcomes in Jordan and Iraq. I argue that colonial strategies of state-building and control conditioned the spread of these new institutions, resulting in the uneven spread of health and education services.

Colonial variation in state capacity in (Trans)Jordan and Iraq

In the wake of World War I, the British established Mandate governments in Transjordan and Iraq. A comparison of the colonial moment in both countries shows that British opted for very different strategies in dealing with local elites and incorporating them into the new national state. The British administration in Mandate Iraq was primarily concerned with creating domestic allies who could help them rule the country as local proxies. The preferential treatment and economic elevation of tribal landowners in particular paved the way for uneven state building through the creation of strong, parallel local orders capable of constraining the central state. In Transjordan, however, local elites were brought into a relationship of economic and political dependency on the state and were less able to contest British-led state building reforms.

Administrative reports from the British National Archives highlight Britain’s twin goals of maximizing control and minimizing expenditure in colonial Iraq. Many reports attest to the British using land as a reward for loyal sheikhs and landowners, particularly those in the southeastern province of Amarah.[2] This de facto relationship quickly became de jure through the passage of a series of laws that effectively transferred the control of huge swaths of Iraqi territory to the landed elite and their clients, rather than the state.[3] Sluglett (1976) argues that this informal alliance between the Iraq Government and the landowning class, or mallāks, resulted in the landowners being “left as far as possible to their own devices.” When the British attempted to enact conscription in the 1920s and 1930s, British intelligence attachés in the provinces polled provincial governors about the feasibility of conscripting fellāh or tribesmen. In areas dominated by mallāks, the response was an unequivocal “no.”[4] At one extreme, the British did not bother to expand the state into the tribal district of Muntafiq. They levied minimal taxes and actually reduced the previous administrative and tax collection responsibilities that the Ottoman government had exacted. Areas dominated by the mallāks received fewer services from the new Iraqi state, creating regions that lagged behind non-mallāk districts.

A good metric of how the landed elite’s relationship with the state conditioned sub-national state capacity is the progress of land title settlement in both Iraq and Transjordan. Beginning in 1933 for both countries, the British encouraged land settlement as an avenue toward reforming a tax collection system that largely depended on land and agricultural production. At the time of the Iraqi July Revolution in 1958, Figure 1 shows that the British-led land settlement program had made few in-roads into provinces controlled by the mallāk elite.

Figure 1: Land title settlement in Iraq, 1958 (percent completion)

In Transjordan, however, British-led property right reforms to dismantle traditional forms of collective land tenure and to create a stronger tax among citizens was achieved primarily through the country’s colonial civil court system. Most citizens were small-holder farmers, but the courts treated claims from powerful landowners and the poor with surprising fairness.[5] Transjordanian land settlement was, de jure and de facto, more equitable and efficient than the settlement that occurred in Iraq. In 1946, the year of Jordan’s independence, most of country’s most densely settled regions in the central and northern parts of the country were surveyed and settled. By 1952, a mere six years later, the whole country had been surveyed (Figure 2).

Figure 2: Land title settlement in Jordan


Developmentally, the even application of property rights in Jordan and the class favoritism shown by the British in Iraq resulted in two very different levels of infrastructural power. Thorough property rights allocation requires that the state be able to survey the land, determine its owner, adjudicate disputes, and collect taxes. The pervasive reforms in Jordan achieved precisely this. In contrast, property in Iraq was not titled as private but rather as various types of state land that were distributed rent-free to favored elites. This process was less rigorous and more personalistic than the Jordanian land settlement, and one unintended consequence of the reform was that landed elites became more powerful than the state.

Iraq’s subnational variation in state capacity has profound ramifications for the provision of public services. In the rest of this essay, I focus on mapping how landlord-dominated regions differed from the rest of Iraq in state capacity and public service provision. As a result of preferential treatment by the British, landed elites became powerful economic and political actors that often enriched themselves to the detriment of their constituents. As a result, feudalistic regions were alienated from the state-building process and resulted in the under-provision of critical state institutions in landlord-dominated areas of Iraq.

Education and health provision in Iraq

Archival evidence suggests that the presence of landed elites favored by the British curtailed the spread of the Iraqi state and its services both during and after the colonial period. I use an original dataset to compare mallāk-dominated regions of Iraq and those where landholding was more equal. The core of my dataset draws on statistical yearbooks published by the the Iraqi government. As no library owns a complete collection, acquiring these yearbooks required extensive archival research in multiple locations, including the British Library, Durham University’s Middle East Documentation Unit, Oxford’s Bodleian Library, the Harvard Lamont Library, and the National Library of Jordan. These annual publications collect statistics from every ministry and sector, and thus provide an unprecedented account of the development of state institutions and the Iraqi economy from the end of the colonial period forward.[6] I digitized, cleaned, and systematized the data for each liwa[7] to create a panel from 1929 to 1965.[8] I measure two social policy outcomes: the percentage of the population enrolled in secondary school and the number of medical doctors per 10,000 liwa inhabitants. The number of doctors is only available from 1941 forward, and although this date falls after the official end of the British Mandate, the British re-occupied Iraq during World War II and maintained heavy influence in government and policy between 1933 and 1941.

Figure 3: Proportion of secondary students by liwa, 1929-1965

Descriptive statistics lend preliminary evidence that the privileged status of the Iraqi landed elite had a negative effect on the spread of state institutions in Iraq during the state-building period and in the post-revolutionary decade. Education and health care provision are frequently mentioned in the British Archive as priorities for the British and the post-colonial state. Despite these priorities, education and health care were slow to enter regions characterized by feudalistic land tenure arrangements. Plotting secondary education across liwas shows the trend that landlord-dominated regions tend to fall in among the lowest performers in secondary schooling (Figure 3). Plots shaded in gray are those dominated by the landed class. As with education, access to health care decreases in liwas dominated by large landowners (Figure 4).

Figure 4: Doctors per 10,000 inhabitants by liwa, 1941-1965

Discussion and conclusion

A useful frame in understanding the ramifications of state services like education and policing for state building is Migdal’s analysis of weak and strong states. Migdal (1988) argues that state social control, or the successful subordination of people’s own inclinations in favor of state rules, is critical for a strong state. A strong state is one where citizens are compliant, active participants in a system they consider legitimate. Weak states, in contrast, jockey for social control with other sub-state groups that spoil the state’s monopoly over the state-building process. While few states in the region could be described as immune to sub-state interests, this essay illustrates how some may have emerged from the colonial experience weaker than others and less able to pursue broad and generous social policies.

The case of Iraq shows how early state capacity is a critical consideration for the development of later social policy outcomes. This essay advocates looking to the past to explain the contemporary limitations of MENA states. The colonial administration’s favoritism toward the landed elite created pockets of weaker state institutionalization, shown here with longitudinal education and health data. Given the colonial origins of many public health and education systems in MENA, questions of who gets what, when, and why may be more helpfully explained through a historical lens and attention to the original design of Middle Eastern states caught between European preferences and domestic elite interests.

[1]   Statistical Yearbooks were published continuously until 1988 but begin to leave out variables of interest or to only report them at the national level. Another challenge is that administrative borders changed frequently in the post-revolutionary period (i.e., after 1958), and thus maintaining consistent geographic units will require additional methods to approximate variables of interest.

[2]   Province

[3]   The Iraqi liwas during this period are Mosul, Erbil, Sulimaniyah, Kirkuk, Baghdad, Dulaim (Ramadi from 1958), Diyalah, Diwaniyyah, Hillah, Kut, Amarah, Karbalah, Muntafiq (Nasiriyyah from 1958), and Basra.

[4]   See Fischbach (2000) for numerous examples of fellaheen winning in court over merchant and sheikh landowners.

[5]   See, for example, Banerjee and Iyer (2005); Besley and Persson (2009); Engerman and Sokoloff (2005); Mahoney (2003); Michalopoulos and Papaioannou (2014); Nunn (2009)

[6]   AIR 40/1419, FO 838/12, and FO 624/201, British National Archive

[7] The first law, the Tribal Criminal and Civil Disputes Regulation of 1925, gave Arab sheikhs and Kurdish aghas the ability to hold their own courts through a separate tribal legal system. The second set of laws gave landowners nearly unmitigated access to land and labor. The Land Settlement Law of 1932 created a new class of land tenure, miri lazmeh, that established a process for confirming the leases of landlords on very generous terms, particularly on newly irrigated lands. The Law Governing the Rights and Duties of Cultivators (1933) also gave landlords feudalistic powers over the fellāheen.

[8]   File AIR 23/120, British National Archive.