M id S Ea dl pe s e ci t & ResearchDigest al Is No World Bank su r t e hA on f r ic a VOLUME 9 L NUMBER 3 L SPRING 2015 99248 The State as Employer of Last Resort in Postrevolution Tunisia In the wake of the Arab Spring, evaluation, promotion, and compensa- IN THIS ISSUE hiring in Tunisia’s public sector more tion—but the authors find large gaps The State as Employer of Last Resort in than doubled while performance between the legal framework and its Postrevolution Tunisia … page 1 application. incentives weakened Recent changes in practices in Tunisia’s Recruitment into the public sector public sector raise questions about B has traditionally been based on com- performance incentives for civil servants etter job opportunities were petitive entry exams (concours) designed an important demand during to select the best candidates. But ad What Drives Weak Job Creation the Arab Spring protests. In the hoc decrees issued after the revolu- in Tunisia? … page 2 wake of the uprisings, however, private tion led to deviations from this system. Knowing which firms create the most jobs sector growth slowed and unemploy- These included decrees providing for a can help in tackling unemployment. A recent ment rose. In Tunisia, where at least general amnesty for civil servants who study dissected patterns in job creation in 30 percent of university graduates had lost their job for political reasons Tunisia were unemployed in 2012, the public under President Zine al-Abidine Ben Macroinsurance for sector stepped in to avoid further so- Ali; the direct recruitment of people Microenterprises … page 3 cial unrest. injured in the revolution and family Can insurance against political and In 2011 and 2012 the government members of those killed in the upris- macroeconomic risk encourage investment by of Tunisia hired 90,000 new recruits, ing; and the regularization of contract microenterprises? increasing the number of public sector and temporary workers. These ad employees by 17 percent, to 616,000 hoc changes allowed the state to act Testing the Effectiveness of Job Matching (excluding state-owned enterprises). as employer of last resort, sharply in Jordan … page 4 While public sector employment in increasing direct recruitment and by- Unemployment is pervasive among educated Tunisia had been growing steadily over passing competitive exams. young people in the Middle East and North the past decade, the January 2011 revo- Performance evaluation is manda- Africa. Could matching job seekers with job lution accelerated this trend, with an- tory for all civil servants, with the pub- openings help? nual recruitment more than doubling lic sector code requiring systematic Predicting Bank Insolvency in the Middle East (figure 1). Low GDP growth and falling evaluation on specific criteria. In prac- and North Africa … page 5 government revenues made the public tice, however, the grading system lacks A recent study developed and tested an early hiring spree fiscally costly, pushing the transparency and objectivity. Grades warning system for monitoring the risk of deficit up to 5.7 percent in 2012. are based mostly on a subjective ad bank insolvency in the Middle East and How was such a drastic increase in hoc assessment by the employee’s im- North Africa recruitment possible? And are the new mediate supervisor and are so inflated recruits likely to become top perform- that they provide almost no differen- Economic Inequality in the Arab ers? A recent paper by Brockmeyer, tiation between employees. Region … page 6 Khatrouch, and Raballand addresses Promotions are supposed to be Arab economies have seen growing income these questions by drawing on both merit based, with candidates ranked inequality between different groups and qualitative interviews with civil ser- on the basis of written exams, supervi- areas. What has been driving the increase? vants and legal analysis of the public sors’ recommendations, and annual Open Skies over the Middle East … page 7 sector’s performance management performance evaluations. In reality, system. The public sector code re- however, most are based on seniority, What effect would more liberal air travel quires the use of performance incen- and this has been especially true in policies have on passenger traffic between Turkey and the Arab states? tives along the entire career path of civil servants—in recruitment, (continued on page 8) 2 World Bank ResearchDigest What Drives Weak Job Creation in Tunisia? Distortions that prevent the most inferior to that of larger firms. desperately desire. At the same time, productive firms from expanding are Between 1996 and 2010 the inefficient producers are able to con- the biggest constraint to job creation Tunisian firm-size distribution re- tinue operating as somehow they man- mained severely skewed toward small age to protect their market share. in Tunisia firms. Net job creation The results suggest W was driven predomi- Job creation in that the binding con- hich firms create the most nantly by jump-start straint to job creation jobs in developing coun- self-employment Tunisia functions in Tunisia is the lack of tries? The jury is still out (one-year-old firms more like a nursery a business climate that on this question. This is unfortunate employing one worker). fosters the expansion because the answer has important Indeed, the start-up of shielded from the sun of the most productive policy implications for how to tackle one-person firms was than like a proper firms. A defunct finan- unemployment. an important contribu- cial sector, cumbersome Examining which firms create jobs tor to aggregate job cre- garden—no matter bankruptcy procedures, in an economy can shed light on the ation. And young firms how many seeds are and policies such as constraints to job creation. Where accounted for most new regulatory barriers to job creation is weak, this could reflect jobs. planted, there will be firm entry and expan- distortions inhibiting the growth of Small firms’ post- few blossoms unless sion and to the hiring productive firms or be due to demand entry performance and firing of workers all constraints, with productivity a poten- in generating jobs is barriers are removed undermine competitive tially even more important determi- feeble, however, and pressures and hamper nant of growth and survival. Moreover, job creation is impeded by a lack of the efficient allocation of resources. As limited job creation may be driven by upward mobility. While small firms on a result, programs to finance new and stagnation among incumbent firms, average tend to create more jobs than existing small or medium-size enter- insufficient entry, or excessive exit. larger ones, the reason is that small prises are ill equipped to address un- An understanding of which of these firms tend to be younger; once firm employment because these firms rarely matters most and who creates jobs age is controlled for, small firms cre- grow. Job creation in Tunisia functions can guide the design of policies that ate fewer jobs than large firms. In other more like a nursery shielded from the promote private sector development. words, small firms create more jobs sun than like a proper garden—no Programs supporting small and because they are young, not because matter how many seeds are planted, medium-size enterprises, for example, they are small. While similar patterns there will be few blossoms unless bar- are predicated on the notion that have been documented for developed riers are removed. small firms create the most jobs. economies, the process of market se- A recent article by Rijkers, Arouri, lection in Tunisia is markedly different. Freund, and Nucifora dissects pat- Developed economies exhibit an up- terns in private sector job creation in or-out dynamic where the most pro- Tunisia. The authors use a unique data ductive entrants expand and the weak- set containing information on all reg- est ones are driven out of business. In istered firms, including the self-em- Tunisia, by contrast, even entrants that ployed. They focus especially on the stay in business don’t manage to grow; contributions of small and medium- small firms are more likely to die but, size enterprises, which account for the conditional on survival, are not more bulk of all employment, as is typical in likely to grow. developing countries. Based on their These patterns are driven by defi- findings, they argue that weak aggre- ciencies in the (re)allocative process; gate job creation was due not to ex- while both profitability and productiv- cessive job destruction and exit but to ity are positively associated with net insufficient dynamism, manifested in job creation, this correlation is very a lack of upward mobility and stagna- weak. The best firms have difficulties tion among incumbents, and to mar- expanding and gaining market share, ket imperfections preventing the best and firm productivity does not rise rap- Bob Rijkers, Hassen Arouri, Caroline Freund, and firms from growing. While entry of idly with firm age. This suggests that Antonio Nucifora. 2014. “Which Firms Create the small firms was an important engine of markets don’t select the best firms, but Most Jobs in Developing Countries? Evidence from job creation, the performance of such instead impede them from creating Tunisia.” Labour Economics 31 (December): firms in creating jobs after entry was the jobs that deserving Tunisians so 84–102. World Bank Research Digest 3 Macroinsurance for Microenterprises Does political and macroeconomic If a second covered shock occurred in microenterprises seldom make fixed uncertainty inhibit investments a subsequent month, the insurance capital investments anyway. Their by microenterprises—and would would pay out an additional month’s loans went largely to working capital— loan installment. investments that are short term and insurance make a difference? The insurance was designed to easily adjusted and thus less affected F cover certain macroeconomic shocks, by uncertainty. The results suggest that irms in developing countries rank including a suspension of the stock despite complaints about uncertainty policy uncertainty and macro- exchange, the imposition of extended as a constraint to business growth, this economic instability as the two curfews, a rise in monthly inflation to uncertainty does not appear to be pre- biggest constraints to their growth in its highest level in five years, or the venting microenterprises from making the World Bank’s Investment Climate removal of subsidies on gasoline or profitable investments in machinery Surveys—ahead of taxes, regulation, liquefied petroleum gas (LPG). These and equipment. corruption, and access to finance. shocks were identified through market While the insurance product did Faced with such uncertainty, firms testing as salient in the wake of the not change the investment behavior may be reluctant to make new invest- revolution and likely to harm business of firms, the demand for the product ments or hire new workers, choosing owners either directly or because of suggests that firms might still value instead to wait and see how the econ- what they would likely signal about the protection it could provide when omy evolves. This concern may be economic and political chaos in the shocks do occur. The product could particularly important in much of the country. also be more valuable for small and Middle East and North Africa, where The authors worked with a sample medium-size enterprises, which may firms have faced dramatic increases in of 2,961 microfinance clients, all of be more likely to be considering lumpy political and macroeconomic instabil- which were finishing one loan and capital investments that they defer be- ity in the context of the Arab Spring. about to decide whether to take cause of uncertainty. Insurance is a natural financial in- out a new loan between April and Insurance penetration is very low in strument to help protect against risk. September 2012. The clients were Egypt, with regulations inhibiting the But while much attention has been randomly sorted into two groups of development of the sector. As a result, given to developing insurance prod- equal size: a treatment group that was many firms have little understanding ucts for farmers, hardly any has been offered the insurance product and a of the concept of insurance. This unfa- devoted to developing them for other control group that was not. These firms miliarity with insurance does not ap- enterprises. A recent paper by Groh were microenterprises, with most hav- pear to explain the results of the study, and McKenzie describes an effort to do ing no paid workers. Half were in retail, but does make it hard to launch new so in the Arab Republic of Egypt. The the rest in manufacturing or services. products and build an understanding authors worked with the Alexandria Median monthly profits were around of how they work. Given the enormous Business Association (ABA), Egypt’s $200, and the average loan was around volatility in incomes for microenter- largest microfinance organization, to $800. prise owners in developing countries, develop and pilot an innovative insur- Demand for the insurance prod- further policy and research efforts are ance product to help protect its clients uct was relatively high: the insurance warranted to develop better ways to against uncertainty in the aftermath was purchased by 36.7 percent of all help insure against some of this risk. of the January 2011 revolution, which firms in the treatment group and by ended almost 30 years of rule by 55 percent of those taking a new loan. President Hosni Mubarak. The insurance had no impact on firms’ The insurance product was de- investment behavior, however. Firms signed for microfinance clients de- in the treatment group were no more ciding whether or not to take a new likely to take a loan, did not take larger 12-month loan from ABA. The insur- loans, and were no more likely to use a ance, which could be purchased only loan to make new capital investments. along with a new loan, cost 0.5 percent (The sample was relatively large and of the value of this loan. If a shock cov- confidence intervals are small around ered by the insurance occurred during zero, so this lack of effect is not a re- the loan period, the payout would be sult of low statistical power.) As a re- two months of loan principal install- sult, profits and sales did not increase. Matthew Groh and David McKenzie. 2014. ments—one to ABA to cover the cur- Indeed, there is some evidence that “Macroinsurance for Microenterprises: A Random- rent month’s loan installment and one revenues fell. ized Experiment in Post-Revolution Egypt.” Policy to the client to compensate for poten- What is the reason for the lack Research Working Paper 7048, World Bank, tial loss in business from the shock. of effect? It appears to be that these Washington, DC. 4 World Bank ResearchDigest Testing the Effectiveness of Job Matching in Jordan Matching job seekers to vacancies potentially interested in hiring gradu- who were matched at least once, and had little impact on employment ates through the program. only 2.1 percent of the matches made. outcomes among educated youth in The pilot intervention involved Furthermore, among the 24 candidates three main steps: who were hired, 15 quit within the first Jordan. What could explain this? •  Testing the job seekers. The job month. Thus only 9 jobs were directly U seekers in the treatment group under- generated through this matching. Not nemployment is pervasive went two hours of computerized tests surprisingly, treatment regressions among educated young people to measure personality type, profi- show no significant impact of the job in many countries of the Middle ciency in English and Excel, and quan- matching on employment outcomes East and North Africa. In Jordan in titative, verbal, and spatial reasoning for graduates. 2010, the unemployment rate was 19 as well as two hours of live, interactive The paper explores several poten- percent for men ages 22–26 with a sessions to measure soft skills. tial explanations for the lack of impact. postsecondary degree, and 47 percent •  Matching job candidates to firms. Job candidates turned down 28 percent for their female counterparts. Why Firms were asked detailed questions of the match opportunities they were doesn’t the labor market clear for edu- about the positions they were con- given and turned down or quickly quit cated youth? sidering filling, and the psychometric 83 percent of the job offers. This leads Much of the discussion of this testing results were used to identify to the conclusion that much of the un- question has focused on such issues three to five candidates who would be employment among educated youth is as shrinking government sectors com- considered a good fit for each posi- voluntary, reflecting an unwillingness bined with a range of regulatory and tion. These candidates were asked to work in many jobs. market failures that inhibit the growth whether they would be interested in This conclusion suggests that the of private sector firms, resulting in the position; if not, they were replaced necessary policy response is more low demand for labor. But this doesn’t by another match. The firm was sent difficult and complicated than if the explain why labor markets don’t clear résumés for the matched candidates problem were simply high minimum through movements along the labor along with a description of why they wages or high search costs. The results demand curve, with wages falling to a would be a good fit. The firm would suggest two directions for future policy point where labor demand equals la- then choose whether to interview a actions: interventions on the firm side bor supply. candidate and then whether to offer to spur the development of a vibrant One potential explanation is of- the candidate a job. private sector that provides a larger fered by search and matching theory, •  Following up. The firms and job number of skilled jobs, and efforts which suggests that persistent un- candidates were contacted to learn the to encourage educated youth to con- employment occurs because of costly outcome of the matching process, and sider a broader range of occupations, search frictions that make it difficult attempts were made to arrange alter- including those now considered less for jobless workers to match with firms native matches if an initial match fell prestigious. with vacancies. These search frictions through. can be greater for youth, who lack Of the 1,011 people in the treat- previous job experience to signal their ment group, 56 percent (564) were quality as employees. Are there inter- matched to at least one job opening— ventions that could overcome these with 55 percent of these candidates be- frictions? A recent paper by Groh, ing matched to more than one—for a McKenzie, Shammout, and Vishwanath total of 1,143 initial matches. But only describes a pilot program that tested 10 percent of these matches (115) led an intervention designed to do so in to a job interview, with the result that Jordan. 58 job candidates were interviewed. In The program worked with a sample 28 percent of matches job candidates of 1,354 recent tertiary graduates said that they were not interested, and seeking jobs: 1,011 individuals (three- in 55 percent of matches firms did not quarters of the sample) were randomly invite candidates for an interview. assigned to a treatment group, whose In the 115 matches leading to a job members the program would attempt interview, job offers were extended Matthew Groh, David McKenzie, Nour Sham- to match to jobs, and 343 were as- in almost half the cases (54). But job mout, and Tara Vishwanath. 2014. “Testing the signed to the control group. It also candidates refused 30 of the 54 job Importance of Search Frictions, Matching, and worked with a sample of 2,279 small offers extended, resulting in only 24 Reservation Prestige through Randomized Experi- and medium-size firms and a booster people being hired. These represent ments in Jordan.” Policy Research Working Paper sample of larger firms that were only 4.3 percent of the job candidates 7030, World Bank, Washington, DC. World Bank Research Digest 5 Predicting Bank Insolvency in the Middle East and North Africa An off-site monitoring system indicators in five categories: Capital a traditional bank run or a wholesale based on bank fundamentals could adequacy, Asset quality, Management bank run, as the recent crisis showed strengthen banking supervision in capacity, Earnings power, and Liquidity on a large scale. For these cases it position. Most stud- would be useful to in- the Middle East and North Africa ies focus on the United A systematic off-site vestigate the determi- T States, where the data nants of bank fragility he global financial crisis revealed are most readily avail- risk analysis based in the region within a substantial differences in the able. Others take a resilience of individual banks cross-country perspec- on an early warning structural credit risk model using market- in the Middle East and North Africa, tive, analyzing bank system is useful to based measures of bank highlighting the importance in bank- fragility in the European vulnerability, such as ing supervision of differentiating be- Union or Eastern the extent that it credit default swaps or tween financial institutions. In many Europe. But none of can help banking default risk measures cases knowledge about the specific the relevant studies has features underpinning the health of focused on the Middle supervisors prioritize implied by stock prices. Also interesting would individual financial institutions has East and North Africa, financial institutions be research that takes a been crucial in identifying problems and the author attempts and informing policy action. This sug- to fill this gap. for on-site inspections systemic risk perspec- tive in analyzing the gests that an early warning system To derive an ear- and identify those sources of bank vulner- built on bank-specific fundamentals ly warning system for can provide a useful monitoring tool, monitoring the risk of needing supervisory ability in the region. Such an approach can helping banking supervisors to identi- bank insolvency in the interventions shed light on the risk of fy vulnerable financial institutions and region, the author esti- contagion by capturing take preemptive steps to aid those mates a pooled logistic regression on the codependence of individual banks’ that are ailing. a sample of 198 banks in 19 countries risk. Earlier financial crises prompted ef- for the period 2001–12. Not surpris- The findings of the paper have im- forts to develop early warning systems, ingly, the findings show that tradition- portant policy implications at a time and the most recent global financial al CAMEL indicators are significant when implementing effective risk- crisis has led academics and policy predictors of bank insolvency in the based supervision remains a challenge makers to intensify these efforts, in- region, in line with previous research in the Middle East and North Africa. creasingly focused on developing in- in different geographic areas. In par- In particular, the results highlight the dicators and methodologies that can ticular, the findings indicate that small usefulness of a simple off-site moni- aid in the timely identification of signs banks with low capitalization, low as- toring system based on traditional of distress. A recent paper by Calice set quality, and high cost-to-income financial metrics associated with the attempts to provide a fundamentals- ratios are more likely to become insol- CAMEL rating system. But the findings based surveillance tool that can serve vent over the next two years. Results should not be construed as detract- as an early warning device in the are robust to the inclusion of addi- ing from the importance of the on-site Middle East and North Africa by iden- tional control variables and to different examination process to a successful tifying the factors that affect the risk of estimation methodologies. The predic- banking supervision program. A sys- bank insolvency in the region. tive power of the early warning system tematic off-site risk analysis based The paper builds on and adds to as measured by both in-sample and on an early warning system is useful the literature on bank default. This out-of-sample performance is also rea- to the extent that it can help banking literature focuses mostly on test- sonably good. supervisors prioritize financial institu- ing the effect of accounting-based Given the definition of technical tions for on-site inspections and iden- variables on the probability of bank insolvency used in the analysis, the tify problematic institutions needing default in discriminant analyses and early warning system developed is supervisory interventions. logistic regression frameworks. The generally able to predict the occur- assumption is that insolvent banks rence of an insolvency event only when present ex ante weaker fundamentals the event originates from widespread than solvent banks. The literature on losses on the asset side of a bank’s individual bank default draws heavily balance sheet—losses that lead to a on the CAMEL rating system, a tool progressive deterioration in the bank’s Pietro Calice. 2014. “Predicting Bank Insolvency widely used by banking supervisors to fundamentals. But insolvency events in the Middle East and North Africa.” Policy Re- monitor the performance and sound- can also originate from the liability search Working Paper 6969, World Bank, Wash- ness of financial institutions based on side of a bank’s balance sheet, such as ington, DC. 6 World Bank ResearchDigest Economic Inequality in the Arab Region Development policies in many Arab The analysis finds moderate levels is particularly strong for poor and economies may have led to greater of inequality, with the Gini coeffi- middle-class households. inequality between different areas cient for the distribution of per capita The analysis of the sources of in- household expenditure ranging from equality between metropolitan and and socioeconomic groups less than 32 in Libya and the Arab nonmetropolitan areas shows differ- T Republic of Egypt to more than 40 in ent patterns across Arab economies. he state-led economic devel- Mauritania, Tunisia, and the Republic In Egypt, the United Arab Emirates, opment model adopted by of Yemen. Arab Spring countries ex- and the Republic of Yemen differ- Arab countries following their hibit varying patterns of income and ences in the returns to household independence led to remarkable inequality, with low income and low in- endowments have a greater effect on improvements in human development equality in Egypt, low income and high inequality than do differences in those indicators and to moderate levels of inequality in the Republic of Yemen, endowments, particularly for middle- poverty and income inequality. Indeed, medium income and low inequality in class and better-off households; this despite the modest economic growth Libya, medium income and medium suggests that although metropolitan in the Arab region, its levels of poverty inequality in the Syrian Arab Republic, households enjoy better endowments, and inequality compared favorably and relatively high income and high the inequality in these countries is with those in all other developing inequality in Tunisia. explained mainly by the higher returns regions. The most important determinant to these endowments in metropolitan In the early 2000s, however, the of economic inequality in Arab econo- areas. In the Syrian Arab Republic and postindependence development mies appears to be the demographic Tunisia, by contrast, the differences in model began to feel the burden of its composition of households, followed household endowments are more im- weight and these gains started to un- by their geographic location. The per- portant in explaining why metropolitan ravel, ushering in rising inequality of sistence and widening of urban-rural households are better off than their different forms. Disparities between and regional inequalities are of serious counterparts. different areas and socioeconomic concern for the prospects for social The findings suggest an urban and groups within economies have wid- cohesion and inclusive growth. These metropolitan bias in the development ened. Inequality of opportunity has inequalities may exacerbate the social policies of many Arab governments, also emerged as a serious concern in and economic instability that pervades which led to growing welfare inequali- the region. Whether these inequali- the Arab world and could undermine ties along rural-urban and regional ties helped lead to the Arab Spring popular support for market economy lines. Furthermore, the lack of employ- is difficult to ascertain, but the de- reforms. ment opportunities and agglomeration mand for greater equity and inclusion To help improve the understanding economies in nonmetropolitan areas has brought the issue of economic of the sources of spatial inequality, the may have led to lower returns to hu- inequality to the forefront for Arab paper uses the unconditional quantile man capital investments in these areas governments. regression method and decomposes than in metropolitan areas. A recent paper by Belhaj Hassine the interregional welfare gap across The market economy reforms attempts to contribute to the the entire distribution into the differ- implemented in some Arab economies understanding of consumption ences in household endowment char- during the past decade appear to have expenditure inequality in the region. acteristics such as age, education, and been more beneficial for households The analysis draws on harmonized employment and the differences in the that had better human capital en- data from 28 household surveys in 12 returns to these characteristics. dowments at the time of the reforms. Arab economies to assess inequality The analysis finds that urban-rural This would have contributed to faster over time and to examine its sources inequality is due primarily to differ- growth in the returns to urban and and structure. Given the importance ences in the distribution of household metropolitan households, particularly of regional disparities and imbalances characteristics. Despite the rural in the returns to their education and in many Arab economies and their development efforts that have been experience. Further reforms may ex- contribution to social unrest, the pursued by Arab governments, urban acerbate inequality if the obstacles to paper devotes special attention households continue to have much balanced regional development are not to the analysis of rural-urban and better endowments of demographic adequately identified and addressed. metropolitan-nonmetropolitan and human capital attributes than inequality across the entire welfare their rural counterparts and this differ- distribution and attempts to draw ence explains most of the welfare gap policy implications for the design of between the urban and rural sectors. Nadia Belhaj Hassine. 2014. “Economic Inequal- strategies to address inequality and The relative effect on inequality of ity in the Arab Region.” Policy Research Working promote equity. differences in household endowments Paper 6911, World Bank, Washington, DC. World Bank Research Digest 7 Open Skies over the Middle East Substantial liberalization of air agreement and the inclusion of Turkey changes would generate very large travel could more than double the in this deeper ACAC agreement. changes in passenger traffic and in the passenger traffic between Turkey The paper exploits data from the number of flights between Turkish and International Civil ACAC cities. Indeed, the and the Arab states Aviation Organization estimates suggest that Even in the age T that document pas- liberalization to the lev- urkey, long a fulcrum between senger flows between of the Internet el of the U.K.-Singapore the West and the East, has deep- city pairs around the ened its economic links with the globe. These data allow physical connectivity agreement would more than double the traffic European Union and is now turning the models to link city matters. Goods between Turkey and the to the Middle East. In this sometimes and country character- turbulent neighborhood it is begin- istics to the existence must be delivered, Arab states. The odds of a direct flight between ning to create dynamic trade links that of a flight and to levels businesspeople must a given Turkish city and echo past relationships, though in a of traffic on existing new international context. Even in the flights. A model that meet, and individuals a city in an Arab state would rise by a factor age of the Internet physical connectiv- predicts flows between must travel to of 1.7–3.0. Put another ity matters. Goods must be delivered, cities can provide more way, the model suggests businesspeople must meet, and indi- textured information in forge bonds and that Turkey’s accession viduals must travel to forge bonds and support of policy analy- catalyze trade. Air to a more liberal ACAC catalyze trade. Air transport is criti- sis than can a model cal in facilitating each of these links, that only predicts flows transport is critical agreement would lead to 25–29 new city pairs especially in a region where travel by between countries. For in facilitating each of sharing a direct inter- land is fraught with difficulty. example, in Lebanon national flight. Many Air travel between countries has liberalization would these links of these new city pairs typically been governed by policies mean more traffic would include a second- negotiated on a bilateral basis, and through Beirut, but in Turkey liberaliza- ary city in Turkey. travel between Turkey and the Arab tion would generate new flights and While these results may seem op- states is no exception. Data from the more traffic through secondary airports timistic, there are three reasons to World Trade Organization indicate that such as those serving Adana, Antalya, suggest that they are not unrealistic. Turkey’s bilateral air services agree- and Konya. First, current levels of air passenger ments with many of its Arab neighbors The scenario for evaluating the traffic in the region are low, and rapid remain quite restrictive by interna- consequences of a deeper ACAC agree- growth is plausible from a small base. tional standards. By contrast, air travel ment considers policy changes that Second, some regional markets have between most Arab states is governed would make this agreement as liberal seen even faster growth rates in recent by a plurilateral air services agreement, as the most liberal air services agree- years. Finally, the changes in policy the Intra-Arab Freedom of the Air ment in the world—such as the bilat- considered are large, especially those Programme of the Arab Civil Aviation eral agreement between the United in the policies governing air traffic be- Commission (ACAC). While the ACAC Kingdom and Singapore. Model esti- tween Turkey and the ACAC countries. agreement is fairly open, there is scope mates suggest that policy changes of The results should nonetheless be un- for further liberalization. that magnitude would increase traffic derstood as the product of preliminary A recent paper by Cristea, Hillberry, among the Arab states by around 30 work to assess the possibilities associ- and Mattoo seeks to improve the un- percent. Moreover, the probability that ated with further reform in the region. derstanding of the links between air two ACAC cities would be connected travel policy and passenger traffic in by a direct international flight would the Middle East. The main aim is to rise by a factor of 1.3. provide quantitative estimates that can Turkey’s bilateral agreements with inform policy choices. Econometric countries in the region tend to have models link the existence of flights much more restrictive policies than between city pairs and the number of those in the present ACAC agree- passengers traveling, respectively, to ment. Thus if Turkey were to join the quantitative measures of policy em- liberalized ACAC agreement as a full bedded in the air services agreements member, the policies governing flights Anca Cristea, Russell Hillberry, and Aaditya Mat- and to other variables. Estimates from between Turkey and the Arab states too. 2015. “Open Skies over the Middle East.” the models are used in two policy would become much less restrictive. Policy Research Working Paper 6937, World simulations: a deepening of the ACAC The paper estimates that such policy Bank, Washington, DC. 8 World Bank ResearchDigest (continued from page 1) Figure 1. Public Sector Recruitment in Tunisia, 2000–13 Recent Policy Research New recruits hired (thousands) Working Papers 50 6927 Economic Development and Female Labor Participation in the Middle East and North 40 Africa: A Test of the U-Shape Hypothesis Paolo Verme 6940 Economics of Transiting to Renewable Energy 30 in Morocco: A General Equilibrium Analysis Govinda R. Timilsina and Florian Landis 6950 New Coincident and Leading Indicators for 20 the Lebanese Economy Samer Matta 6996 Egypt: Inequality of Opportunity in Education 10 Lire Ersado and Jeremie Gignoux 2000 2005 2011 2013 7023 The Reduction of Child Mortality in the Source: Data from the General Committee for the Civil Service. Middle East and North Africa: A Success Story the past few years. In the aftermath of salary and other group-specific allow- Farrukh Iqbal and Youssouf Kiendrebeogo 7026 Inequality of Opportunity among Egyptian the revolution the government sought ances. These allowances as well as Children to avoid discontent in the public sec- premiums for particular professional Lire Ersado and Meltem Aran 7031 Female Labor Participation in the Arab tor by speeding up promotions—and groups have proliferated since the rev- World: Some Evidence from Panel Data in the simplest way to do so was to base olution, increasing the complexity and Morocco Paolo Verme, Abdoul Gadiry Barry, and Jamal them on seniority. Indeed, some agen- opacity of the compensation system. Guennouni cies resorted to across-the-board au- Although the weaknesses of perfor- 7043 Updating Poverty Estimates at Frequent Intervals in the Absence of Consumption Data: tomatic promotions, in which all those mance management in Tunisia’s public Methods and Illustration with Reference to a at a certain level were upgraded to the sector have been known for decades, Middle-Income Country Hai-Anh H. Dang, Peter F. Lanjouw, and Umar next one. past reform attempts failed to bring Serajuddin Compensation schemes also suffer about change. An analysis of the re- 7127 Education Attainment in the Middle East and North Africa: Success at a Cost from weaknesses that limit perfor- form strategies of 1989 and 2007 shows Farrukh Iqbal and Youssouf Kiendrebeogo mance incentives. Public sector com- that the present challenges were 7132 Transparent Government and Business pensation is uncompetitive compared already identified in 1989. Relevant Regulation: “Open for Business?” Carolin Geginat and Valentina Saltane with private sector pay—salaries for reforms were proposed, but few have 7135 Economic Effects of the Syrian War and the engineers, for example, are about twice been implemented. Understanding Spread of the Islamic State on the Levant Elena Ianchovichina and Maros Ivanic as high in the private sector. In addi- the political economy of reform in 7224 An Evaluation of the 2014 Subsidy Reforms tion, the compensation has virtually Tunisia’s public sector thus remains an in Morocco and a Simulation of Further Reforms no merit-based component. Most civil important task for future research. Paolo Verme and Khalid El-Massnaoui servants receive a performance bonus, 7225 The Quest for Subsidy Reforms in Libya Abdelkrim Araar, Nada Choueiri, and Paolo but this bonus is essentially fixed be- Anne Brockmeyer, Maha Khatrouch, and Gaël Verme cause it is linked to the performance Raballand. 2015. “Public Sector Size and evaluation in which almost all employ- Performance Management: A Case Study of Post- Working Papers can be downloaded at http://econ.worldbank.org ees attain the highest mark. Moreover, Revolution Tunisia.” Policy Research Working To download the World Bank Research E-Newsletter, the bonus is dwarfed by the base Paper 7159, World Bank, Washington, DC. go to http://econ.worldbank.org/research_newsletter The World Bank Research Digest is a quarterly publica- The Research Digest is financed by the Bank’s Editorial Committee: Indermit S. Gill (managing editor), tion disseminating findings of World Bank research. Research Committee and managed by DECDP, the Aslı Demirgüç-Kunt, and Shiva S. Makki. Editor: Alison The views and interpretations in the articles are those research support unit of the Development Economics Strong; production: Roula Yazigi. For information or of the authors and do not necessarily represent the Senior Vice Presidency (DEC). The Research Digest is free subscriptions, send email to research@worldbank. views of the World Bank, its Executive Directors, or the not copyrighted and may be reproduced with appropri- org or visit http://econ.worldbank.org/research_digest. countries they represent. ate source attribution. The World Bank 1818 H Street, NW Washington, DC 20433, USA Printed on Recycled Paper