logo logo

The European evaluation of the Association Agreements: guarantees and content still insufficient

All the versions of this article: [English] [français]

PNG - 90.4 kb

FTDES | 30 May 2020

Press release – The European evaluation of the Association Agreements: guarantees and content still insufficient

On 7 April 2020, the “evaluation consortium”[1] of the Euro-Mediterranean Association Agreements[2] published its interim report[3]. This evaluation, commissioned by the European Commission, has already been criticised by 24 organisations from the six Mediterranean countries concerned and from Europe[4], last December. Some organisations had also submitted their comments on the terms of reference[5]. Despite all these contributions and criticisms, as well as the consultations held on the subject of this study[6], we note that this evaluation does not deviate from its trajectory. In addition, on May 8, we submitted specific questions[7] to the evaluation team to improve its content, and we note some marginal improvements in the interim report[8].

However, these marginal improvements are still far from allowing us to consider the results of this study satisfactory for a real evaluation of the Association Agreements. The stakes are all the more significant in that this study is supposed to assess the impact of the previous free trade agreement in order to inform the DCFTA negotiations. In the light of this new report, it is therefore essential to repeat today the criticisms we made last December and in 2017:

– The macroeconomic model used is still inconsistent with its objectives and imposed by the European Commission, which reveals a violation of the principle of independence, as the evaluators only analysed the results provided by the European Commission. Moreover, it is a model with neo-liberal assumptions that are known for their consistently positive results in terms of growth, prices and wages, which is contradicted by other models.

– This model is complemented by statistical analysis and case studies. These are too limited and superficial to compensate for the model’s shortcomings, notably the fact that it does not assess the impact of the agreement over time and with empirical data.

– The inputs and involvement of civil society are still not taken into account.

– The methodology is still focused on the assessment of the consequences on trade flows, whereas it is the consequences of these agreements on the way economies are structured and therefore on their social, budgetary or environmental consequences that should really be placed at the centre of the assessment.

– The case studies in the sustainable development section are very approximate and superficial. They use indicators that do not seem appropriate[9] and give results that are contradicted by empirical evidence[10]. The sustainable development section does not follow either the UN guidelines[11] or those of the European Commission itself[12].

– The fact that the published reports are only in English is very damaging, since many people, especially in North Africa, do not master this language. At the very least, a summary should have been available in French and Arabic.

– The quality of the report and analyses suffers terribly from the demand to evaluate six agreements at the same time in a comparable manner. The report itself acknowledges this[13], as the analyses on each country are very incomplete and do not go into an in-depth analysis.

Thus, the lack of independence, the priority given to the commercial aspects instead of the human aspects, and the superficiality of this report prevent it from being recognised as a credible evaluation of the Association Agreements. Consequently, the Tunisian Forum for Economic and Social Rights:

– Recalls the contributions made by civil society, cited in this press release, and calls for them to be taken into account.

– Calls once again for the reformulation of the study to put human, economic, social and environmental conditions and rights at its centre.

– Calls for country assessment reports, more comprehensive than the general report, in English, Arabic and French.

– Calls for the final report to be fully published in Arabic, English and French.

– Recalls that while this evaluation study cannot be considered as measuring the impact of the Association Agreements unless it is radically transformed, the response to the questions and remarks of civil society is essential and will allow the final report to be as informative as possible.

Questions to the evaluation team on the interim report of the Euromed Association Agreements ex-post evaluation

General questions:

1- The report states that the evaluation was difficult to conduct because of the number of FTAs to evaluate. Indeed, one may find it hard to grasp the consequence of an individual FTA. Could you conduct in depth analyses for each country?

2- Why didn’t you schedule to evaluate the impact of the FTAs in the evolution of inequalities in the partner countries?

3- As future steps, the study “will provide additional information on impact on consumers, on state revenue and on the informal economy”. These should be essential parts of the evaluation, subject to in-depth investigation on their evolution and on the FTA influence, don’t you agree? Could you make an in-depth study of them instead of simply providing additional information?

4- “Because of the nature of the exercise, the results cannot be understood as the ‘effects of the implementation of the Euro-Med FTAs’ but rather as the ‘value the Euro-Med preferences’, expressed in terms of GDP, welfare, trade, output and environmental emissions. As such, they cannot be directly compared with the actually observed changes in trade and other data.” (p.97). What can really be inferred from the study in terms of the actual changes triggered by the AA? Could you make the above statement as well as the answer to this question apparent in the final report?

5- One can notice that most time frames (especially graphs) presented are in the span 2007-2018 while a lot of the FTAs, especially the ones with Tunisia and Morocco, entered into force way earlier (1998 and 2000 respectively). Could you adjust all the figures and graphs in order to get a complete picture? When the graphs begin 10 years before the AA entering into force (1988 then, at least for Tunisia) and the point of entry into force of the agreement is made clear thanks to a bar or an arrow, this is very helpful.

6- In many parts of the report, especially in the case studies, it is said that value chains should be studied in greater details to understand more the processes that have happened. Could you lead such detailed analysis, for the case studies (including sustainability ones) at least? By doing so, could you provide an analysis on whether or not local production was replaced by imports in the sectors that showed the most variation as well as on the type of actors that benefitted or suffered from the FTA (who are winners and losers along the value chain)?

Economic analysis:

7- Before turning to the model, the study tries to grasp what has actually happened. However, this part mainly concentrates on the evolution of tariff preference margins, which is an interesting indicator but is far from telling the whole story. Could you complement this analysis with the study of other indicators? For instance you could analyse statistical time series, in order to see which industrial/agricultural exchanges grew or fell, how the growth of these exchanges changed, then compare this with the evolution of national productions, its share of exports and of exports (generally and to the EU for each sector), its share of national sourcing, its situation in value chains and value captured and retained in the national territory… All of this by comparing the situations prior and after the FTAs.

8- Why didn’t you make a total of the model’s estimations on output? Could you make such totals apparent?

9- When I made these totals, I found the following (in million €): -380 for Egypt, -43 for Jordan, -591 for Morocco, +785 for Tunisia, +269 for the EU. This leads me to ask: how come GDP growth is expected in all of the countries when there is an actual decline in output for some of them?

10- SMEs: the interim report states that there will be an investigation on success stories of integration to trade flows by SMEs. Despite being interesting, investigation in success stories do not fit in such an evaluation. For instance, if we hypothetically say that there is one success story for 100 companies which shut down because of the FTA, one will be very far from understanding what happened. Looking at success stories then appears uninformative. It does not answer a criticism that had been made against the AAs, which may have destroyed a large chunk of the industrial tissue in some countries. Hence, could you instead analyse the extent to which SMEs and the overall industrial tissue of the industrial sectors of the SMCs had to face stronger competition, and how it evolved as a result?

Economic case studies:

11- Why didn’t you use national data in case studies? Could you make use of national data in order to provide a deeper and more complete analysis for each country?

12- Could you address some lacking elements, such as the evolution of tariffs compared with the actual evolution of exports and imports and the influence of these?

13- How come Tunisia is able to maintain high tariffs for industrial imports under the agreement while it is supposed to make them drop to zero (as is the case for other countries)? Could it be a reason of its better relative performance under the FTA?

Human Rights :

14- What guidelines did you use in order to formulate the human rights assessment? In a letter of the European Commission, they indicated that you might follow the “Guidelines on the analysis of human rights impacts in impact assessments for trade-related policy initiatives” of the Commission. If so, could you make clear the different processes indicated in these guidelines, namely screening, scoping and detailed assessment, which are not apparent in the interim report?

15- The UN Guiding Principles on Human Rights Impact Assessments of Trade and Investment Agreements are also essential to follow. Of particular interest would be the following: “Based on the results of the human rights impact assessment, a range of responses exist where an incompatibility is found, including but not limited to the following:(a) Termination of the agreement; (b) Amendment of the agreement; (c) Insertion of safeguards in the agreement; (d) Provision of compensation by third-State parties; (e) Adoption of mitigation measures” (p.8). Would you use the UN guidelines as a whole, and make use of this specific guideline?

Case study employment:

16- How come the employment case study seems to analyse the economic consequences of the FTAs deeper than the actual economic analysis?

17- An equivalence is made between sectoral output and employment for the CGE analysis. This is relevant “to the extent that labour intensity does not change”. Could you complement your analysis by studying these changes?

18- Why didn’t you include a study on the employment situation in each country? Employment patterns are very complicated and depend on several factors, thus not analyzing the trends within each country seems to give an analysis “out of the blue”.

19- Could you also go further into the analysis, by studying whether the people who came to work in the textiles and clothing and chemicals, plastics and rubber sectors were entering the labour market, or if they changed jobs, what were their comparative labor conditions (wage levels, hours worked, health, overall working condition…) both compared to before entering the sector and the progressive evolution of these conditions?

20- Besides, how can the study conclude that a trade balance improvement results in employment benefits? Indeed, the authors do state that many factors are in play, mainly the technological content, “the extent to which imported products compete with domestic products”, the extent of trade diversion, the extent of actual activity that is created or destroyed as a results of these evolutions and to what extent do these added/removed activities employ workers. Could you analyse these very factors in order to give a more adequate picture of the situation?

Environmental impact analysis:

21- Could you argue on why did you chose to use the “Environmental Performance Index, developed at Yale University”? Why did you use an aggregate indicator? What are the advantages and flaws of this indicator? What are the other available indicators and how did you benchmark them?

22- What were the other available ways to measure the state of the environment?

23- Could you analyse the environmental impacts by having a look at the sectors who are supposed to have grown and shrank in each country? Making an analysis of the specific impact of these most important sectors (in terms of land use, water use, water pollution, air pollution, CO2 emission, quantity of waste, quality of waste disposal), could you provide a deeper and more comprehensive analysis?

24- When breaking down the various EPI criteria, and the decomposition into “environmental health” and “vitality of the ecosystem”, one can observe several paradoxes. For example, the higher the water consumption, the better “water quality” is, and therefore “environmental health” is stronger. At the same time, it drops the “water resource” level and therefore the “vitality of the ecosystem” is lower, but overall the final indicator is higher. Overall, it appears that the indicator will be higher when consumption, pollution and environmental services are operating at high regime. Could you justify such a mechanism, given that it has not been proven that a higher GDP leads to a lower environmental impact, but studies tend to go the opposite way?

25- In terms of CO2 emissions, the model yields surprising reduction results overall, while it also results in growth. We know empirically that until today higher growth leads to higher emissions. We may think the observed reduction is due to a contraction in output, but the case of Tunisia contradicts that, since this country sees an important reduction of CO2 emissions while seeing its output and its GDP growth increase the most according to the model. How do you explain such a result? Could you also explain giving an insight on how the model is built?

26- If you believe this is especially due to the contraction/expansion of specific sectors, could you go further into the analysis by comparing it to the reality ie the actual emissions of the sectors that shrank to the sectors that grew, especially given that the model does not correctly assess the impact of the sectors’ evolution (as you state and as you show in some case studies)? In a nutshell, could you provide empirical evidence?

[1] The consortium is composed by Ecorys, FEMISE (Forum euro-méditerranéen des instituts de sciences économiques) and CASE (Center for Social and Economic Research).

[2] 6 agreements are evaluated, between the European Union and Algeria, Egypt, Jordan, Lebanon, Morocco and Tunisia.

[3] The interim report is available at

[4] Link to the declaration :

[5] See

[6] A consultation was held in each country and an online dialogue was held the 30th of April 2020 for organisations from Europe. A majority of participants were from the business sector.

[7] See our questions below. The evaluation team told us that they will answer these questions.

[8] A significant step was to recognize that the model used does not evaluate the impact of the agreements but gives an indication on their current value.

[9] In particular, the Environmental Performance Indicator (EPI) the study used is biased. It mechanically improves with a greater use of natural resources for instance. All the same, trade in “environmental goods” is analysed in a lengthy manner while such trade tells little on the reality of the environment quality of a country.

[10] For instance, the study states that the agreements would have allowed a decrease in CO2 emissions, while allowing for GDP growth. However, empirically GDP growth is a determining factor for CO2 emissions increases.



[13] Page 350

 source: FTDES