A Tale of (Almost) 1001 Coefficients: Deep and Heterogeneous Effects of the Eu-Turkey Customs Union

In view of the deferred start of negotiations for the modernization of the Customs Union between the EU and Turkey (CU-EUT), we look back and analyse the ex post trade consequences of the CU-EUT. Employing up-to-date econometric best practices for regional integration agreements, we quantify both total and heterogeneous trade effects of the CU-EUT. In contrast to most previous studies, our results indicate a significantly positive, large, and robust impact of the CU-EUT, implying an additional increase in EU-Turkey manufacturing trade by 55-65% compared to the previously active Ankara Agreement. We also provide evidence that the CU-EUT significantly increased Turkey’s trade with third countries. Additionally, a substantial heterogeneity in the CUEUT effect is found across different industries as well as for each of its member countries and the direction of trade. We link the heterogeneity of our up to 911 coefficient estimates to differences in initial trade costs and show that it cannot be ascribed to reductions in bilateral tariff rates.

Can regionalism do what multilateralism has so far failed to do-promote greater openness of services markets? Although previous research has pointed to the wider and deeper legal commitments under regional agreements as proof that it can, no previous study has assessed the impact of such agreements on applied policies. This paper focuses on the Association of Southeast Asian Nations (ASEAN), where regional integration of services markets has been linked to thriving regional supply chains. Drawing on surveys conducted in 2008 and 2012 of applied policies in the key services sectors of ASEAN countries, the paper assesses the impact of the ASEAN Framework Agreement on Services (AFAS) and the ambitious ASEAN Economic Community Blueprint, which envisaged integrated services markets by 2015. The analysis finds that over this period, ASEAN did not integrate faster internally than vis-à-vis the rest of the world: policies applied to trade with other ASEAN countries were virtually the same as those applied to trade with rest of the world. Moreover, the recent commitments scheduled under AFAS did not produce significant liberalization and, in a few instances, services trade policy actually became more restrictive. The two exceptions are in areas that are not on the multilateral negotiating agenda: steps have been taken toward creating regional open skies in air transport, and a few mutual recognition agreements have been negotiated in professional services. These findings suggest that regional negotiations add the most value when they are focused on areas that are not being addressed multilaterally.

OVERVIEW
Since the WTO's Doha Agenda failed to deliver meaningful services liberalization, many countries are turning to regional fora in the hope of greater success. That raises the question of whether regionalism can do what multilateralism has failed to do, i.e. promote greater openness of services markets. While previous research has pointed to the wider and deeper legal commitments under regional agreements as evidence that it can (e.g. Marchetti and Roy (2008), Fink and Molinuevo (2008), Mattoo and Sauve (2011), no previous study has assessed the impact of such agreements on applied policies. The ASEAN region is particularly interesting because it is widely believed to be at the frontier of what Baldwin (2011) has called "globalization's second unbundling," with regional integration of services markets linked to the and how far were they from meeting the market integration goal set out in the Blueprint?
Our stark conclusion is that the recent commitments scheduled under AFAS may have created greater regional policy-certainty but did not, in general, produce significant liberalization (because they were not sufficiently ambitious), and the 2015 deadline stipulated in the Blueprint did not instill the intended sense of urgency. One consequence of these agreements -which cover primarily intra-ASEAN trade -could have been that ASEAN countries would have been more open vis-à-vis each other than vis-à-vis non-ASEAN countries. For the most part, they were not. For the seven broad sectors (and relevant modes) covered by the questionnaire, there was surprisingly little difference between policy treatment of intra-ASEAN and extra-ASEAN trade. That is, the supposedly preferential policies vis-à-vis other ASEAN countries were virtually the same as non-preferential or "most-favored-nation" (MFN) policies vis-à-vis non-ASEAN countries. 1 Alternatively, the agreements could have promoted openness generally vis-à-vis all countries.
That too seems doubtful. First, because the ASEAN countries had on average more restrictive policies than any other region of the world except the Gulf states. The average Services Trade Restrictions Index (STRI) for the region was 60 percent higher than the global average. But The second reason that the agreements did not seem to have promoted openness generally was because the ASEAN countries' did not undertake significant liberalization in the four year period studied. While there were some instances of market opening, there were also instance of reversal of liberalization. For the six ASEAN Member States for which the same surveys were conducted in 2008, there was little change in the overall policy regime as of 2012 (regional average STRI fell only about 16 percent from its high level). As a consequence, even though actual openness was greater than that promised by current AFAS commitments, it was still not close to the ambitious goals specified in the Blueprint. But it is important also to recognize that the limited instances of reversal of liberalization could be because the AFAS commitments have served the valuable purpose of reducing policy risks.
There are two exceptions to these conclusions: in air transport, where ASEAN countries' have taken some steps towards regional open skies; and in certain professional services, where mutual recognition agreements have been negotiated. Even in these areas, the regional integration efforts were incomplete: in professional services, domestic regulations have not yet been aligned with the ASEAN MRAs; and in air transport further liberalization is necessary to achieve a truly integrated regional air market. Nevertheless, these initiatives suggest that regionalism could have 1 "Most-favored-nation" or MFN means the country which is the recipient of this treatment must receive equal trade advantages as the "most favored nation" by the country granting such treatment.
incremental value when it focuses on areas which are not being addressed multilaterally (Mattoo and Fink, 2004).
There are some broad caveats to our analysis. Market access in many of the countries was not predictable due to a discretionary licensing regime. From banking to transport, entry was restricted by the explicit and implicit limits on new licenses and the licensing process tended to be opaque and discretionary. In several ASEAN countries, licenses and foreign equity ownership were decided on a case-by-case basis, subject to requirements or approvals that involved several regulators and ministries. Some countries in certain sectors had no regulation at all, especially the lower income countries in the region and pertaining to the supply of services through the cross border and consumption abroad modes. In some of these cases, the supply of services was allowed in practice, while in others it was prohibited. In general, the high level of discretion and the absence of regulation created a less predictable policy environment and made it hard to define and assess the policy regime.
Section 2 describes the nature of services trade policy data as well as how it was collected and verified. Section 3 presents the ASEAN policy patterns and places them in a global context. Section 4 takes a closer look at the policy measures used by ASEAN countries, highlighting certain aspects of the regulatory environment. Section 5 assesses whether ASEAN countries liberalized their policies between 2008 and 2012. Section 6 examines instances of where ASEAN countries became more open vis-à-vis each other and provides two examples. Section 7 compares the regional and multilateral commitments of ASEAN countries with actual policy. Section 8 concludes.

SERVICES TRADE POLICY DATA AND MEASUREMENT
A detailed description of the original World Bank Services Trade Restrictions Databaseincluding details on the data collection process, the policy measures covered and the questionnaire used in the data collection-is provided in Borchert, Gootiiz and Mattoo (2012) and in supplementary material available at http://iresearch.worldbank.org/servicetrade. The global policy patterns of services trade policy emerging from the Database are presented in Borchert, Gootiiz and Mattoo (2013). Here we focus on the approach taken to updating information in 2012 for the six ASEAN countries covered in the original 2008 survey and to collecting information for the four ASEAN countries not previously covered. 2 The ASEAN survey in 2012 focused, as did the earlier surveys, on policies that affect international trade in services -defined, as is now customary, to include the supply of a service through cross-border delivery, consumption abroad, by establishing a commercial presence, or by the presence of a natural person. The perspective is one of a foreign supplier who wishes to provide services to a particular country, and we focus mainly on policy measures that discriminate against foreign services or service providers. transportation. Second, the questionnaire is designed to identify differences in intra-ASEAN and extra-ASEAN policy regime in services, and in particular, instances of regional liberalization and preferences. Third, we examine more closely than in earlier surveys whether there is in fact a regulation or policy in place for each specific subsector mode to take into account the conditions in countries like the Lao People's Democratic Republic and Myanmar. We also assess the implications of the absence of explicit regulation or policy, such as whether there are any implicit limits on the number of licenses allocated.
The seven major services sectors are disaggregated into further subsector-modes (as shown in Annex quantitative restrictions, such as those that limit the total number of providers, could hurt trade by preventing foreign entry even though they also limit domestic entry. Secondly, regulations such as qualification and licensing requirements ostensibly address the asymmetric information problem in certain services sectors but can impose a disproportionate burden on foreign providers such as professionals who have already met these requirements in their home countries. Thirdly, in some sectors the absence of regulations, such as those that ensure all (domestic and foreign) entrants have access to essential facilities such as ports and telecommunications networks, can be seen as a "sin of omission" because without such access entry may not be feasible. To cover each possible sin of commission or omission in all sectors is virtually impossible, but we attempt to include at least those which are likely to have a significant trade impact.
For each mode, the measures differ depending on the sector. In general, for mode 3 we have a core set of measures across sectors which are supplemented with sector-specific measures, for example limits on the size of loans in retail banking and restrictions on the international gateway in telecom. The core set of measures that pertain to mode 3 fall into the following four broad categories: requirements on the legal form of entry and restrictions on foreign equity; limits on licenses and discrimination in the allocation of licenses; restrictions on ongoing operations; and relevant aspects of the regulatory environment.
Measures governing mode 1 are slightly different in that they typically stipulate conditions under which cross-border trade may take place, rather than conditions imposed on the service provider.
Mode 4 measures, covered only in professional services, focus on qualification and (re-) certification requirements as well as entry and immigration rules, all of which strongly affect the movement of service supplying individuals. The challenge in evaluating policy measures is to assess whether prudential or regulatory measures affect contestability of the market by restricting entry of foreign suppliers (Findlay and Warren, 2000). While we make an effort to capture the important licensing regulations in professional services where they have a significant impact on trade, in future work more could be done to improve the coverage of such measures in areas like financial services. Finally, to understand how the policy was measured and became the services trade restrictiveness indices (STRI), please see the detailed note on the scoring in the Appendix.
First-hand information for ASEAN member states was collected by administering a questionnaire in 2012 which was completed by local law firms familiar with the policy regime in the respective countries and sectors. The information on policies received was evaluated, and its restrictiveness assessed, by a team of World Bank economists. To ensure data accuracy, the policy information was reviewed by government officials between March and May of 2013.
Upon receiving government comments, the policy information and scores have been revised.
This paper is based on the data that have been reviewed and validated by the government officials.
It is notoriously difficult to measure policies affecting services trade because of their variety and complexity (see, for example, Hoekman (1996) and the overview by Deardoff and Stern (2008)).
We rely on a measure of the restrictiveness of a country's policy regime for any subsector-mode, the "services trade restrictions index" (STRI), which has the weakness of being subjective but the virtue of being simple, transparent and robust (Gootiiz et al. 2013). 6 This measure is most 6 The OECD has also developed a measure of services trade restrictiveness drawing upon the more detailed data available for industrial and more advanced developing countries (OECD 2009(OECD , 2011 convenient to depict overall patterns in policy, across countries, modes, and sectors.

HOW OPEN ARE THE SERVICES MARKETS OF ASEAN COUNTRIES?
The comparison of STRIs shows that ASEAN was on average one of the more restrictive regions in the world. The average Services Trade Restrictions Index (STRI) for the region was 60 percent higher than the global average. Figure 1   sectors. 7 It is useful to look more closely at the nature of these policies. 7 We focus here on the five main sectors. Annex Figure A1 includes also education and medical services.
Financial services: Bank sector policies in Thailand, Philippines, Malaysia, and Vietnam were more restrictive than in the other countries, because they restricted greenfield entry and operations. In Thailand, the limit on foreign ownership in a "local bank" was 49 percent. A foreign-owned subsidiary faced no limit on foreign equity participation, but there was a limit on operations: the number of branches and ATMs allowed per subsidiary was 20. Foreign bank branches could operate up to 3 branches or off-premise ATMs without a location limit. In Philippines, greenfield entry was not possible since the license limit of 10 had been reached, and for acquisition, the foreign ownership limit was 60 percent. In Malaysia, primary entry as a branch was not allowed and entry through a subsidiary was temporarily not allowed, as no new licenses were being issued, although there was no limit on foreign ownership in a subsidiary. For acquisition, the limit was 30 percent and there was a restriction on expanding through additional branches; 10 microfinance branches were allowed per bank and further branches were allowed based on the effectiveness of these branches in serving microenterprises. Vietnam allowed wholly foreign-owned subsidiaries but imposed a limit on the acquisition of banks. To acquire existing banks, the foreign ownership limit was 30 percent for aggregate foreign investment and 20 percent for a single foreign credit institution.
In automobile and life insurance, Malaysia, Thailand, Lao PDR and Myanmar had restrictive policies. Myanmar was still drafting its regulations on the insurance and reinsurance sector, and it was not possible to enter at that stage. In Thailand, the licensing regime was discretionary; the Minister of Finance gave approval for the license and equity participation. In Lao, there was a 49 percent limit on foreign ownership, and the licensing regime seemed burdensome as approval from the Ministry of Planning and Investment, Ministry of Finance and Ministry of Industry and Commerce were required. Malaysia did not allow entry via a branch. Foreign ownership in a subsidiary could be 100 percent, however, no new licenses were being issued but were announced from time to time on an ad hoc basis. The foreign ownership limit in acquiring a share of an existing insurance company was 70 percent. In cross-border reinsurance, Malaysia had a restrictive policy: companies need to demonstrate domestic unavailability in Malaysia before obtaining services abroad. The Philippines required 10 percent of reinsurance to be ceded to the National Reinsurance Corporation of Philippines. Note: The services trade restrictions index (STRI) at the regional level is calculated as a simple average of individual country's STRIs. The STRI in the cross-border air passenger transport subsector is excluded. Regional abbreviations: GCC -Gulf Cooperation Council, ASEAN-ASEAN Member countries, SAR -South Asia Region, MENA -Middle East and Northern Africa, AFR -Sub-Saharan Africa, LAC -Latin America and Caribbean, OECD -High income OECD, ECA -Europe and Central Asia. The financial STRI includes scores for retail banking mode 1 and mode 3, auto mobile, life, and reinsurance mode 1 and mode 3 respectively. Telecom STRI includes scores for fixed line and mobile. Retailing STRI includes scores for retailing mode 3. The transport STRI includes STRIs for air passenger international mode 3, maritime international mode 1 and mode 3, road freight mode 3, and rail freight mode 3. Professional services STRI includes scores for accounting, auditing, legal advisory on domestic law and foreign law in mode 1, mode 3, and mode 4. For comparability, the STRI scores for education, medical services, and some other professional services subsectors are excluded.  Philippines, a foreign retailer needed to bring in a paid-up capital of USD 2.5 million or more, provided that investments for each store must be not less than USD830 thousand. In Vietnam, establishing an outlet beyond the first one was considered on a case-by-case basis and approval depended on the number of outlets, market stability, population density, and consistency of the investment project with the master plan of the city, where the shop was planned to be set up.
Malaysia also had a minimum capital requirement that foreign retailers needed to meet. In these cases, domestic retailers did not have the same requirement.
Transportation: Transportation services were relatively restricted in ASEAN countries as they were in other parts of the world. In cross-border (mode 1) maritime shipping, we examined restrictions on both private and government cargo. Thailand, Vietnam, Philippines, and Malaysia had restrictions on foreign ships carrying government cargo but no limitations on private cargo.
On commercial presence (mode 3), for the types of transport covered by the survey (maritime, air, road, and rail), the majority of Member States mentioned that the control must be held by local companies. In air transport, the member states signed the ASEAN Multilateral Agreement on Full Liberalization of Air Freight and Air Passenger services. It was difficult to assess how much more openness the regional air services agreements offered beyond the existing Bilateral Air Services Agreements (BASAs), which are discussed in more detail below.
Education and medical services: These services were covered most comprehensively, as all modes of supply were included in our survey: cross border (mode 1), consumption abroad (mode 2), commercial presence (mode 3), and presence of natural persons (mode 4). Not surprisingly, most countries were fairly open in mode 1 and mode 2 types of trade in education and medical services. In mode 3, Thailand, Indonesia, Myanmar, and Philippines had restrictive policies. In these countries, the control of such institutions was required to be held by nationals and in the case of Myanmar, the medical and higher education services were run by state-owned institutions. In Philippines, medical services were run by the state and the educational institutions were required to be owned and operated by Philippine nationals only. In mode 4, Lao PDR and the Philippine required medical and educational services to be provided by the nationals. Other countries were surprisingly open in the supply of services through mode 4.

Professional Services:
The supply of accounting, auditing, legal advisory service on foreign and domestic laws, architectural, engineering, and management consulting services were covered through mode 1, mode 3, and mode 4. Although countries differed in their policies, it appeared that most countries had fewer restrictions on management consulting, accounting, legal advice on foreign law, architecture, and engineering services than on auditing and legal advice on domestic law. In many of the countries, the cross -border supply of services (mode 1) was unregulated and open. In mode 3, the countries had restrictions on ownership, organization, and practices. Indonesia did not allow investment in most of its professional services sectors. In mode 4, Thailand and Philippine were quite restrictive: Thailand did not allow entry via mode 4 in all of the professional services sectors covered and the Philippines allowed entry subject to restrictive conditions including reciprocity and labor market tests.

IS ASEAN POLICY RULE-BOUND OR DISCRETIONARY?
Besides the overall policy pattern by sector, we looked at a few of the key policy measures the survey covers in each sector: legal form of entry and ownership, licensing regime, and regulatory environment. The central features that emerged are the restrictiveness of policy and the high degree of discretion in the policy environment relating to the entry and operation of foreign firms.

Legal form of entry and foreign ownership
As we saw in the previous section, the restrictiveness of legal form of entry and ownership varied by sector and country, but certain broad trends emerged. In general, higher foreign ownership was allowed in a greenfield subsidiary than in entry through acquisition (Annex Tables A6.A, A6.B and A6.C). Many countries allowed full foreign ownership in a subsidiary, but full foreign ownership did not actually mean liberal conditions of entry, since licensing could still be restrictive. For example, even though Malaysia allowed 100 percent foreign ownership in banking and life insurance, new licenses were not being issued. In Thailand, insurance sector licenses were subject to review by several authorities, including the Office of Insurance Commission (OIC), on a case by case basis. In case more than 49 percent of foreign equity was desired, the approval of the Ministry of Finance (MOF) was required upon the recommendation by the Commission; the Minister had the power to grant a license with the approval of the Cabinet.
Compared to Greenfield entry, there were stricter limits on foreign ownership via acquisition of an existing entity, especially if the entity was state-owned. Myanmar and Philippines did not allow foreign acquisition of state-owned entities in most of the sectors covered. However, because acquisition was not subject to new licensing requirements, conditions of entry through this legal form may in fact have been more liberal than greenfield entry. Across countries, the foreign ownership limit was more restrictive in transportation sectors. The only country that allowed full foreign ownership of a state-owned entity was Singapore.

Licensing regime
The licensing regime is vital but it was difficult to assess whether licensing measures were applied for prudential or protectionist reasons. In most countries, licensing and market entry criteria were publicly available but fulfilling publicly available criteria did not ensure that a license was granted. All countries except Vietnam indicated licensing was not automatic in at least several sectors. Instead, licenses tended to be issued on a case-by-case basis. Measuring the discretionary element in licensing was, of course, difficult.
In only a few countries did we observe an explicit licensing limit or a hard quota-type of restriction but the discretion of the licensing authority could be used to implement implicit limits. Many countries also maintained different licensing criteria for foreign and domestic firms, but most such differences were relatively minor, such as an additional document or minimum capital requirement.

Regulatory environment
Our survey also covered several aspects of the regulatory environment, of which we describe three: whether the regulator was required to provide reasons for license rejection; whether there was a right to appeal the decisions of the regulatory authority; and whether regulators provided prior notice of regulatory and policy changes. Countries in which the regulators were required to provide reasons for rejecting licenses would presumably have had less room for discretion.
Having a right to appeal was also connected to the licensing regime and indicated whether there the private sector had recourse to a remedial process. The survey results showed that Myanmar, Malaysia, Philippines, Singapore and Thailand did not require regulators to provide reasons for license rejection. Appeals were not allowed in Cambodia, Myanmar, and Malaysia. Prior notice helped the private sector prepare for policy changes and may even have allowed for private sector input into policy. Indonesia, Philippines, Thailand, Myanmar, and Lao did not have processes for prior notice to the private sector.
In low-income countries, where there was less institutional capacity, the complete absence of a formal policy or regulation was not uncommon. The most number of subsector-modes that were not covered by any specific regulation or policy appeared in Lao PDR, followed by Cambodia, Myanmar, and the Philippines, and it was a phenomenon mostly observed in modes 1 and 2. The absence could have had a restrictive impact, since it reduced transparency and predictability of the policy regime and increased the potential for discretion. But in many of these cases, in practice the supply of a service was allowed, with Vietnam a notable exception in this respect.
Some other dimensions of the regulatory environment that the survey covered are described below.

DID ASEAN MEMBERS BECOME MORE OPEN BETWEEN 2008 AND 2012?
As noted above, the surveys were conducted in both 2008 and 2012 for six ASEAN countries: Cambodia, Indonesia, Malaysia, Philippines, Thailand, and Vietnam. Over this period, the data shows that six countries' policies on average became more liberal but the change is very small ( Figure   3). To identify the policy change, we need to look at the subsector mode level, because the country level score is a weighted aggregation of subsector-mode scores. For example, Indonesia's overall STRI increased by 7 points between 2008 and 2012. Even though the restrictiveness index went down in 5 subsector-modes, it went up in 3 subsector-modes, including the relatively important retail sector.

WAS ASEAN INTEGRATING FASTER INTERNALLY?
One of the goals of the survey, as described in Section 2, was to identify instances where ASEAN countries treated services and/or services providers from their regional partners differently from those whose provenance was outside the region. In fact, neither the law firms that collected the policy data nor the governments that verified the data could identify any meaningful instances of differential treatment. For the seven broad sectors (and relevant modes) covered by the questionnaire, the supposedly preferential policies vis-à-vis other ASEAN countries were virtually the same as non-preferential (or "most-favored-nation" (MFN) policies vis-à-vis non-ASEAN countries.
In professional and transport services, liberalization initiatives naturally tended to be among two or a few countries because the regulatory framework favored reciprocal arrangements, such as recognition of qualifications and negotiation of traffic rights. ASEAN countries had taken initiatives in both these areas and to illustrate their impact we assess below the intra-ASEAN openness in architectural and engineering services via mode 4 and air transport services via mode 1.  Table A7). We found two problems: in some states, the restrictive domestic regime had not been reformed to align it with the relatively liberal MRAs; in other states, the liberal domestic regime was already more liberal than the MRAs.
The first problem was the lack of domestic regulatory reform needed to support the specific The survey revealed that in some other respects, member states already had quite liberal regimes for foreign licensed architects and engineers (Annex Table A8). In these respects, it appeared that being an ASEAN licensed professional conferred no additional advantage since the MFN regime was already liberal. To illustrate this aspect, we compare one condition of the MFN regime with the comparable condition in the MRA: work experience (Table Annex A9). The ASEAN MRA on architectural services required at least 10 years of experience; but the MFN regime shows that four countries (Indonesia, Vietnam, Myanmar, and Lao PDR) did not require any work experience, and two countries (Philippines and Malaysia) required 2 and 5 years of experience, and one country (Singapore) required 2 to 10 years of experience. In engineering services, we observed the same pattern, where the MFN regime was more relaxed than the ASEAN regime. Hence, although the MRAs are potentially an important step in the regional integration in professional services, there still appeared to be limited benefits of being registered as an ASEAN professional -a conclusion which accords with the findings of Aldaba (2013) and Hirawan and Triwidodo (2012).

Cross-border air transport
Compared to other sectors, ASEAN member states appeared to have made progress in the regional integration of their air transport markets. 8 They had signed "multilateral" air transport agreements that were more liberal than their previous restrictive bilateral air service agreements. However, regional integration was incomplete, as there were number of areas that needed to be further liberalized to achieve a truly integrated regional air market. In fact, some members had moved ahead and individually concluded bilateral agreements with certain OECD countries that could be even more liberal than the agreements with other member states.
Most bilateral routes within ASEAN had been liberalized by the ASEAN multilateral agreements. The only exception were routes into and out of those member states that had not fully accepted the air transport agreements and these remained governed by restrictive bilateral air services agreements (BASAs). The formal ASEAN agreements on air transport liberalization were the "Multilateral Agreement on Air Services" (MAAS), the "Multilateral Agreement for Full Liberalization of Passenger Air Services" (MAFLPAS) and "Multilateral Agreement for Full Liberalization of Air Freight Services" (MAFLAFS) and their respective Implementing Protocols. These multilateral agreements go beyond the BASAs in two important aspects: the agreements allow 3 rd , 4 th , and 5 th freedom rights for air carriers between designated secondary cities and all capital cities of ASEAN Member States; 9 instead of substantial ownership and control by the nationals, the "community carrier" concept is in principle allowed. This means an airline can be substantially owned and effectively controlled by ASEAN interests taken cumulatively or in the aggregate (Tan, 2009). This provision allows airlines to attract capital infusions and management expertise from multiple sources within ASEAN.
However regional integration appeared incomplete in certain key respects. To achieve a truly integrated aviation market, further liberalization, such as the seventh freedom (the right to fly between two ASEAN countries while not offering flights to one's own country), eighth freedom (the right to fly between two or more airports within an ASEAN country while continuing service to one's own country), and ninth freedom (the right to fly inside a ASEAN country without continuing service to one's own country) are all necessary. A single aviation market such as that which exists in the European Union (E.U.) liberalizes such operations fully and allows market competition throughout the region. But in ASEAN countries, domestic carriage remained highly sensitive for large countries with a large domestic population. Typically, such operations were reserved exclusively for local players, and most ASEAN governments upheld that status quo.
As far as ownership and control were concerned, although in theory a "community carrier" was allowed to operate, in practice, it was still difficult. There was no certainty that the "community carrier" could fly into all member states in the region, as the Member States still needed to provide an approval before the carrier could operate. Unless a significant number of member states first declared their unequivocal approval for such a model, the current uncertainty could discourage any investor thinking of establishing such an airline. At the time of this study, investors complied with the traditional "substantial ownership and effective control of nationals" rule.
Finally, even though the agreements were in force among most ASEAN member states, key states such as Indonesia and Philippines remained outside the scope of the agreements.
Indonesia's position on the ASEAN agreements could be traced to its leading carriers' lobbying of their government to continue protecting their international operations against those of rival airlines from neighbouring ASEAN states. Through their industry group, the Indonesian National Air Carriers Association (INACA), the major carriers had traditionally opposed efforts to open up the ASEAN air travel market. The Philippines' reluctance was related to the lack of airport slots and infrastructural constraints.
However, there are several factors that may provide the momentum for change beyond 2015.
First is the growing confidence of Indonesian carriers such as Garuda and Lion Air. As these airlines expand their services and increase their competitiveness and appeal to passengers, there may come a time when they feel more secure and see less of a need to resist greater liberalization. Second, there is the pressure created by the provincial governments, tourism authorities and business community to allow greater direct access into secondary cities. Third, there is the pressure created by the agreements with larger countries such as China. The Member States are realizing that without a truly single market agreement internally, negotiating with large countries like China may be difficult for the ASEAN Member States. Fourth, innovative airlines have sought to get around the restrictions, including those that are cast in the bilateral and multilateral agreements. One example is how AirAsia pioneered the cross-border joint venture model -while still imperfect, it allows AirAsia to get around the "seventh freedom" prohibition and to operate region-wide from multiple hubs using a common, well-recognized brand. (2013)

REGIONAL AND MULTILATERAL COMMITMENTS AND GOALS
Having examined applied policies, we are in a position to make three sets of comparisons: between regional commitments and goals, and actual policy; between multilateral commitments and offers, and actual policy; and, finally, between the regional and multilateral dimensions.

The regional dimension
10 The other parties to MALIAT are Chile, Cook Islands, New Zealand, Samoa and Tonga. 11 These horizontal agreements do not alter the capacity provided for in the existing bilateral agreements.
The ASEAN Framework Agreement on Services (AFAS), signed in 1995, is one of the first regional trade agreements in services. AFAS is closely related to the GATS and follows its main principles, disciplines and approach to liberalization. It contains liberalization commitments that aim to reduce restrictions to services trade between the ASEAN Member States and calls for the It is evident from Table 1 (and Annex Figure A2) that all ASEAN member countries' applied policies were more liberal than their AFAS commitments, though the size of the gap varied across countries and sectors. Indonesia and Vietnam are examples of countries where there was virtually no gap, whereas Singapore, Cambodia and Myanmar had policies that were much more open than their commitments. In terms of sectors, Table 2 shows that the gap was particularly large in financial (especially banking) and education services with commitments more than twice as restrictive as policy, and quite small in transport and medical services.

Comparing the regional and multilateral dimensions
To bring all the dimensions together, ASEAN countries' GATS commitments and Doha offers were far more restrictive than their AFAS commitments (with a gap of about 23 STRI points).
Thus, there is no doubt that the countries -especially Myanmar -have displayed a far greater willingness to widen and deepen their legal bindings in the regional than the multilateral context.
The two exceptions are again the recently acceding countries, Cambodia and Vietnam, which have made WTO commitments close to their AFAS commitments. But the ASEAN countries' AFAS commitments themselves remained more restrictive than applied policies, as we saw above. And the gap between applied policy and Blueprint 2015 was still large (about 20 STRI points).    Did not submit an offer.
None, except subject to requirement for local shareholding of up to 49%.
By 2015, allow for foreign (ASEAN) equity participation of not less than 70%.
No restrictions except when acquiring a state owned entity, the foreign equity limit is 49%.

Indonesia
Local services: Provided exclusively by PT Telecom until 2011. International: Provided exclusively by duopoly, expires 2005. Foreign equity limit is 35% and must be in form of a joint venture. At the end of each period, government decides whether to permit additional suppliers.
Offer is similar to GATS commitment.
Only through joint venture with local private sector. Foreign equity limit is 49% By 2015, allow for foreign (ASEAN) equity participation of not less than 70%.
There are no restrictions except foreign ownership limit is 49 percent.

Lao PDR
Only through acquisition of existing operators and foreign equity participation limit is 49% for 5 years after the date of accession. Thereafter, commercial presence is allowed with foreign equity limit of 60%.
Did not submit an offer.
Local and national long distance services can be supplied only on a facilities basis for public use. 100% foreign owned or joint venture enterprise is allowed.
By 2015, allow for foreign (ASEAN) equity participation of not less than 70%.
Foreign ownership limit in a state owned entity is 49.9%.

Malaysia
Entry allowed only through acquisition. Foreign equity limit is 30%.
With respect to network facilities and services provider: only through acquisition and foreign equity limit is 30%. The management must be controlled by Malaysians. For Telekom Malaysia, the foreign equity limit is 30% in aggregate with no single country holding more than 5% of the equity at any one time.
Only through acquisition of shares of existing licensed public telecommunications operators, foreign equity participation limited to 51% in such providers.
By 2015, allow for foreign (ASEAN) equity participation of not less than 70%.
Entry is possible only through acquisition (no new license is allowed). The foreign equity limit is 49%.

Myanmar
No commitment. Did not submit an offer. No commitment.
By 2015, allow for foreign (ASEAN) equity participation of not less than 70%. Progressively remove other Mode 3 market access limitations by 2015.
Foreign Investment Law (2012) allows commercial presence of foreign service suppliers. No restrictions on the foreign equity participation.

Philippines
Franchise from the Congress and Certificate of Public Convenience and Necessity from the National Telecommunications Commission required. Foreign equity limit is 40%. All executives and managers must be citizens and limit on the share of foreigners in BOD is 40%.
Offer is similar to GATS commitment.
Entry is subject to following requirements and conditions: franchise from Congress of the Philippines; certificate of Public Convenience and Necessity from National Telecommunications Commission (NTC); foreign equity permitted up to 40%; and resale of private leased line is not allowed.
By 2015, allow for foreign (ASEAN) equity participation of not less than 70%.
Entry as a branch or acquisition of state-owned entity is not allowed. Foreign ownership limit is 40%. Nationality requirement for BOD is 60%.

Singapore
Up to two additional operators will be licensed in 1998 for the provision of these services commencing By 2015, allow for foreign (ASEAN) equity participation of not less than 70%.
Foreign majority owned or controlled providers may only offer services on a resale basis, and such offerings are limited to certain defined services. No other restrictions.

Vietnam
Facilities-based: Upon accession, joint ventures with telecom service suppliers duly licensed in Vietnam will be allowed. Foreign equity limit is 49%. Non-facilities based: Upon accession, joint venture with telecom service suppliers duly licensed in Vietnam with foreign equity limit of 51% is allowed.
Did not submit an Offer.
None, services must be offered through commercial arrangements with an entity established in Viet Nam and licensed to provide international telecommunication services.
By 2015, allow for foreign (ASEAN) equity participation of not less than 70%.

CONCLUSIONS
The main contribution of this paper is the collection and analysis of rich applied policy information for ASEAN countries. Four gaps in data limit the scope of the present analysis and should be the focus of future data collection and research. First, we do not have adequate data on the existing market structure -e.g. the number of firms, their market share and ownershipacross sectors and countries, which means that our policy measures capture the restrictions on entry into markets but do not capture the prevailing extent of competition between domestic and/or foreign firms. Second, we do not have good data on outcome variables such as prices, quality or diversity of services, which makes it hard to infer the restrictiveness of policies by econometrically analyzing their impact on outcome variables of interest. Third, we are able to capture only limited information on the state of prudential and pro-competitive regulation, which makes it hard to assess how far these ostensibly non-discriminatory measures offer de facto protection to domestic service providers. More importantly, this gap makes it hard to assess how far the gains from market-opening depend on the state of complementary regulation, and we emphasize that a mechanical elimination of trade barriers without reform of complementary regulation is not necessarily desirable. Finally, we capture only limited information on the implementation of policies. For instance, we make an effort to identify certain aspects of the processes involved in licensing services providers, such as transparency and accountability, but the process remains opaque and it is hard to determine whether the processes in themselves offer protection to domestic providers. In some cases, the absence of laws and regulations makes it difficult to assess actual practice, and we do not know whether a de jure vacuum signifies de facto openness or prohibition.
Despite these limitations, we were able to reach some clear conclusions. First, the ASEAN member states had on average more restrictive policies than most other regions and the pace of recent reform has been slow. Furthermore, with regard to the explicit restrictions, there was little sign of preferential treatment of any ASEAN member state of other Member states. The absence of preferences is not a problem; the absence of reform is. Member states needed to reduce rapidly the remaining explicit barriers to foreign entry and ownership -ideally on an MFN basis -to achieve the Blueprint Goals by 2015.
Second, regionalism offers a potentially valuable avenue for liberalization in areas where multilateral cooperation is difficult, such as in professional services and transportation. ASEAN member states had made some progress in deepening regional integration in these areas, however, the efforts were incomplete. ASEAN member states needed to reform domestic regulations in professional services to better align them with the MRA provisions, and to make the MRA provisions as liberal as their respective MFN regimes. In air transport, there was a need to further liberalize the freedom of rights by allowing the 7 th freedom and eventually even cabotage, and making the community ownership of designated airlines automatic rather than discretionary.
Third, successful liberalization requires supporting reform of domestic regulation, ranging from prudential regulation in financial and professional services to pro-competitive regulation in telecommunication and transport services. In these areas too, there was scope for regional coordination and cooperation, to reap economies of scale in regulation and to prevent the fragmentation of the regional market because of divergent national regulation (Mattoo and Sauve, 2011).
Finally, the reform process needs to be monitored, transparent and informed by sound analysis.
For all these reasons, ASEAN countries must move to remedy the weaknesses in the current state of data identified above. In particular, an effort must be made to collect better data on the implementation of reform in all dimensions -ranging from liberalization to improvement in regulation -as well as the consequences of reform for market structures and market outcomesrelating to the prices, quality, diversity and access to services. Such data would facilitate analysis of both the gains from reform and design of reform, which could make future reform socially desirable and politically feasible.

Regionalism in Services:
A Study of ASEAN ONLINE APPENDIX

ANNEX SECTION 1: MEASURING SERVICES TRADE POLICY, STRI
It is notoriously difficult to measure policies affecting services trade because of their variety and complexity (see, for example, the overview by Deardoff and Stern, 2008). We develop a measure of the restrictiveness of a country's policy regime, the "services trade restrictions index" (STRI), which has the weakness of being subjective but the virtue of being simple, transparent and robust.
This measure is most convenient to depict overall patterns in policy, across countries, modes, and sectors. It builds on a relatively long tradition of restrictiveness indices, ranging from simple counts of policy barriers (Hoekman, 1996) to more complex weighted averages, where weights reflect prior (usually subjective) assessments of the relative restrictiveness of specific policy barriers; work currently being undertaken at the OECD 12 uses an elaborate version of this method, which is described in OECD (2009a).
We construct a single measure of overall openness for any subsector-mode combination, e.g. one for the cross-border supply of bank loans and another for accepting bank deposit by establishing commercial presence abroad. This measure avoids the pitfalls of the approaches that assign fixed weights to all types of restrictions (entry, operational, regulatory) and that treat the restrictions as additive. For instance, if foreign suppliers are not allowed to enter in the first place, then that restriction is binding and other restrictions on operations and regulatory environment simply do not matter. Similarly, a foreign equity limit of 49 percent already precludes foreign corporate control and so adding to it a further (frequently encountered) requirement that the majority of board of directors be nationals would amount to double counting. inconsistent, e.g. license length or license allocation mode, were not considered for scoring as cross-country differences would reflect response rates or interpretation differences rather than differences in restrictiveness.
It is convenient to assign a value to each of these five categories of regimes on an openness scale from 0 to 1 with intervals of 0.25. 14 We call the resulting score a services trade restrictions index Annex Table A1 show, most policy regimes have more than one provision in place per sub-sector and mode of supply, in which case the assigned score (shown in the right-most column) reflects the overall restrictiveness of all measures evaluated 14 At this level, basic STRI scores are no more than 'labels' attached to the five ordered categories of restrictiveness. However, as soon as these scores are further processed, either by aggregation or by use in a quantitative model, the specific values assume a cardinal meaning that implies the five categories are 'equidistant' in terms of restrictiveness. The working paper version of this article (Borchert, Gootiiz and Mattoo, 2012b) discusses an alternative approach of ranking policy bundles purely ordinally.

(STRI). As the examples in
simultaneously. 15 Since the STRI focuses mainly on the set of measures which discriminate against foreign services and providers, the greatest level of openness is associated with a value of zero. However, since the STRI does not adequately cover complementary areas of nondiscriminatory prudential and pro-competitive regulation, and since it is likely that the results of liberalization depend on the state of these types of complementary regulation, we cannot say that a zero level of STRI is necessarily immediately desirable from a broader welfare or development perspective.
Annex Virtually open 0.25 Vietnam: Life insurance -mode 3 "Entry as a branch is not allowed. No restrictions on foreign ownership in Greenfield subsidiary or acquisition of existing entity." Existence of major/non-trivial restrictions 0.50 Thailand: Air passenger domestic -mode 3 "The limit on foreign ownership is 49 percent, with effective control by Thai nationals. At least 40% of Board of Directors must be national." Virtually closed 0.75 Malaysia: Reinsurance-mode 1 "Reinsurance companies must demonstrate domestic unavailability in both Malaysia and Labuan before obtaining services abroad." a.

Completely closed 1
Philippines: Architecture services -mode 3 "Commercial presence is not allowed." Notes: As is apparent from the examples shown, most subsector-mode combinations are characterized by multiple provisions, in which case the regime assignment reflects the overall restrictiveness of all applicable measures.
Once a score has been attached to each regime, STRI values can be aggregated across sectors and modes of supply. Let denote the basic scores on a 5-point scale per sub-sector and mode of supply as described in Table 2. In order to arrive at an aggregate STRI of country c, , we 15 The Database Guide (Borchert, Gootiiz and Mattoo, 2012a) contains in its Section 4 three examples-from Burundi, Thailand and India-that illustrate how a portfolio of several measures is being assigned to one of the five basic scores. In principle, policy measures can be divided into two tiers. The first tier measures include those that affect market entry decisions most significantly, such as a limit on foreign ownership and the number of licenses. The second tier measures are those that affect operations of service providers, such as restrictions on the repatriation of earnings. The second tier measures do not contribute to overall restrictiveness when first tier measures are prohibitive. In contrast, if the first tier measures are not prohibitive then second tier measures are also considered in determining the overall restrictiveness score. jmc s c STRI begin by taking weighted averages across modes of supply , whereby the set of modal weights is specific to sector j. The sectors differ in the relative importance of alternative modes for delivering a specific service. For instance, in a 'consumer service' such as life insurance, a higher modal weight is attached to commercial presence than in the reinsurance sector in which cross-border provision amongst firms is the dominant mode of supply. Formally, the sectoral scores are given by Sectoral scores are then aggregated across all sectors using weights that reflect the relative importance of constituent services sectors in domestic value added. Sector weights are based on services sectors' standardized share in total services output for an 'average' industrialized country. 16 Overall country-level scores are obtained as The complete weighting schemes used to aggregate modes, subsectors and sectors, respectively, can be found in Annex Table A.7, including further details regarding the sectoral weights. All scores at any level of aggregation are available from the 'STRI' section of the database; in particular, the full set of baseline values is accessible so that users are free to devise alternative aggregation schemes if they so wish.
We recognize the subjectivity of this approach but given data constraints and the wide range of sectors covered, there is no obviously superior method of quantification. A demonstration that the STRI assessments are broadly corroborated by alternative methods of quantification can be found in the working paper version of this paper (Borchert, Gootiiz and Mattoo, 2012b). The subjectivity of the STRI is somewhat mitigated by the extensive consultations we have conducted with the private sector and regulators in making the assignments of weights to specific categories and developing a scoring rule sheet, which sets out how measure specific restriction is scored. We 16 A sense of how sectors are over-/underweighted in low-income countries can be gleaned from the fact that the share of financial and business/professional services tends to rise with income whereas the share of retail distribution and, to some extent, telecommunications services tends to decline with income. However, for the STRI to be comparable across countries, we need to use one uniform set of weights for all countries (see Annex 1 for further details).
believe that the adopted approach is at this stage more suitable than any fixed algorithm to turn the rich and difficult-to-quantify aspects of policy information into a broadly plausible if somewhat imprecise restrictiveness scores. In Paul Krugman's words, it has the virtue of being "roughly right rather than precisely wrong."

Weighting Schemes for the STRI Index
Annex Table A2 below documents the sets of weights used to derive aggregate, country-level STRI scores, , from basic scores per sub-sector and mode, . Modal weights sum up to unity within any given subsector, e.g. 'accounting' (all subsectors are listed in Table 1).
Subsectors are aggregated to the sectoral level, e.g. 'telecommunications,' using simple averages.
Sector scores are aggregated to the country level using standardized weights based on the constituent services sectors' share in total services output for an 'average' industrialized country.
The service sector output shares are taken from Hoekman (1995, p.37/Appendix 1) and scaled so as to make the weights of all sectors covered in the Services Trade Restrictions Database add up to unity. We recognize that services sectors command a different share in total services output in different countries, especially across developing and developed countries, and are at least in part influenced by policy restrictions. As an empirical regularity, the share of financial and business/professional services tends to rise with income whereas retail distribution and, to a lesser extent, telecommunications services occupy a larger share in poorer countries. However, comparability requires the use of one uniform set of weights for all countries. We chose to use the shares for an 'average' industrialized country because industrial countries tend to be more open and so shares are less likely to be distorted by restrictions. Notes: The sector weights are used for constructing country STRI, there are no sector weights reported for education and health services, because these sectors were not covered in 2008 surveys and for comparison with 2008 country level STRI, the 2012 country level STRIs don't aggregate STRIs for education and health. The STRI for midwives and physical therapists services via mode 4 was aggregated with medical dental services via mode 4 with equal weights.
For most sectors the deviations between the set of weights used and weights representative of low-income countries are not large; if anything, scores in the retailing sector are underweighted whereas professional services scores are somewhat overweighted.

ANNEX SECTION 2: ASEAN MULTILATERAL AGREEMENTS IN AIR TRANSPORT 17
The discussions on regional air services agreements started in November

What is allowed under the agreement?
Overall, the objectives of the agreements remain fairly modest -market access relaxations stop simply at the third, fourth and fifth freedoms, and do not extend to the seventh, eighth and ninth freedoms. 18 More specifically, the MAAS Implementing Protocols specify the following "third", "fourth" and "fifth" freedom market access rights: In terms of air traffic volume and market potential, Protocols 5 and 6 have much greater economic significance as these cover the ten capital cities and are not limited by sub-regional proximity. Specifically, Protocol 5 provides contracting states' designated carriers with unlimited third and fourth freedom opportunities between their own capital city and all the other ASEAN capital cities. Protocol 5 further provides that such rights shall be allowed by 31 December 2008 (although, as noted above, Protocol 5 was actually adopted only in May 2009). On its part, Protocol 6 lays down a deadline of 31 December 2010 for a contracting state's designated carriers to operate full third, fourth and fifth freedom rights from their capital city to other contracting states' capital cities (e.g. a Malaysian carrier from Kuala Lumpur to Hanoi with fifth freedom pick-up rights in Phnom Penh).
The MAFLPAS Implementing Protocols address the following market access rights:  Protocol 1 -Unlimited Third and Fourth Freedom Traffic Rights Between Any ASEAN

Cities  Protocol 2 -Unlimited Fifth Freedom Traffic Rights Between Any ASEAN Cities
The MAFLPAS agreement was designed to supplement MAAS and to cover the rest of the ASEAN cities. Hence, MAFLPAS Protocol 1 allows for unlimited third and fourth freedom operations for state parties' carriers between two non-capital cities, or between a non-capital and a capital city (capital-to-capital operations remain governed by MAAS Protocol 5). MAFLPAS Protocol 2 provides for unlimited fifth freedom operations involving the non-capital cities. By its terms, Protocol 2 can also cover flights involving capital cities, except when all three points are capitals, in which case MAAS Protocol 6 governs.

Ownership and control
How have the ASEAN multilateral agreements sought to deal with these ownership and control restrictions? In essence, when both market access and ownership and control are freed up, the Member States can achieve the true single air market. On top of prohibiting seventh freedom and domestic operations by foreign carriers, the current ASEAN regime also does not allow carriers like AirAsia from going into, say, Indonesia, either to establish a wholly-owned subsidiary or to buy over an existing local airline fully. In comparison, EU permits any E.U. national to move into another E.U. country and establish a fully-owned airline there, and fly it between any two points within the E.U. (even domestic points).
Interestingly, both MAAS and MAFLPAS provide alternatives to the traditional "substantial ownership and effective control" rule. They allow "ASEAN community carrier", which means an airline can be substantially owned and effectively controlled by ASEAN interests taken cumulatively or in the aggregate (Tan, 2009). This means that airlines can now attract capital infusions and management expertise from more sources across ASEAN. However, there is no certainty that the "community carrier" can fly into all member states in the region, because the Member States receiving an application from such a carrier must provide an approval before the carrier can operate. This is a great disincentive for any investor thinking of constituting an airline as such, unless a significant number of member states first declare their unequivocal approval for such a model. For now, investors comply with the traditional "substantial ownership and effective control of nationals" rule. Indeed, no community carrier has thus far been set up in ASEAN, and new airlines like Malindo and Thai Vietjet Air continue to employ the traditional joint venture model that requires majority ownership and effective control by local national interests. The member states should work toward a regime that allows for carriers bearing a community ownership structure to be recognized automatically, instead of at the discretion of each individual member state. The solution is to allow member states to retain the traditional national ownership and control restrictions for their own designated carriers if they wish to, without affecting other carriers' ability to be set up as community carriers.

How far have the member states implemented these agreements and the relevant commitments?
The agreements are in force among most ASEAN member states. However, key states such as Indonesia remain outside the scope of the agreements. Both MAAS and MAFLPAS are in force after having received the acceptance of the minimum number of three ASEAN member states for each agreement. At the same time, the respective Protocols' separate requirements for entry into force have been satisfied. All the Protocols are thus in force for those member states that have ratified them. As shown in Table A3, all member states have accepted MAAS Protocols 1 to 4, but Protocols 5 and 6 have not yet been accepted by Indonesia and the Philippines. In the case of MAFLPAS and its Protocols 1 and 2, Indonesia and Lao PDR are not yet state parties (Table A4).
The following tables summarize the member states' ratification status as at October 2013: Annex In the light of such developments, there are encouraging signals that Indonesia's policy on the ASEAN agreements could be evolving. Indeed, the Indonesian government is reportedly considering the acceptance of MAAS Protocols 5 and 6. When this happens, it will be a huge boost for the ASAM integration project and the entire region.
In comparison, the Philippine government's position is slightly different. The Philippines has actually embraced MAFLPAS Protocols 1 and 2 to open up access to its secondary cities in a bid to spur regional development. Yet, it has kept its capital, Manila, restricted and has not accepted MAAS Protocols 5 and 6. The government justifies its decision based on the shortage of landing and take-off slots and overall runway congestion at central Manila's Ninoy Aquino International Airport.
While the concern over congestion at Ninoy Aquino International is understandable, the attempt to link traffic rights and airport slots is problematic. Indeed, these are separate matters that should be kept distinct. In particular, the lack of airport slots should not prevent member states from ratifying the ASEAN agreements to liberalize market access rights and to signal their support for ASEAN's market integration efforts. Linking slots to access rights is also a negative precedent in that it encourages air rights negotiators to use congestion and lack of slots (which may be within the competence of other government agencies) as pretexts to delay their acceptance of regional agreements.
For its part, it is unclear why Lao PDR has not ratified MAFLPAS Protocols 1 and 2. It is likely that internal consultations are ongoing within Lao government agencies and that ratification will happen soon. It should be also noted that Cambodia has very recently in 2013 submitted instruments of ratification for MAFLPAS and Protocols 1 and 2, becoming the latest member state to accept these agreements. Notes: As an exception to the modal aggregation rule outlined above, air passenger transport subsectors are first aggregated within mode 3, i.e. air passenger domestic and air passenger international, then the resulting modal score is aggregated with mode 1 using the modal weights as shown.  IDN  KHM  LAO  MMR  MYS  PHL  SGP THA  VNM  Banking  99  100  100  100  100  0  100  100  100  Insurance auto.  80  100  49  0  100  100  100  49  100  Insurance life  80  100  49  0  100  100  100  49  100  Fixed telecom  49  100  100  100  49  40  100  100  70  Mobile telecom  65  100  100  100  49  40  100  100  70  Retailing  0  100  0  100  100  100  100  100 Banking  99  100  100  100  30  60  100  49  30  Insurance auto.  80  100  49  0  70  100  100  49  100  Insurance life  80  100  49  0  70  100  100  49  100  Fixed telecom  49  100  100  100  49  40  100  100  70  Mobile telecom  65  100  100  100  49  40  100  100  70  Retailing  0  100  0  100  100  100  100  100  100  Air transport  49  49  49  49  49  40  100  49  49  Maritime ship.  49  49  NA  0  100  40  100  49  49  Maritime aux.  49  100  NA  0  40  100  49  51  Road freight  49  100  49  0  49  40  100  49  51  Rail freight  0  100  NA  0  0  0  NA  49 49 Note: Zero means foreign ownership is not allowed; NA means not applicable due to different reasons, such as the country is landlocked or has no railway system. Empty cells mean data is missing.  Registration/License: Obtained a current and valid professional registration or licensing certificate to practice architecture in the Country of Origin issued either by the Professional Regulatory Authority (PRA) of the ASEAN Member Countries and in accordance with its policy on registration/licensing/certification of the practice of architecture or the Monitoring Committee  Registration/License: Professionals should have been assessed within their own jurisdiction as eligible for independent practice. The assessment may be undertaken by the Monitoring Committee (MC) or by the Professional Regulatory Authority (PRA) within the country of origin.  Work Experience: Acquired practical and diversified experience of not less than ten (10) years of continuous practice of architecture after graduation, of which at least five (5) years shall be after licensure/registration and at least two (2) years of which shall be in responsible charge of significant architectural works  Work Experience: Gained minimum of 7 years of experience (since graduation), of which at least two (2) years shall be in responsible charge of significant engineering works.  Cambodia Entry is allowed subject to meeting certain conditions: Educational and work experience requirements must be met. Foreign degrees and work experience recognized, the number of years of work experience is not available. There is a restriction on the employment of foreign employees, which is applicable to all firms. The maximum percentage of foreign employees in any firm is set at 10%. Exceptions may be granted. Initial stay allowed is 1 month, can be extended to 12 months.

ANNEX TABLES Annex
No restrictions except 90 percent of firm employees need to be nationals.

Thailand
Not allowed. Not allowed.

Vietnam
Must meet educational requirement, degrees from foreign countries recognized. Work experience or training not required. Professional exam in local language is required. Labor market test is required.
No restrictions except for labor market test required.

Indonesia
No sector specific regulation governing this subsectormode. There is no additional qualification requirement. Work experience or training not required.
No sector specific regulation governing this subsector-mode. There is no additional qualification requirement. Work experience or training not required.

Philippines
Foreign citizens may be allowed to take licensure exam if he/she can meet reciprocity requirement and obtained educational from universities recognized by the Government of Philippine. Foreign nationals need a special/temporary permit from the Board of Architecture and the Professional Regulatory Committee (PRC). Must be qualified to practice architecture in his/her own country. Foreign nationals shall be required to work with a Filipino counterpart. Work experience of 2 years is required. LMT is required.
Foreign licensed professionals maybe allowed to take the engineering license exams, practice, or be given a certificate of registration or be entitled to any privileges under the pertinent professional regulatory laws, provided that the country of which he/she is a citizen permits citizens of the Philippines to practice within its territorial limits under the same rules and regulations governing citizens thereof. This provision pertains to agriculture, geodetic, mechanical, metallurgical, chemical, civil, electrical, mining, naval architecture and marine, sanitary and electronic and communication engineering. LMT is required.

Myanmar
No regulation or policy exists, but in practice entry is allowed. Domestic regulations are being drafted. Foreign licensed professionals can provide services automatically without additional requirement for qualification.
No regulation or policy exists, but in practice entry is allowed. Domestic regulations are being drafted. Foreign licensed professionals can provide services automatically without additional requirement for qualification.

Lao PDR
Foreign licensed professionals are qualified automatically without additional requirement. No educational, work experience, and training requirement. Entry as a SSE is not allowed. The limit on the length of stay initially allowed is 4 years. Extension of stay is allowed.
Foreign licensed professionals (FLP) are qualified automatically without additional requirement. No educational, work experience, and training requirement. Entry as a SSE is not allowed. The limit on the length of stay initially allowed is 4 years. Extension of stay is allowed.

Malaysia
Must reside in Malaysia not less than 180 days in a calendar year. Must be qualified in the country where he/she normally practices. Need to meet labor market test. The length of stay initially allowed is 5 years. Extensions of stay are allowed. Work experience of 5 years is required, foreign experience is recognized. Entry through ICT is not allowed.
Must reside in Malaysia not less than 180 days in a calendar year. Must be qualified in the country where he/she normally practices. Need to meet labor market test. The length of stay initially allowed is 5 years. Extensions of stay are allowed. Work experience of 5 years is required, foreign experience is recognized. Entry through ICT is not allowed.

Singapore
FLPs can provide services subject to certain conditions, (regulated by Architects Act). Must need educational requirement, degrees from certain countries are recognized: Universities from Australia, England, Scotland, Wales, Ireland, Northern Ireland, Germany, Japan, France, Canada, China, Hong Kong SAR, China, New Zealand, USA. Must be qualified to practice in any foreign country. Work experience (can be in any country) requirement varies from 2 to 10 years. The length of stay initially allowed is 2 years. Extension of stay is allowed.
For professional engineering services in civil, mechanical and electrical engineering, FLPs can provide services subject to certain conditions: Must be qualified and licensed to practice in Singapore for registration as professional engineers in the above branches.