Do we need EU regulation to open up trade in Europe? Mutual recognition versus Harmonisation

Originally published in Liberaal Reveil, the magazine of the Dutch Telders Foundation

How to open up markets in the European Union? There basically are two ways to do it. Take the example of architects. How do we make sure an architect from Poland can be active in France? 

Roughly speaking, one way is for France to remove all protectionist aspects of its regulation for architects so Polish architects can operate in France, while of course still respecting all regulations concerning health, safety and consumer protection. Thereby both countries “mutually recognise” each other’s standards. Trade is about trust.

Another way is for the European Union to harmonise rules for architects, so these are the same all over the EU.  

Of course, in reality, there are many shades of grey. There are minimum- and maximum- harmonisation, and to achieve mutual recognition, sometimes it’s even needed to issue EU directives, as was actually the case for architects.[1]  For the sake of clarity, in what follows underneath, “harmonisation” will be understood as “the creation of new rules at the European level”, whereas “mutual recognition” will be understood as “merely scrapping protectionist elements of national regulations”, recognizing that in reality we often see a mix.

Hereunder I provide an overview of how opening trade in Europe became more and more about harmonisation, why that is not a good thing, and what to do about it.

  1. History:

The Treaty of Rome, signed in 1957, foresaw a transition period whereby countries only retained their veto in common decision making for a certain period of time. When the transition period came to its end and majority voting was due to be introduced, in 1965, France refused to send any representatives to Brussels, as a means of protesting this, leading to the  “Empty Chair Crisis”.[2] With the so-called “Luxembourg compromise”, it was agreed that in questions of vital interest member states retained a veto right[3]. This was a non-legally enforceable political veto, watered down[4] in the 1980ies, which could be used in areas of majority voting. It shouldn’t be confused with circumstances in which member states still retain a legal veto right. 

In 1979, a very important principle was established by the European Court of Justice (ECJ) in its Cassis de Dijon – ruling[5], declaring that under European law, if a company is allowed to make a product freely available for sale in one European Community country, then it must be allowed to do so in all member states. The product involved was in this case a French liquor, which had an alcohol percentage of 15%, while German law banned any fruit liquor with an alcohol percentage lower than 25%. As not many German liquors would violate this particular German law, the Court ruled that such a restriction was de facto protectionism through the back-door.[6] The French company won the case, but France wasn’t overly happy with the result. The Court, which left some room for exceptions[7], had hereby established more firmly the so-called principle of mutual recognition, and the French government could already imagine that many of its own regulations would be branded protectionist and illegal under European law. 

Given that returning to a context of national borders made even less sense in the 1980s than in the 1950ies, the obvious solution for this was to draft more rules at the European level, if having all these rules at the national level wasn’t possible then. As Jens-Peter Bonde, an MEP for Denmark for 30 years, puts it: “The verdict forced the member states to agree on common standards to which they would otherwise not have agreed. It paved the way for decisions by qualified majority under the so-called Internal Market, introduced by the Single European Act in 1987.”[8] 

So in order to easy rule-making at the European level, veto powers were being scrapped by this “Single European Act”, which set the objective to establish a European “single market” by 1993. A laudable aim, but in order to achieve this, the Treaty extended Qualified Majority Voting to new areas[9], something which was repeated in every European Treaty ever since. It also introduced the so-called “cooperation procedure”, empowering the European Parliament, an obsessive cheerleader of ever more regulation.  

Like many who support the creation of a single market, Margaret Thatcher supported the Single European Act, believing harmonization was needed to open borders. It has now become clear for even the proponents of this approach that there is a heavy cost to this. Parallel to the growth of the body of European harmonised rules, the approach to open up trade in Europe through mutual recognition was continued[10].  Hereunder follows an overview of the downsides of harmonisation, the benefits of mutual recognition and what can be done to boost the use of the latter.

  1. The Problems with Harmonisation

–          1. The high cost of harmonisation’s “one size fits all” – approach 

When making laws for 28 diverse economies, covering 500 million people, there is a greater risk that regulation doesn’t suit local needs and imposes a high cost as compared to when regulation is decentralized. Also, making new rules at the European level increases the risk that these contradict national rules, leading to regulatory overlap. Even if in theory European rules then trump national ones, this increases legal un-clarity and uncertainty.

We can witness this when looking at figures. Large-scale research[11] by Open Europe has concluded that the cumulative cost of EU regulation introduced between 1998 and 2008 for all 27 EU member states is €928 billion, which is 66% of the €1.4 trillion which all national and EU regulation introduced during that period has costed. By 2018, over a 20 years period, the cost of EU regulation introduced since 1998 is estimated to have been risen to more than €3.017 trillion. This is over €15,000 per household in the EU. [12]


It must be said that this investigation only focuses on the cost of regulations, derived from so-called “regulatory impact assessments”, which is an imperfect estimate. It ignores any benefits of regulations, which in theory may be greater than the costs. However, more recent Open Europe research concluded that for 24 of the 100 EU rules that are most costly to the UK, estimated costs outweigh estimated benefits[13], again according to official government estimates, which already have a tendency to present the situation slightly rosier than assumed. 

That 66% or two thirds of the impact of regulation in the EU comes from the EU level  should moreover give some pause for thought. Are two thirds of problems in Europe really of such a nature that a common solution for 500 million people is needed? Don’t think this is all the result of some conspiracy by eurocrats in the European Commission in Brussels. Member states play an important rule in EU decision making, and are often very keen to decide rules in Brussels, rather than at home. For example, the ban on traditional light bulbs[14] would have been facing much more opposition in London and Berlin than it did, because it was decided at the European level. In Brussels, there is less scrutiny by the media or national parliaments.

When national electorates found out, a lot of criticism against the measure emerged, resulting in Germans hoarding traditional light bulbs. But it was too late. For a politician, there is a great incentive to prefer policy making in Brussels over the national capitals, if one wants to have an easy time implementing policy. 

In Brussels, the European Parliament[15] is supposed to be a check on the legislative process, given that there is a lack of sufficient scrutiny by the media and national politicians. However, instead of restraining the Commission’s machine, it acts as an extra motor to encourage it. The very least one could expect from this institution is that it would refuse to provide discharge to the EU Budget at least as long as the EU’s own auditor, the Court of Auditors, refuses[16] to provide a positive statement about EU spending for 20 years in a row now[17], due to the unacceptable high level of errors in EU spending.  To refuse discharge has been the position of the Netherlands, the UK and Sweden in the last four years[18]. The Parliament on the contrary consistently calls for a bigger EU budget. The EP gained a lot more powers with the Lisbon Treaty. Unsurprisingly, it is possibly even worse in ignoring its role of restraining the regulatory machine than it is in controlling the use of EU funds. Only in 2013, the institution voted to ban chocolate cigarettes.[19]


That two thirds of the impact of rules in the EU is coming from the EU level is a much more rock-solid claim than all kinds of other figures floating around, which either ignore local regulation and thereby overestimate the power of the EU, or do not take into account indirectly applicable EU regulation.[20] These are of course all rough estimates, but very few dispute that the European regulatory machine is out of control. The European Commission itself admits this, as it has on regular occasions, in 1996, 2005 and 2013[21], launched large scale attempts to cut EU red tape. These laudable attempts haven’t been very successful, unlike some attempts at the national level. It emerged this year for example that the UK government’s cuts to red tape over the last five years were entirely undone by new EU regulations which came into force in the last two years alone[22]

There is some cause for hope, however. The second man in command at the new European Commission, former Dutch Foreign Minister Frans Timmermans, is responsible for “better regulation.” No-one should be naive about what one man can do to change the culture at the European level, but it is the most high-profile attempt at that level to do something about this problem.[23] His initial attempts to single out 80 legislative proposals for withdrawal or reworking[24] were already met by fierce opposition[25], not least by certain member states, who for example opposed his idea to ditch the proposal for an EU-ban on plastic bags. It only goes to show that the root of the problem is not just in Brussels, but also in national capitals, who sometimes like to abuse the European Union as a means for imposing regulations they couldn’t pass as easily at the national level.

–          2. It’s harder to change EU rules than is the case with national rules:

Industry lobbyists are often quite keen on harmonisation. They see the immediate benefits of countries having to adopt rules which lobbyists can easily influence at the European level. They however forget that once these rules are in place, in order to change them, one needs to go through the whole European legislative machinery again: the European Commission must issue a proposal, in most cases the European Parliament must have its say and 28 governments need to agree on changing it. The long-running discussion of changing the so-called “posted workers directive” can be an indication. This directive allows workers to work in other EU member states for a certain period as long as they comply with the social rules of their home state. While favoured by Britain and the central-and Eastern member states, social democrats in the old member states have been long been calling for changes. Only at the end of 2013 a minor deal was reached to change it. Industry lobbyists may think they’ve scored a victory when a regulation is finally drafted in a way they like, but this may backfire when as a result of technological or economic changes a regulatory change is needed. Such a regulatory change is then much more likely to happen at the national level.

–          3. There is a higher risk of “regulatory capture” with harmonisation than with mutual recognition

For interest groups of all brands, it is much easier to shape legislation when new rules are  being discussed than when the focus is only on removing protectionist elements of national regulations. The phenomenon of “regulatory capture”, whereby interest groups have a vast influence over policy making, can be witnessed both at the European and national level, but it is most harmful when it happens at the European level. If a new innovative company, like taxi ride share provider Uber, for example, wouldn’t be active on the Italian market, due to the lobbying power of the Italian taxi sector, that is bad enough, but if Uber would succeed to break into only one European country, using its success there to convince other countries after that, also Italy would ultimately have to open its borders. If Uber would be banned by the EU, this dynamic wouldn’t exist.

–          4. Regulation is better if tested in 28 different jurisdictions, rather than only in one

The chance that one gets it right after 28 attempts is higher than after only one attempt. The examples of again Uber or Airbnb, a website for people to rent out lodging, can illustrate this. Everywhere in the world, and also in Europe, national and regional regulators are struggling to adapt their old legislative framework to the advent of these newcomers. Governments will learn from each other and copy the regulatory response of the member state which is most successful in its venture to integrate new technologies. If the EU would have the power to decide the regulatory treatment of new technologies, the chance choosing incorrectly is so much higher. This is currently visible in the discussions on whether to allow genetically modified organisms (GMOs), where the EU does have competence and has imposed stringent regulations[26], strangling this potentially lucrative industry, while the rest of the world is steaming ahead. Luckily, more powers over this choice were recently handed to national authorities[27], but only after years of unfruitful discussion, also boosted by the EU’s adherence to the so-called “precautionary principle”, of which some claim[28] that we wouldn’t have enjoyed inventions as the aspirin, if this principle had been applied back in the days.

–          5. Harmonisation is questionable democratically, given how it boosts the power of technocrats

One particular problem with drafting harmonising regulations which are meant to cover 28 different countries with 500 million people is that a lot of details need to be worked out. As a result of that, a very opaque system of implementing regulations, called “comitology”[29] has been developed at the European level, providing bureaucrats with often enormous powers. A Dutch PhD thesis investigating comitology came to the conclusion that not less than half of the content of regulations was being decided through comitology, which is supposed to be only about implementing rules. So to recap: two thirds of the impact of rules is decided at the EU levels, according to official regulatory impact assessment data, and half of the content of these rules is decided by bureaucrats in obscure committees after the actual decision has been made by the European Parliament and the Council of Ministers. Welcome to Europe.

The last big revision of the EU Treaties, the Lisbon Treaty, contained some reforms of the Comitology system, but these are deemed to have increased[30] the powers of the European Commission, an institution which already enjoys vast powers and even a monopoly to propose EU legislation. Most infamously, the proposed EU ban on (non-standardised) jugs of olive oil, was passed through comitology, before it was withdrawn[31] after having been ridiculed in the media.

In EU-regulation we can see a lot of use of ‘technical standards’, which supposedly aren’t of a politically sensitive nature, but these have a worrying tendency to become politicized, which was the case with EU rules on banker bonuses.[32]

–          6. Harmonisation can easily lead to competence creep, dangerously wide interpretation of rules and more powers for unelected judges in Luxembourg

In the case of mutual recognition, not much more is needed than to identify which elements of national regulations are protectionist. This may lead to political and even judicial discussions, but nothing compared to similar discussions in the field of Harmonisation, which brings new regulations into life, which then need to be interpreted. This can lead to situations whereby member states do not apply the rules in the same way, leading to unfair competition.

When the EU banned battery cages for hens, for example, the measure was implemented by the UK and some other countries, Spain, France, Poland admitted not be ready to drop battery cages, despite having had 13 years to prepare for the change, leading to unfair competition, something the European Union is in theory not supposed to create, but to erase[33].

Another consistent feature is that the rulings of the European Court of Justice (ECJ) in Luxembourg have radically changed the meaning and scope of EU rules – in various EU policy areas, typically expanding the reach of EU rules[34], not least because being activist is in the DNA of the ECJ, which has been dubbed “the least accountable and least representative institution of all”.[35] Just like the European Parliament and the European Commission, the ECJ’s powers have been boosted by the Lisbon Treaty.[36] One way to restrain the ECJ would be to create an intergovernmental “EU Subsidarity Court”, to which the ECJ would be subject to in matters of distribution of competences. 


III.             The benefits of Mutual Recognition:

–          1. Absence of expensive regulation

All the costs related to harmonisation listed above wouldn’t exist if the path of mutual recognition of norms was chosen to open up trade in Europe. Most importantly, there is of course the high cost of many of these rules. In today’s world, with its global supply chains, the benefits of releasing European industry from crumbling regulations are perhaps even higher than assumed, as benefits would extend outside of Europe. As already mentioned, it’s true that sometimes regulations also bring about benefits, but given that official government data already indicate that this isn’t so in one in four cases, it is hard to underestimate the economic benefits which would result from a fundamental cleaning up of the European regulatory rulebook.

–          2. More regulatory competition, which isn’t necessarily less regulation or a “race to the bottom” of regulatory standards:

As explained above, mutual recognition allows for competition between states in order to create the best regulatory environment. Member states can still issue very stringent regulations, but as long as they’re not protectionist. Germany can ban cigarettes, if it wants to. It just isn’t allowed to ban “French” cigarettes. The absence of harmonisation doesn’t necessarily mean that there will be less regulation. In discussions about EU rules for capital standards for banks, a prominent example of EU harmonisation, the UK and Sweden had to push very hard in order to be allowed to impose capital standards on banks that were more stringent than prescribed in the original proposals for harmonisation[37]. If this issue would have been left to the member states in a framework of mutual recognition, the liberty of both countries to regulate its banks more fiercely than France and Germany would not have been endangered.

Separately, it should be noted that there isn’t much compelling empirical evidence for fears that regulatory competition would lead to an undesirably low level of regulation[38]. If there is one thing we cannot accuse national politicians in Europe of, it is a reluctance to issue rules and regulations. A study by Grant Thornton[39] concluded that Poland was Europe’s undisputed leader in volume of new legislation, at least when calculating in number of pages, closely followed by France.


–          3. Without harmonisation, there is a lower risk of regulatory capture

Special interest groups would have a harder time to get their way if so many regulations weren’t decided at the European level. They surely would still lobby fiercely in favour or against removing protectionist elements in national regulations if mutual recognition would be the dominant practice. In itself, there is nothing wrong with lobbying. On the contrary. It’s hard to imagine how poor law making would be without the input of the industry which will be burdened by those laws, especially when it comes to very technical product standards and regulations. Of course, industry would prefer that no burdensome harmonising measures would be taken, but given that they are, they may as well try to shape them.

  1. How to boost the method of mutual recognition again and improve harmonisation?

There are many ways to support the use of mutual recognition and to stop the EU’s regulatory machine. In order to get rid of many burdensome regulations, there is only one, very complicated way: renegotiating at the EU level, apart from hoping that technological development will make some of these regulations redundant. Perhaps making it easier to abolish old EU rules than to make new ones would be an option, but this may be legally tricky. There are many ways to prevent expensive EU rules entering into force in the future, however. For some of them, Treaty change is needed.

–          1. Restoration of vetoes for member states in the EU Council of Ministers

If 28 member states can exert a veto, it will become harder to agree on new EU legislation, thereby forcing mutual recognition as an alternative approach to the forefront. This would increase cross-European support for EU-rules and would be an extra safeguard against bad rules. There would be a flipside, however: it would also become harder to abolish EU rules.   

–           2. Restricting what the Commission can do

One way would be for the member states to set an exhaustive and tightly‐defined agenda for the EU Commission, which has a monopoly of initiative to prepare EU rules. Anything not included in this, would be off-limits and a competence of member states. The ongoing attempt to harmonise taxes through harmonizing the tax base[40], for example, is a proposal which is constantly being recycled by the Commission but of which member states have always made clear they don’t think it’s any of the Commission’s business.

–          3. Red cards for national parliaments:

Equipping national parliaments with a veto, a ‘red card’, would give the current ‘yellow card’ system whereby national parliaments can force the Commission to rethink proposals binding teeth. The yellow card system has only used a few times. It was introduced by the Lisbon Treaty, stating that if one third or more national parliaments object to an EU proposal on subsidiarity grounds (within an eight week window), then the Commission has to reconsider the proposal. In the case of the yellow card against the so-called “European Public Prosecutor’s Office (EPPO)”[41], the Commission basically ignored[42] the opposition from eleven national parliaments. So did the European Parliament, which is even less surprising. Respect for subsidiarity is unlikely to materialize in practice as long as it remains the responsibility of EU institutions to enforce it.

–          4. A Subsidiarity Court

The creation of an intergovernmental “EU Subsidarity Court”[43], to which the European Court of Justice (ECJ) would be subject in matters of distribution of competences, could help to restrain the judicial activism of the ECJ[44] which leads to EU ‘competence creep’.[45]

Unlike now, it would become possible to appeal ECJ rulings, but only on the grounds of subsidiarity issues. Important is that this Court would be directly composed by either the presidents of national Constitutional Courts or representatives appointed by governments. In the latter case, not much would change, but if for example the President of the German Constitutional Court would have a seat in there, there would be an institutional link with the national judicial order.

Of course, this could as well fail to serve as a safeguard, but at the moment the ECJ is completely unrestrained, apart from the possibility for member states to nominate a judge for a period of six years, although even this is subject to formal approval by the other member states.

–          5. A Subsidiarity Commissioner

To a large extent, the idea of a “Subsidiarity Commissioner  has now been implemented with the arrival of Frans Timmermans as second in command in the Juncker-Commission, responsible for “Better Regulation”. It is his job to scrutinize whether legislative proposals respect the subsidiarity and proportionality principle and to become unpopular with his colleagues, as for example Energy Commissioner Canete, who saw 62 of his 64 proposals axed[46] by Timmermans last year. If Timmermans is successful, he should ultimately become popular with the public, of which only 19%  is happy with the EU at the moment.[47]

–          6. Improve harmonisation

In certain cases, harmonisation is needed and even in the cases where it isn’t needed, there are ways to minimize the burden it imposes. Suggestions made by Open Europe include “one in, one out” arrangements, sunset clauses, common commencement dates for regulation, expiry dates, improved impact assessments, the creation of an independent impact assessments board and simplification of procedures to scrap regulation.[48]


In short, harmonisation or the creation of new rules at the European level should be limited to exceptional cases. The focus should instead be on mutual recognition of national rules, by scrapping protectionist elements from these rules.

The EU’s attempts to open up borders between member states for the delivery of services illustrates this point. The original version of the services directive contained the idea to install the so-called “country of origin principle”, meaning that service provides who would comply with the regulations of their home country would be welcome to operate in other member states. This was effectively an application of mutual recognition. As a result of opposition by Belgium and France, fearing the arrival of the so-called “Polish plumber”, it was watered down in favour of a system which allowed loopholes[49] for member states not to open up their services markets, which they have been gratefully exploiting. As a result, implementation has been patchy at best, with many barriers still in place. A new attempt to open up services in the EU should therefore include the “country of origin principle”. Open Europe calculated that if only the twelve countries who in 2012 declared to be willing to open up their borders for services would do so amongst each other, through “reinforced cooperation”, about half of the economic benefits would materialise as compared to when all EU member states would do that. The “Polish plumber” would still not be welcome in Germany or France, who wouldn’t have to do something they don’t want, while he or she would be welcome in Sweden, Italy and the UK.[50]

At the moment, a very similar discussion is raging with regards to the Transatlantic Trade and Investment Partnership (TTIP), a trade deal under negotiaton between the European Union and the United States. A lot of TTIP-critics complain that it will impose deregulation through the back-door. It’s too early to tell if this will be the case, but TTIP-negotiators know what they should do: not imposing regulation or deregulation, but removing barriers to trade. TTIP shouldn’t be about keeping China out of the trade system and imposing common Western standards on the rest of the world. It should be about helping businesses to expand their transatlantic activity, so wealth is created which enables to boost trade between China and the West even further. When the principle of mutual recognition has prevailed in TTIP, by limiting it abolishing trade barriers, it’s high time to apply it again at the European level.

Pieter Cleppe represents independent think tank Open Europe in Brussels


[2] and



[5] and

[6] The ECJ judged that such measures fall under European Treaty articles like current art. 34 TFEU which prohibit “quantitative restrictions and measures of equivalent effect”

[7] The ECJ ruled that exceptions to this principle as described by current art. 36 TFEU can be made only in order to serve the public interest (eg, the protection of health, the environment or consumers.


[9] Currently specified in the articles from article 114 TFEU on

[10] To make the mutual recognition principle fully operational, the European Parliament and the Council adopted Regulation (EC) No 764/2008, which concentrates on the burden of proof by setting out the procedural requirements for denying mutual recognition

[11] and and

[12] Especially as the cost of EU rules was accelerating by 50% between 2005 and 2008. For the record: EU GDP amounts to around €13trillion per year.

[13] These regulations include the Temporary Agency Workers Directive and the Energy Performance of Buildings Directive






[19] MEPs supported that “Imitation tobacco products which appeal to minors and consequently form a potential gateway to using tobacco products shall be prohibited.” The Council later forced the EP to abandon this position  







[26] Who are estimated to add between 10 and 20 million pounds to the cost of developing a GM trait in a crop, which is deemed prohibitive for the public sector and for small and medium sized businesses. The UK House of Commons recently slammed the EU treatment



[29]“Comitology” refers to a system whereby implementing powers are sometimes attributed to the Commission aimed at implementing legislation uniformly in the Member States. In exercising its implementing powers, the Commission is assisted by representatives of the Member States, sitting in committees. and  





[34] For example, the SiMAP and Jaeger rulings on working time spent resident on call and compensatory rest have made the EU’s Working Time Directive hugely more burdensome for the UK’s National Health Service. Similarly, the ECJ ruled last year to scrap the insurance industry’s derogation from the EU’s Gender Directive, which had the effect of preventing the insurance industry from offering different premiums based on gender, even if it can be statistically proven that men and women present different degrees of risk. P.16:




[38] P 22

[39] Of course one would need to look at the actual impact of the rules to get a clearer picture, but it may be an indication of some sort.

[40] The so-called “Common Consolidated Corporate Tax Base“ and




[44] If its budget can be any guide: this has tripled as compared to 10 years ago



[47] German voters have the least amount of trust in the European Commission and European Parliament of 13 national and EU institutions tested. They also want national parliaments to be able to block EU laws.


[49] Allowed to impose requirements “where they are justified for reasons of public policy, public security, public health or the protection of the environment”.