Search Menu


Issues list



Cover image



The ‘Public Order’ Parachute in Combating Racism and Xenophobia

The European Commission has initiated infringement proceedings against virtually half of the Member States, accusing many of, inter alia, failing to take the necessary measures to ensure that racist and xenophobic hate crimes are effectively criminalised. The article looks at the right of a Member State to limit prosecution for incitement of violence or hatred to acts that are carried out in a manner likely to disturb the public order. The authors argue that application of the ‘risk to public order’ criterion if interpreted appropriately, will in most cases reduce the threat of ‘taking it too far’. They argue also that there is a risk of confusion between ‘public order’ and ‘public nuisance’, due to the national criminal courts being more familiar with the latter. This could lead to unreasonably loose application of criminal punishments and pose a risk breaching the nullum crimen nulla poena sine lege certa principle. Perhaps because of having suffered censorship and absence of fundamental rights under the Soviet Union, Estonian society voices strong concerns about criminalisation of hate speech. In the authors’ view, these concerns may be reduced by narrow interpretation of the ‘risk to public order’ notion.


EU law; internal market; xenophobia; racism; public order; freedom of expression; freedom of speech; framework decision; fundamental rights; criminal law

1. Introduction

The fight against racism and xenophobia and its relationship with other freedoms has been a hot topic both in academia and at family dinners. Newspapers report that an Austrian man, Helmut Griese, at the age of 63, was fined approx. 700 euros by a local judge for yodelling while mowing the grass in his back yard. *1 The accusation was of this being ridiculing of the religious beliefs of the Muslim family living next door. In 2021 a Belgian quartet was convicted of hate speech after wearing a ‘Stop Islamisation’ banner at the market of Mechelen. Basing the decision on Belgian criminal law, the ‘Correctional’ Court of Mechelen convicted four members of the right-wing organisation Voorpost to six months in prison for inciting hate and violence during a protest last year. The people in question had banners with the following text (freely translated): ‘Is this the future of Flanders? No, thank you!’ and ‘Stop Islamisation’. The banners also showed women wearing a burka or niqab. Combining these two slogans and the women displayed on the banners was a bridge too far for the Belgian court. *2 The warning label ‘[C]ontains language and attitudes of the time that may offend some’ at the beginning the beloved BBC comedy 'Allo 'Allo! caused bewilderment among nostalgic TV viewers. *3

At EU level, these discussions show links to the Framework Decision (FD) on Combating Racism and Xenophobia (2008/913/JHA) *4 , to synchronise rules on crimes with a racist or xenophobic background. The FD stands on the founding principle that nothing less than a criminal offence with a maximum sentence of at least 1–3-year imprisonment is sufficient to address serious racist or xenophobic acts. With the initial proposal dated 26 March 2001, it certainly took a long time to negotiate. *5 The FD refers to race, colour, religion, descent, or national or ethnic origin as characterising types of groups public incitement of violence against which must be penalised. It calls for the criminalisation of incitement of violence or hatred; the public dissemination or distribution of writings, images, or other materials with racist or xenophobic content; and the public endorsement, denial, or gross downplaying of genocide or crimes against humanity.

In October 2020, the European Commission decided to send letters of formal notice to Estonia and Romania. According to their assessment, the national laws do not fully and accurately transpose the EU rules on combating certain forms of expression of racism and xenophobia through criminal law. *6 By the end of 2021, similar proceedings had been started against 13 Member States, or 48.15% of them all, and the list may not yet be complete. Among those the Commission has decided to open infringement proceedings against are Germany, Hungary, and Luxembourg. One of the general accusations presented by the Commission is a failure by the Member State to take the necessary measures to ensure that racist and xenophobic hate crimes are effectively criminalised. *7

This article focuses on the possibility of the Member States limiting the criminalisation to only conduct that is either carried out in a manner likely to disturb public order or threatening, abusive, or insulting. *8 In our opinion, including this exception in national criminal law would in most cases reduce the threat of ‘taking it too far’ in relation to the risk of people getting punished in cases that do not by nature deserve intervention by criminal law and that of excessively limiting the freedom of speech and expression. It would also reduce the risks related to some fundamental principles of criminal law.

The ‘Estonian angle’ is presented to the reader to provide context related to ‘public order’ from the standpoint of international law being confused with ‘public nuisance’ within the context of national law. Namely, Estonian penal law perceives ‘public order’ in the context of a public nuisance, thus expressing a concept vastly different from the understanding of the term in international law. The criminal law’s sections about ‘hate speech’ will be applied not by professors of international or EU law but by national criminal-court judges. This generates a severe risk of spill-over from the world of misdemeanours. We argue that said risk can be mitigated if the courts restrict themselves to the narrower interpretation of ‘public order’ in cases of hate crimes.

2. The Estonian ‘hate speech rules’ and their application

Hate speech is something that Estonia has condemned at Constitutional level. *9 Art. 12(1) of the Constitution states that

[e]veryone is equal before the law. No one shall be discriminated against based on nationality, race, colour, sex, language, origin, religion, political or other beliefs, property or social status, or on other grounds.

According to Art. 12(2) of the Constitution,

[t]he incitement of national, racial, religious or political hatred, violence or discrimination shall be prohibited and punishable by law. The incitement of hatred, violence or discrimination between social strata shall also be prohibited and punishable by law.

Of course, the Constitution itself does not turn any act into a punishable one. In the logical structure of laws, criminal offences, among them a number of misdemeanours, have their home in the Penal Code. ‘Hate speech’ as a term in its own right is not expressly used in the Estonian Penal Code but is employed in informal communication (in the vernacular sense). Instead, similarly to the FD, Section 151 speaks of incitement of hatred. The current Subsection 151(1) of the Estonian Penal Code *10 (in Section 151, titled ‘Incitement of hatred’) reads:

Activities which publicly incite to hatred, violence or discrimination based on nationality, race, colour, sex, language, origin, religion, sexual orientation, political opinion, or financial or social status if this results in danger to the life, health or property of a person are punishable by a fine of up to three hundred fine units or by detention.

The so-called qualified offence is provided for in the second subsection, which refers to

(2) [t]he same act if:

1) it causes the death of a person or results in damage to health or other serious consequences; or

2) committed by a person who has previously been punished by such act;

is punishable by a pecuniary punishment or up to three years’ imprisonment.

Similarly to many other Member States’, Estonia’s law distinguishes between misdemeanours and criminal offences. Accordingly, the first subsection of Section 151 characterises a misdemeanour for reason of the foreseen punishment of a fine or detention up to 30 days. The second subsection foresees a pecuniary punishment or up to three years in prison, hence representing a criminal offence. Accordingly, incitement of hatred is a criminal offence in Estonia only if it either causes the death of a person or damage to health or other serious consequences or is committed by a person who has previously been convicted for incitement of hatred (per Subsection 151 (2)).

The offence of hate speech has gone through several changes since Estonia regained independence in 1991. Until 2004, publicly inciting hatred or violence was a crime punishable by pecuniary punishment or up to three years’ imprisonment. On 30 June 2004 it was downgraded to a misdemeanour punishable by a fine of up to three hundred ‘fine units’ or by detention, and on 16 July 2006 it was downgraded once again, with incitement of hatred or violence becoming punishable as a misdemeanour only on the condition that it has caused a danger to a person’s life, health, or property. Since then, Section 151 has remained, in essence, the same. However, this does not mean that it has remained uncontested. Amending Section 151 has been discussed many times (2005, 2012–2013, 2020, 2022), in most cases with the aim of converting it back to a criminal offence and replacing the criterion of danger to a person’s life, health, or property with a less restrictive characteristic such as violation of public peace, a systematic nature to the act, or a threat to public order. *11 A recent proposal to amend Section 151, made in 2022, covered moving the result – danger to life, health, or property – from Subsection 151(1) to Subsection 151(2). *12 In the explanatory memorandum, the following was stated:

Incitement to hatred that does not lead to a clearly identifiable and causal consequence still jeopardizes the legal interests protected by rendering this an offence and has potential to contribute to dangerous hostile attitudes in the society. That is why incitement to hatred without a consequence too should be a misdemeanour. *13

Statistics show that hate speech offences are not widespread in Estonia. In this country with approximately 1.3 million inhabitants, 25,800 crimes registered per annum, and a crime index of 23.38 in 2021 *14 , the section penalising hate speech is rarely applied. Ten instances of criminal offence in this regard were registered between 2003 and 2020, the most recent in 2015. During this period, six people were convicted, five of them for a crime committed before 30 June 2004. The last conviction was handed down in 2005. In all, 26 misdemeanours were registered between 2004 and 2020, with 16 proceedings being terminated and 13 convictions made. *15 The only judgement the Supreme Court has made on hate speech was one in 2006 regarding the wording of Section 151 of the Penal Code, which entered into force in 2004. *16

It could be that Estonian society is very tolerant and serious cases of hate speech are rare. It could equally well be that the cases that fall under the hate speech section of the Penal Code are defined too narrowly (discussed shortly below). No empirical research is available to help. Since 2016, the Ministry of Justice has collected data on other crimes committed with the motive of hatred, finding that 38 such crimes were registered in 2016–2020, motivated by hatred among other things. Most of these offences were criminal offences qualified under Section 263 of the Penal Code, on aggravated breach of public order, under Section 121 of the Penal Code, on physical abuse, and Section 120 of the Penal Code, on threat. *17 In addition, the 2018 Victims’ Survey shows that almost 2.6% of the 1002 respondents had been victims of crime because of their nationality, race, colour, religion, disability, or sexual orientation. *18 These data do not reveal the spread of hate speech in the society. Data collection is further complicated by the fact that the Penal Code does not include racist and xenophobic motivation as an aggravating circumstance, although it can be taken into consideration by the courts in the determination of the penalties.

3. The ‘Estonian problem’

Being inhabitants of a country forcefully included in the Soviet Union for fifty years, Estonians vividly remember the absence of freedom of expression and the strict censorship involved. Perhaps a fear of taking the first steps back in that direction drives the resistance to criminalisation of hate speech in general terms. In other words, there is strong opposition in Estonian society to making the hate speech rules stricter. The debate in and beyond the parliament has been fierce. Supporters of criminalisation argue that a need to criminalise hate speech arises not only from EU law but also from the Constitution. However, this argument can be contested, as the Constitution states only that such activities should be forbidden and punishable, without specifying the field of law that should impose such a ban with relevant consequences. In any case, even the supporters of criminalisation tend to feel that the pressure from the EU to expedite the process is too strong to allow constructive discussions and well‑considered decisions.

The infringement proceedings initiated by the European Commission against Estonia added fuel to the fire surrounding what is seen as tightening the screws around freedom of expression. One of the root causes of this confrontation is the requirement in Art. 3(2) of the FD for Member States to provide criminal penalties as a minimum and indicating thereby that lesser penalties are insufficient. As we discussed above, the FD requires the Member States to impose criminal penalties of a maximum of at least 1–3 years of imprisonment for hate speech offences, a criterion that Subsection 151(1) of the Penal Code does not meet. In addition, and again as discussed above, when it comes to hate speech as a crime, Subsection 151(2) of the Penal Code foresees preconditions of ‘death’, ‘damage to health’, or ‘other serious consequences’– conditions not present in or permitted by the FD – as prerequisite for criminal liability.

So it is rather likely that Estonian law does not meet the minimum standards foreseen by the FD, yet the Estonian Government has consistently rejected the Commission’s arguments in the infringement proceedings. *19 The issue itself became political long before the official infringement proceedings started. In 2015, Minister of Justice Urmas Reinsalu repeatedly expressed the view that it is not wise to criminalise hate speech. *20 He referred to people who think that criminal law is a way to rearrange people’s attitudes when stating:

For such people, I’d recommend more readings of «The Gulag Archipelago» where Aleksandr Solzhenitsyn is writing about a teacher at military academy jailed for laughing while reading the party paper. A mental murder. *21

The minister continued by stating that

[i]nciting hatred, hostility and discrimination is provided by penal power as it is, but it needs to contain [the] objective necessary elements of offence – real danger must have occurred. If we will undertake to rapidly enforce fresh regulation and take the labelling attitude, what we may achieve is very forceful contradictory emotions which may pose danger in society.

The infringement proceedings were perceived as a slap in the face and added fuel to the fire. The Reform Party, in opposition at the time, filed a proposal for a bill to criminalise hate speech. Its leader, Kaja Kallas, who was to become prime minister in 2021, explained:

The government coalition has made inciting hatred against various minorities increasingly commonplace. However, this should not be the case in a state based on the rule of law[,] and incitement of hatred and public calls for violence should be punishable in criminal procedure, even if victims do not immediately follow. *22

The leader of coalition party Isamaa, Helir-Valdor Seeder, criticised the hate speech bill of the Reform Party, saying that it would curtail free speech and free media:

Isamaa believes that a free society does not need to solve playground issues via a militia and a criminal code. A democratic society honours the rights of the people and of the media, to express themselves freely.

The leader of Isamaa further explained:

The attempted further criminalization of hate speech represents a creeping introduction of censorship. The rights of people and the media to free speech may not be put under ideological pressures. Opinions and expression cannot be influenced solely by the law. Reform's proposals would be a welcome means of intimidation, one which would provide the chance to silence many fundamental debates. The state cannot dictate to people what and how they can think or speak. *23

Estonia’s Prosecutor General Andres Parmas also intervened and emphasised that freedom of speech is a very important value for society. He also expressed fears, stressing the limited resources of the Prosecutor’s Office:

[For me a]s the head of the prosecutor’s office, it cannot be ignored that such a change in the law means an additional workload for both the prosecutor's office and the police and the court. This resource [burden] must be borne by the prosecutor's office at the expense of other crimes. *24

The bill submitted by the opposition Reform Party with a proposal to criminalise hate speech was voted down by members of the Riigikogu in December 2020, 50 against 39. *25

In April 2022 after the proposal to amend Section 151 was dropped from the bill banning aggression symbols, Minister of Justice Maris Lauri lamented:

Are we really doing this in such a way that use of symbols is punishable but if people are attacked with words, insulted, or victims of agitation to be beaten, there is nothing wrong? *26

In summary, infringement proceedings are ongoing, Estonia has not accepted the allegation of being in breach of its EU obligations, and a large proportion of its politicians and general population strongly disagree on whether hate speech should be criminalised. *27

4. The clarity (certa) issue

Although there are conflicting moods in the society, it is a fact that international pressure to criminalise hate speech exists. Just recently, in late 2021, the Commission launched its initiative to extend the list of Euro-crimes to hate speech and hate crime via the Treaty on the Functioning of the European Union. *28 Therefore, the need for and possibility of criminalising hate speech must be discussed with open cards in consideration of all relevant circumstances.

From the perspective of criminal law, criminalising ‘hate speech’ can only happen when the phenomenon is defined in a manner that is in line with the principle of legal certainty, also known as the principle nullum crimen nulla poena sine lege certa. According to this principle, the characteristics of the punishable act must be comprehensible to the addressees of the law and the person applying the law at least to the extent that their content can be understood via interpretation. *29 In short, the wording of the law should enable people to choose the right behaviour and the courts to avoid arbitrariness. As one of the fundamental principles of criminal law, the certa principle requires that substantial vagueness in defining crimes be avoided at all cost. In sum, whether the criminal norm fulfils the legal-certainty principle substantially depends on the composition and wording of the norm.

The ‘Estonian problem’ is that the law contains preconditions for a hate crime that the FD does not recognise. If these presumptions were removed, possible contradiction with the principle of certainty would arise. Let us be more precise: currently, in order for hate speech to be punished as a misdemeanour, the law requires the presence of specific results – danger or real damage. In light of the FD, these preconditions should be removed. Here, the European Commission against Racism and Intolerance in 2015 recommended that the Estonian authorities introduce without delay in ‘parliamentary proceedings a draft amendment to section 151 of the Criminal Code, removing the restriction whereby an offence cannot be deemed to have taken place unless it is proven that it entails a risk to the health, life or property of the victim’. *30

Partially in line with the goal for the FD to limit itself only to particularly serious forms of racism and xenophobia by means of criminal law, Estonian law already criminalises hate speech if it causes the death of a person or damage to health or other serious consequences or is committed by a person who has previously been convicted for incitement of hatred.

However, criminalisation of incitement of hatred without any reference to a (potentially) harmful consequence *31 would widen the space of potentially punishable activities tremendously and lay a burden on the judges handling criminal cases to determine what incitement to hatred, violence, or discrimination covers. Everything here points to a severe risk of breaching the principle of legal certainty.

This would create ambiguity in the society as to which cases may be prosecuted and which not and have serious potential to bring with it arbitrary decisions. Common sense foresees a flow of complaints to the police from people about what they read, see, or hear in their day-to-day life. It would create both additional fear of expressing one's opinion and additional opportunities to silence opponents whose views are not welcome. Until 2006, Section 151 did not require a specific consequence, and questions were raised in relation to its wording. What activities does incitement to hatred, violence, or discrimination cover? Does the incitement have to be addressed to a specific group? Must the group or persons to whom the call is addressed also be specified? The answers to these questions, now more than 15 years old, are provided in only one Supreme Court judgement and were by no means unanimous. *32

Accordingly we see the risks involved in criminalisation of hate speech in a generalist manner as outweighing the potential added benefits. It may lead to stronger division of the society instead of greater understanding and harmony. Tensions in society, which already exist without the criminalisation of hate speech, could be exacerbated. We also see that such risks could be significantly alleviated with the introduction of a threat to public order as a precondition for the crime. Therefore, we feel that the FD contains the filters necessary to limit its impact to cases with a ‘real danger’ present, which will be discussed below.

5. The risk of misunderstanding what ‘public order’ is

While making a claim that the FD contains the filter necessary for conforming to the principle of certainty – public order – we point to the vast difference between what is understood as public order in national penal law and the concept of public order on EU level. Why would such a point be relevant? Because judges dealing with criminal law may be misled by the term ‘public order’, which is very familiar to them but in a wholly different context.

In essence, the term ‘public order’ in national criminal law refers to general requirements for behaviour in public places – in other words, to creating a ‘public nuisance’. The case law of the Criminal Chamber of the Supreme Court explains ‘public order’ in the context of misdemeanours and petty crimes as ‘relations between persons enshrined in customs, good morals, norms or rules in society, which ensure for everyone public confidence and the opportunity to exercise one’s rights, freedoms and obligations’. *33 The Estonian Law Enforcement Act, which regulates the protection of public order, resonates with the case law of the Criminal Chamber by defining public order as ‘a state of society in which the adherence to legal provisions and the protection of legal rights and persons’ subjective rights are guaranteed’. *34 Here, the link between the protection of public order and law of misdemeanours is strong, as the latter deals with those acts against public order that have gone ‘too far’.

It is obvious that the above-described concept of public order is rather vague, as it seems to cover law and order as such in the society. Such vagueness stems from the nature of law enforcement and the need for flexibility of administrative law, the emphasis of which lies in prevention. Offences against public order are also provided for by the Penal Code. As previously stated, these offences also reflect the general nature of law enforcement; i.e., these are that infringement of the law and order that the legislators have considered serious enough to call for different degrees of punishment. In general, these are rather small-scale sorts of offences, such as breach and aggravated breach of public order (a misdemeanour and criminal offence, respectively), violation of the requirements for holding public meetings (a misdemeanour), an unauthorised public meeting (a criminal offence), and trespassing (a misdemeanour or criminal offence). As such, they are less serious and deemed less reprehensible than the serious offences against public order and security in the terms in which we understand it in EU law. In comparison, for instance, to public security as a universal legal interest, the local understanding of public order is related more to the values of a particular society. *35

Accordingly, on national level internally the term ‘public order’ is related primarily to minor offences that disturb other people. Rumbling in public, insulting passers-by, playing music too loudly and disturbing the neighbours, fighting in public, or (in extreme cases) just being drunk in public – these are breaches of public order that are dealt with daily as misdemeanours by judges of district courts. These very same judges decide criminal cases. Therefore, if a new crime whose precondition is a threat to the public order is introduced to them without further explanation and clarification, they intuitively, on the basis of their professional experience, consider this crime to belong to the category ‘public nuisance’. Also, there is no case law to point them in a different direction.

6. Public order in EU law

The Commission did not include the term ‘threat to public order’ in the original proposal for the FD. Accordingly, the travaux préparatoires do not contain much to help us understand the origin of the thought and meaning of the exception. Comparing the various language versions of the FD seems to indicate that the use of the term is consistent. The English version refers to ‘a manner likely to disturb public order’. In French the wording is ‘exercé d’une manière qui risque de troubler l’ordre public’, in German ‘die in einer Weise begangen werden, die geeignet ist, die öffentliche Ordnung zu stören’, and the Spanish version refers to ‘del orden público’.

The fact that both the French and the English version, as well as other language versions, refer to ‘public order’ in a consistent manner leads us to believe that the FD does not invite us to discuss the broader concept of public policy and what the differences in their scope are, if any, in different legal cultures. *36 There is a certain difference in the scope of public order and l’ordre public, which is often translated as public policy. For example, in the context of free movement of persons, the European Court of Justice (ECJ) refers to public policy and l’ordre public interchangeably. *37 The term ‘public order’ is not synonymous with ‘public policy’. *38

A.G. Sharpston makes a convincing argument identifying public order as something narrower and more concretely limited than public policy in the broader concept. In addition to public order, public policy encompasses acts that are considered to be against the policy of the law. *39 Public order is narrower in nature and ‘broadly covers crimes or acts that interfere with the operations of society’. *40

In the case of the FD, both languages, English and French, consistently refer to ‘public order’: ‘It is settled case-law that the meaning and scope of terms for which EU law provides no definition must be determined by considering their usual meaning in everyday language, while also taking into account the context in which they occur and the purposes of the rules of which they are part.’ *41 Noted, a recital does not in itself constitute a legal rule and thus has no binding legal force of its own. However, the preamble of an EU law instrument may explain the measure’s content. *42 The preamble of the FD does exactly this, emphasising the limited scope of the harmonisation intended by stating (emphasis added):

This Framework Decision is limited to combating particularly serious forms of racism and xenophobia by means of criminal law. Since the Member States’ cultural and legal traditions are, to some extent, different, particularly in this field, full harmonisation of criminal laws is currently not possible. *43

An indication of targeting harmonisation of ‘particularly serious’ situations allows for an interpretation wherein ‘threat to public order’ serves the purpose of eliminating the less serious forms of activities from the scope of the rules introduced. This leads us to a narrower context. *44

Most cases of the ECJ address public order in the context of the internal market and free movement, a context wherein ‘the protection of national security and public order also contributes to the protection of the rights and freedoms of others’ with particular reference to everyone's ‘right not only to liberty but also to security of person’. *45 This in turn should obviously exclude a liberal interpretation of what does and does not constitute a threat to public order. Showing up drunk in a public space may disturb public order for the purposes of national misdemeanour law, but it does not include a threat to the ‘security of person’ within the meaning of EU law. When labelling someone as a threat to public security, we are discussing the existence of a real threat such a person represents to public security. *46

Hence, in the Oteiza Olazabal case, restrictions to a person living close to the Spanish border were considered justified with a reference to potential terrorist ties and the risk of him living in the proximity of the ETA. *47 Evidently, the risk of realising terrorist ties is something that constitutes a risk of a highly serious nature.

If the ‘public order’ referred to by the FD can be interpreted with a parallel to the cases dealing with free movement of persons, the ECJ has held that a mere disturbance of the social order is clearly insufficient. *48 Public order, as employed in the Bouchereau judgement, presupposes the existence of ‘a genuine and sufficiently serious threat to the requirements of public order affecting one of the fundamental interests of society’. *49 If interpreted in keeping with these examples, the existence of a risk to public order as a precondition for a crime significantly changes the picture.

7. Conclusions

It is fair to say that an attempt to internationally define hate speech crimes involves an element of clash of cultures. Some societies readily criminalise a wide range of disturbing statements since they feel that in history similar statements have caused catastrophes in their country or that the statements cause unfair suffering to people in connection with dramatic events from history. Others avoid labelling such statements as hate speech as far as possible, and criminalising these statements, as their history has revealed to them the fragility of freedom of speech and expression.

Estonia belongs to the latter set on account of the fact that for almost 50 years it was involuntarily incorporated into the Soviet Union. In consequence, to Estonian society any attempt to restrict freedom of expression in a generic manner sounds like an attempt to reduce plurality of opinions in the society. Dissenting opinions may not be pleasant to hear, but they fuel the development of society. Estonians know that very well because without such opinions having flourished there might not be an independent Estonia today. Therefore, the idea of criminalising hate speech has not been warmly and unanimously welcomed. External pressure to do so has not helped but, to the contrary, increased conflicts in the society.

The need to fight serious forms of xenophobia and racism is obvious, and criminal penalties may be necessary. Accordingly, the society needs assurances and defensive mechanisms if it is to be sure that only particularly serious forms of racism and xenophobia are combated by means of criminal law. This expectation is in line with the raison d’être of the FD.

Were hate speech crime defined merely as ‘incitement of hatred, violence or discrimination’, it would most likely fall short of the principle of legal certainty. Very intense questioning about the meaning of incitement would arise. We do see a solution, whereby the criminalisation of hate speech is combined with the strict precondition of applying only to cases of actions carried out in a manner likely to disturb the public order. We reiterate that the law could define the crime of hate speech as incitement of hatred, violence, or discrimination carried out in a manner likely to disturb the public order. The minimum requirements of the FD would be fulfilled, and the rest of the articles of national law, foreseeing lesser offences as, for example, misdemeanours, would no longer contradict the FD. Instead they could be seen as gold-plating in the desired general direction.

This would resolve the bigger risk, leaving us still to deal with the national-law puzzle of how to distinguish the ‘public order’ in the FD and EU law from the very same words contained in the very same penal code referring, for example, to activities similar to being drunk and disorderly. There remains the puzzle of how to help the same judges who decide on the minor offences treat the same terms very differently when they decide on criminal cases of hate speech. Once again, risks of legal certainty are on the table.

If the judges opt for the liberal, wide interpretation equating public order to public nuisance, vast numbers of cases of hate speech could arise overnight. The public-nuisance concept would give hate speech the character of a petty offence, which would even further widen the circle of punishable acts of incitement of hatred. Any public outrage, outburst of resentment, or insult would immediately fall within the sphere of hate speech and thus of criminal punishment as all these activities go against law and order. Expressing thoughts that are unpleasant or inconvenient to other people or to the public – for example, reading insulting poems naked on a town square – could become punishable. This would go strongly against the principle of ultima ratio, whereby criminal sanctions should be the last resort in dealing with unwanted behaviour.

This would be detrimental to the society for several other reasons. Potentially, arbitrary decisions could follow and confusion be created among citizens about what activities are in fact prohibited, leading them to fear expressing themselves. It has potential to create a system in which penalties are based not on people's actions but on their attitudes. A dramatic increase in hate speech cases would demand vast resources. The resources needed to deal with hate speech would come at the expense of other cases. At the same time, lesser forms of punishments for smaller matters could easily resolve this negative impact. Thirdly, conflicts in the society would increase further as many would feel unfair pressure on freedom of expression.

In conclusion, if the crime of hate speech were to be limited in terms of a threat to public order, this has to be done in a manner ensuring that national courts understand that it is not a case of public nuisance but of real danger. The latter characterises the EU’s approach to public order. How the legislators technically approach this is another matter. If the EU’s approach to public order is applied, the situation with legal certainty is significantly improved and the solution is much more in line with the fundamental principles of criminal law. Accordingly, it would be perfectly acceptable to apply the FD’s public-order condition since freedom of expression would remain protected.

The old man could continue yodelling, and people would remain free to express their opinions and fears. And risks to public order still could be adequately addressed.