Which courts or tribunals must refer? (art 267(3)) – obligation!

      • Rationale for imposing this duty is to prevent body of national case law that is not in accordance with EU law from being established!
      • 2 interpretations:
        1. The abstract theory: only bodies whose decisions are never subject to review
        • ➔ MSs would want this – don’t want to be subject to too many mandatory prelim refs
        1. The concrete theory: depends on whether the body’s decision is subject to appeal in the type of case in question.
      • ECJ’s jurisprudence seems to support the concrete theory.
        • Costa v Manchester taxi , 1964: although magistrate’s decisions were capable of being appealed in some instances, there was no right of appeal in the particular case – allowed prelim ref!
        • Lyckeskog case, 2000: just cos appellate court decision was subject to prior declaration of admissibility before it could be appealed to the Sup Ct did NOT mean that the appellate court’s decisions came within art 267(3) – procedural impediment to appeal did NOT make it a final court. It is Sup Ct is the final court under obligation.
        • ➔ hence, question is whether there is any possible legal remedy, regardless of practicability/subjection to restrictions.
        • What if lower national court makes prelim ref, which is reversed on appeal by higher court?
          • Cartesio case, 2008: ECJ supported ability of lower courts to refer to ECJ, even in face of opposition from higher court. It is up to the referring court alone (and not the appeal court) to assess relevance and necessity of prelim ref. For clarity and legal certainty, ECJ will abide by decision to make prelim ref, which will have full effect as long as it hasn’t been revoked/amended by the referring court.
          • Facts: about registration of company under Hungarian law (before lowest courts) – ECJ held that review of what was essentially administrative act is still inter-party procedure, between company which wanted the registration, and lower court seeking to deny it.

MHSWR 1999

Impose a duty on employers to make a suitable and sufficient assessment of the risks to health and safety of employees and others affected by their work.

 Reg 2 requires company to have written H & S policy

 Reg 3 employers are duty bound by risk assessments. Purpose is to help the employer to determine what measures should be taken to comply with his duties under the relevant statutory provisions. Just visit the Notaries in London for more information.

 Reg 5 steps taken to ensure risk assessments carried out and the risks identified are sorted out.

 Reg 7 requires the employer to have a competent assistant in applying the provisions of H & S law, i.e. a health and safety official, where dismissal for carrying out any activities in this capacity would render the dismissal automatically unfair.

 Reg 10 every employer must give his employees comprehensive and relevant info on the risks to their health and safety identified by the assessment and the preventive and protective measures.

 Reg 13 ensures that every employer provides adequate H & S training to their employees.

 Breach of the MHSWR can lead to criminal prosecution by the HSE not civil action.

Breach of the other Regs (ie noise, COSHH etc) can lead to civil liability and possible

prosecution by the HSE.

Plan of Attack

1) S.2/3/4 HSWA 1974 (and others if relevant)

2) Regs 3/4/5 MHSWR (and others if relevant)

3) Relevant specific Regs – if breach can give rise to civil liability and/or prosecution by the


4) Common Law – duty of care owed by employer to employee to take reasonable care of his employees’ health and safety. Duty to protect people by having a safe premises, plant and equipment and a safe system of work, inc work related stress.

5) Contractual obligations – employees may also have claims on the basis of breaches of contractual obliges (express or implied) owed by employers to employees. Eg it’s possible for an employee whose complaints of excessive stress are not properly responded to could resign and claim constructive dismissal on the basis that their employer had breached the implied duty of trust and confidence.

Also, courts have been willing to imply similar terms to those existing under common law into the contract of employment. In Walton’s and Morse v Donnington 1997, the EAT held that there was an implied term in the contract that an employer will provide and monitor for employees, so far as reasonably practicable, a working environment which is reasonably suitable for the performance by the employees of their contractual obliges. Here they held the employee had been constructively dismissed when her employers failed to provide her with a smoke free environment, they also held there had been a breach of the implied duty of trust and confidence when her employers failed to deal adequately with her problems relating to those who smoked around her.

Liability for breach of statutory duty

Even though HSWA and MHSWR expressly excludes civil actions, the majority of Regs under

the HSWA do not exclude civil actions thus they may be tried to found a civil claim if an

employer fails to comply with the Regs. To be successful in a claim for breach of statutory


duty, an injured employee must show that the breach is actionable in civil court, duty is owed

to C by D, C’s loss is within the mischief of the legislation, D is in breach of the duty and the

breach caused the loss.

Vicarious Liability

Defined as, ‘the liability of an employer to persons injured by the wrongful acts of his

employees, if committed in the course of their employment’. From the Lister case liability

now turns on whether there was a close connection between the employee’s act and the

nature of the duties he was employed to do.

Breach of HSWA

Vast majority of breaches are remedied by verbal advice and letters issued by enforcing authorities, which are quick, informal and usually effective. More formal methods include improvement and prohibition notices and prosecution.

The HSE can issue:

 Improvement notices are usually served relating to important issues but not where imminent danger exists, e.g. requiring a safety policy risk assessment of a machine or process. The notice must state the legal requirements being broken and require the person on whom it’s served to remedy the contravention within the stated time period, which may not be less than 21 days from the date of its service.

 Prohibition notices are served where activities involve a risk of serious personal injury. Service is effected on the person carrying on or in control of the activities. Notice gives info regarding the complaint but orders the activity to stop until notice complied with. Notice can take immediate effect, or must take effect at the end of the 21 day period allowed for appeal.  Criminal prosecution against companies or its officers for contravening Regs under

HSWA, obstructing or failing to comply with any requirements imposed by inspectors, failing to comply with an improvement or prohibition notice etc. Where person found guilty of failing to discharge a duty under ss 2-6 of HSWA, he is liable on summary conviction to a fine not exceeding £20,000, or on conviction on indictment to an unlimited fine. Terms of imprisonment are also available for failing to comply with court orders, improvement notices or prohibition notices.  HSE carry out a name and shame policy, where they attract adverse publicity to the accused, forcing them to improve their heath and safety policies.

 Breach of HSWA cannot lead to civil liability.

Framework Regulations

Purpose they tell e/ers what to do to comply with the HSWA

they’re basically guides to approaching Health and Safety on a practical level

it identifies the risks, etc

the Regs are criminal as well so you’re prosecuted rather than sued

Some of the more common ones:

The Management of Health and Safety at Work Regs (MHSWR) 1999

The Noise at Work Regs 1989

The Provision and Use of Work Equipment Regs 1998

The Manual Handling Operations Regs 1992

The Control of Substances Hazardous to Health Regs 1999

The Workplace Regs 1992

For the full list and details see Appendix E p305 RB.

Regs intended to supplement the general duties imposed by HSWA 74 with more specific requirements. For more information, follow the abogados Florida blog

Dworkin Taking Rights Seriously, Chs. 2 and 3

  • Under Hart’s analysis, rules become law is either they are valid i.e. have passed all the criteria which the system sets out OR because it is accepted (not enough that people should conform to it by habit but must positively accept it). The rule of recognition is considered by Hart as “ultimate” and therefore cannot be law because it is valid according to the tests of some other law (which, by definition, must be inferior to the “ultimate” rule- that of recognition). Instead it must be law because it is “accepted” by the people. This is distinct from Austin who considered that power was a pre-requisite for making a law, whereas Hart considered that it must have authority.
  • There is a difference between rules and principles. Rules are applicable in an all or nothing way i.e. if a rule is (1) valid and (2) its facts apply to the case, then it determines the outcome. If either 1 or 2 fails, then it contributes nothing to the decision. By contrast, principles do not tend to stipulate a punishment when they are broken, and nor are they conclusive e.g. we have a principle that a man should not profit from his own wrongdoing, and yet adverse possession does just that. Principles merely state goals of morality/fairness/justice. Also, there can be two conflicting principles within a legal system, but this does not mean one is invalid (it simply may mean that each one is supreme in certain areas). However if 2 rules conflict, one must be invalid, to be decided by other rules e.g. preference for the rule with higher authority/enacted later etc.
  • Positivists believe that where a case isn’t clearly covered by a valid legal rule e.g. because of vagueness of a rule, then the answer cannot simply be found by applying the law. Instead judges/officials must reach beyond the law and use their discretion. This discretion is defined as existing where the judge is simply not bound by authority. Hart called for this doctrine of discretion because law is “open textured” (i.e. it is often vague or susceptible to many meanings). If the judges’ decisions are determined by their principles, then the principles are in effect becoming law. Dworkin argues that it would be absurd to deem principles incapable of binding judges in a legal sense, since it would make no sense to say that judges have a mere moral obligation to consider all the principles. If they have such a duty it must be legal. Positivists might counter that although principles have influence on judges, they are not themselves binding in that they do not completely determine what is to happen in a given situation. However this is merely to say that principles are not rules. Also, if 2 principles conflict the judge must take his pick, just like where 2 rules conflict. Other positivists might say that since a principle cannot be shown to be authoritative as a rule can (e.g. by locating it in statute/court decision) it cannot be legal. However Dworkin argues that whether principles are binding can be determined, albeit less conclusively, by looking to things like community practice etc.
  • There is a conflict between whether principles can be binding and the rule of recognition. Principles are not recognised as satisfying the test for legal validity under the positivist model and therefore cannot count as laws. According to positivism, it would be impossible to claim that principles can be laws, since it would be impossible to find a sufficient level of institutional support (from case law or statute) that would satisfy the rule of recognition. Dworkin says the rule of recognition only works for clear cases, but for complex ones principles are invoked before the court to determine the outcome of the case.
  • Hart argues that for a social rule to exist, there has to be (1) a generally observed practice, (2) an understanding that a rule to this effect exists, and (3) sanction for non-conformists. He applies this to the rule of recognition, saying that (1) Officials often impose laws/regulations e.g. tax collection; (2) they point to the “rule” that courts give effect to the laws of parliament; and (3) that judges are considered at fault and verbally sanctioned if they fail to conform to the rule. However, while the example of judges obeying parliament is true as a social rule, the example of a churchgoer removing his hat is a “normative rule” (i.e. it ought to be the case) and only exists in fact if the individuals concerned do in fact have the duty that they suppose themselves to have.
  • Dworkin says the main difference between himself and Hart is that Hart believes that in every legal system a social rule exists that sets the limits of a judge’s duty to recognise any other rule or principle as law. Dworkin says the very fact that judges constantly disagree about what other rules they are to take into account and to what extent they are subordinate to parliament demonstrates that such a rule clearly cannot exist, even when using Hart’s 3-step formula.
  • Dworkin rejects the idea that “institutional support” can found a secondary rule, as the law of recognition would require. He says that it is not morally or politically neutral (i.e. it depends on what the majority of officials at any given time think the best model of government is, such as separation of powers). Thus a theory of law founded on this basis could not be separate from political or moral theories.



  • Dworkin Law’s Empire, Ch. 4


  • Conventionalism is the idea that past political decisions can be made explicit and non-contestable. Legal conventions (a sub-species of social conventions) designate which institutions have which powers, e.g. the convention that HL decisions bind those of the lower courts. Conventionalism also disputes the view that there is always law to enforce. In hard cases the law may be incomplete and the convention of precedent may not apply, and so judges have discretion in the “strong sense” and the constitutional authority to resolve such issues. Conventionalism thus reaches a similar conclusion to positivism. The difference is that positivists claim strong judicial discretion to be evident from analysing the language of the law, whereas conventionalists say this is really what happens, on the evidence, and offers the best account of it so as to guide future practice.
  • The theory also says that judges must treat as law what convention dictates is law, e.g. in the UK conventionally acts of parliament are law and therefore judges have to follow them.
  • Why should past political decisions is decisive of present rights? One answer is that it provides fair warning. The doctrine also states that the past yields no rights tenable in court except those that are made uncontroversial by what everyone knows and expects. If convention dictates the answer, then that must be it. If convention is silent, then judges shouldn’t pretend their decisions come from it (i.e. from statutes or precedents). What about where we want the court to flout convention in order to change the law, as in Caldwell for example? Dworkin says this would contradict the doctrine by allowing “law beyond convention”. Conventionalism demands that in “hard cases” the judge’s discretion should be guided as far as possible by conventions, e.g. by asking what parliament would have done (sue to the convention of parliamentary supremacy).
  • Do we have the conventions that conventionalism requires? Application of statute? Not the uncontroversial/literal application- courts have to interpret in cases of unfairness etc. Who favours which interpretation will realistically depend on moral/political beliefs of each person. There is no convention for dealing with vague precedents or statutes. Also, if every hard case that goes to the appellate court is to be decided explicitly not on the basis of convention (since hard cases are ones not resolved by convention) then the number of exceptions to the doctrine is sufficient to damage our faith in the doctrine per se.
  • Conventionalism fails precisely because judges pay more attention to statutes/precedents than the doctrine allows them to do. They seek to apply them even where an explicit understanding/interpretation of the statute/precedent is not possible and has to be inferred. Judges do not just give up on statutes in hard cases.
  • Dworkin challenges the idea that legislation or precedent should be read in a certain way by virtue of that way being “convention”. If this were true, any criticism of the merits of a statute or precedent would be out of place (if convention is the justification, then merits are irrelevant). In fact, an interpretation comes to be established by consensus, not convention, so that merits are relevant (as a successful argument that an interpretation has no merits can overturn a consensus- though NOT a convention). So many of the UK’s constitutional features are debated so often that they cannot be regarded as existing because of convention. (NB this belief that rules exist because of consensus is perilously close to Hart’s rule of recognition!).
  • Is the ideal of “fair warning” sound and is it enough to justify conventionalism? Surprise per se is not unfair even when my prediction was reasonable (e.g. I bet on a horse thought to be the best which loses). Instead it is only unfair when I have been encouraged to predict a state of affairs by those who deliberately defeat it. However since we do not profess conventionalism, it is not unfair to go against it e.g. there is no consistency among judges in applying conventionalist principles (see above) therefore people aren’t going to rely on it.
  • Conventionalism cannot be justified on the grounds of surprise being inefficient or undesirable since it doesn’t really prevent surprise e.g. in hard cases where conventionalism demands that the judges accept that rules don’t exist and just decide the case in their discretion which is unpredictable. If we really wanted predictability we would use “unilateralism” i.e. only remedies where the law is clear, and hard cases cannot succeed i.e. no discretionary jurisdiction.

Austin’s conception

  • In Austin’s conception of a society there are no legal limits on the power of a sovereign. The only limits on his ability to legislate are his own morality, fear of popular rebellion etc. What about constitutions or courts? Are countries with these not legal systems. NB There is a difference between having legally unlimited power e.g. absolute monarch, and being supreme within a system e.g. British Parliament.
  • After attacking Austin’s model, Hart redefines what law is: The union of primary (rules governing standards of behaviour) and secondary (rules governing the making/changing etc of primary rules) rules. Primary rules impose duties, secondary rules confer power i.e primary deals with “actions”, secondary deals with “operations”.
  • Hart says there is a difference between being obliged and being under an obligation. Austin treats both as the same, in that a person is obliged to obey the law if he will suffer some punishment as a result. This would mean that a gunman’s order to hand over your wallet is law: plainly wrong. Equally this interpretation would mean that where a law requires a person to do something but there is no risk of punishment (e.g. if he has escaped the jurisdiction), then he is no longer under an obligation to comply- clearly wrong. Instead, “to be obliged” really just refers to a person’s motives for doing something, whereas to be “under an obligation” refers to the existence of a social rule, entailing a general demand for conformity and social pressure/sanction is applied to a non-conformant to a great extent. Only being under an obligation can come within law, not simply to be obliged.
  • What are the defects of a non-legal society (i.e. where things are merely done by social pressure, uniformity of beliefs etc)? Firstly it could only successfully exist where there is a small commune and people are able to function thus i.e. couldn’t work where social pressure is an inadequate form of coercion; Secondly social rules are not codified and therefore there would be uncertainty; Thirdly the rules would be “static” i.e. hard to change, as there are no rules governing how to create or change the rules of society. Fourthly is the inefficiency of social pressure as a means of changing the rules e.g. no method of determining if a person has broken the rules and no agency to administer the sanction. Primary rules of obligation and secondary rules of operation resolve these problems.
  • The operation by which primary laws are made and whether something is a primary law depends on the recognition of it as such by officials, judges etc. This is the rule of recognition. The rule of recognition sets out criteria by which the validity of a law can be tested. E.g. “is the council’s by-law valid? Yes. Why? Because it is made in compliance with the powers conferred on it by the minister.” This demonstrates a rule of recognition that councils only have as much power as the executive grants them. The rule of recognition is a question of fact: It exists in fact and for this reason is the “ultimate rule” as it is the rule that tests whether all others constitute laws.
  • Rule scepticism is the idea that talk of rules is a myth, since in reality the decisions of the courts (and the prediction of them) are the laws (not merely interpretation/application of the rules). Hart disagrees since without secondary rules governing the courts’ functions and powers, the courts might not exist at all, or there decisions would be indistinguishable from those of ordinary citizens. Therefore it makes no sense to talk of the courts as being all-powerful in this way.
  • Kelsen views laws as instructions to officials to do X in Y circumstances.