Not only the restorative practices but the researchers come from different cultural backgrounds. They are aware of self-selection bias, but random controlled trials have not yet been used. In several countries (notably Germany, but not England and Wales) they have the advantage of national statistics. Finland (pop. 5m) had 11,120 cases in 2008 at a cost to central government of only €6.3m, Norway (pop. 4.5m) had 9,000 in 2007.  

Pause to consider how many cases countries with populations of about 50m should have if they are to be comparable. Some include recidivism and cost-effectiveness in their studies, but there is awareness that some of the important factors, such as participants’ sense of justice, are difficult to measure. Research is sometimes premature, before a programme is well established; early research in England and Wales was particularly uncoordinated. There is some awareness that mediation is also a preventive approach, as well as ‘restoring social peace and the values of the norms that have been violated’ (p. 102) - but can it do that when mediations are normally held in private? 

The main referral agency in several countries is the prosecution service. In Belgium prosecutors must consider mediation in juvenile cases, and may do so with adults. In Germany prosecutors must consider mediation at every stage, and in some countries there are also referrals by the police. Some, like Italy and Poland, have only ‘islands’ of restorative practice. In Norway and Poland mediators must be volunteers (paid a small honorarium); in Poland employees of the criminal justice system cannot be mediators, but in Italy most mediators are professionals working in their own time.  

Some countries deal primarily with adult offenders, mainly violence (nearly 50% in Germany, and 25 % domestic violence in Austria). Some have many civil (neighbourhood) cases (Norway, the Netherlands), and may include mediation in schools.  

Some legislation provides for restorative justice only as one of an array of diversion measures.  Central direction is provided by the government (Norway), an NGO (Poland), or is lacking (Italy). In Finland it is organized by the Ministry of Social Affairs and Health, not the Ministry of Justice.  

Three main types of research are used: action research, descriptive, and evaluative. The main practitioner of action (‘accompanying’) research is Austria, although Hoyle et al.’s research in Thames Valley, England, is also in that category. Researchers can be go-betweens between professions. Descriptive research in Germany started with four model projects; the idea spread widely, but projects varied from fewer that 10 to more than 100 cases per year. One study found ‘a gap between ideal and reality’ in Bavaria (p. 128) – it is not alone in that! Some countries do not require mediators to be trained. Evaluative research in Finland in the 1990s found that 75% of referred cases led to mediation, of which 80% led to agreements, of which 80% were fulfilled. Heterogeneous financing arrangements caused problems, and led to legislation in 2006; it is intriguing to note that a project on mediation in domestic violence was funded in part by the Slot Machine Association.

There is no mention of feedback from mediators to assist crime reduction and social policy; the authors recognise that recidivism is affected by factors outside the restorative justice process, but research often does not take this into account. Inadequate record-keeping and obstacles to access to data can be a problem. Although some countries have produced standards, there is little mention of complaints procedures. 

The book is written in good English, with explanations of countries’ varying procedures;  occasionally Continental terminology slips through, such as ‘general prevention’ and ‘special prevention’ for ‘general deterrence’ and ‘individual deterrence’ (all of which are mostly wishful thinking by legal theorists anyway). A useful six-page list of research reports is given, mostly in English or with English summaries. It is unfortunate that there is no index – much needed in a work like this – and the detailed contents list is an inadequate substitute. Despite the delay in production (most of the data only go up to 2008 or earlier, and it was not published until 2010) this is a valuable compendium, and a welcome addition to the English-language literature which relates mainly to Anglophone countries. 

It is a source book for ideas about different styles of restorative practice and of research, showing the value of criteria other than recidivism and methodologies other than random controlled trials. It should be useful to policy-makers, although as Vanfraechem and Aertsen say, with pointed understatement, ‘Political agendas on crime do not always display the highest degrees of rationality’ (p. 271). The sponsor, the European COST programme, has performed another useful service.