The idea of participatory budgeting (PB) is currently experiencing a renaissance. This is linked to the possibilities of simple mass communication over the Internet. More than 20 years ago, PB began flourishing for the first time. With support from the Bertelsmann Foundation, citizens were to be made more interested in the affairs of their municipalities. This goal was to be achieved by involving citizens in areas where key political decisions were being taken for the future: budget formulation. Initially there were some success stories, but most of these did not last.  In many cases the municipalities tried to present the budget, which is usually complicated and difficult material, in a way that citizens could understand, and in some cases invested significant human resources in communicating this material to citizens at public meetings.
On the other hand, in some cases one might gain the impression that rather than providing citizens with a real opportunity for participation, administrators were rather seeking to demonstrate to a broader public what a difficult and arduous job they had to do. Regardless of how seriously participation was actually taken, in many cases the initiators soon gave up. The press spokesman of the municipality of Monheim am Rhein, Michael Hohmeier, put it like this:
‘We were happy if 20 people came to our information events’.  Four out of six pilot municipalities involved in a project supported by the Bertelsmann Foundation and the Ministry of the Interior of the German federal state of North-Rhine Westphalia discontinued PB as soon as the project came to an end in 2004 (Castrop-Rauxel, Hamm, Monheim am Rhein and Vlotho).
Participatory budgeting – ‘Leaped up like a tiger, landed like a bedside rug?’
In the current debate on the pros and cons of PB, three key points of criticism stand out:
1. The original objective of involving citizens in the complex process of municipal budget formulation was not achieved, and has since been abandoned. Nowadays, it is only collections of proposals for measures and (investment) projects, or voting procedures on proposals made by policymakers and administrators, that operate under the label ‘participatory budgeting’.
2. The number of citizens who participate is low. The vast majority are middle-aged men who left school well-qualified and now have well-paid jobs. The question therefore arises of whether the measures identified through PB possess sufficient legitimacy to warrant their adoption by democratically elected local councils.
3. The low figures for participation raise another problem: how sustainable is citizen participation through PB? In many cases we can see that PB has already been discontinued. The question as to whether ‘participatory budgeting’ as a whole is gradually fading away seems warranted. This is because, as with employee suggestion schemes, in PB fresh ideas are not forthcoming every day. In fact, there is a risk that the debate will keep returning to many similar proposals or even the same ones.
Criticism no. 1: An ambitious idea – but the reality is different
PB in Germany goes back to an initiative of the Bertelsmann Foundation and the Ministry of the Interior of the federal state of North-Rhine Westphalia, which in the year 2000 launched the pilot project ‘Municipal participatory budgeting’ involving six pilot municipalities. In their foreword  the then Minister of the Interior, Dr. Fritz Behrens, and Prof. Marga Pröhl of the Bertelsmann Foundation, state that: ‘All citizens are affected directly or indirectly by the budget decisions of their council. It is all the more astonishing that citizens have little interest in the preparation of budget statutes, and make barely any use of their existing rights of participation’. Yet this is not quite so astonishing, as dealing with municipal budgets is usually unpopular even among the (honorary) members of local councils. This is because it is a time-consuming ‘science’ that is barely comprehensible to people with little or no training. The party political groups like to delegate this to ‘budget specialists’ (as is also the case at the federal and state levels). According to the authors of the foreword, PB should now change this. This is summarised as the salient outcome of the pilot project: ‘Among citizens there is a huge need for information about the budget, and a strong willingness to participate in preparing it’.
Sobering experiences with information events
As mentioned above, in Monheim am Rhein a maximum of 20 people came along to the budget information events. The author personally attended similar events in the cities of Tübingen and Giessen, and the municipality of Hassloch. If we also subtract from the number of attendees the local council members and the administrative staff, then 20 people at Monheim still seems to be a ‘good turnout’. In Tübingen this low willingness to participate had already been noted in previous years. After several ‘sobering’ years, this became a reason for the mayor there, Boris Palmer, to consider discontinuing the event, as in North Rhine-Westphalia.
Four of the six original North Rhine-Westphalian pilot municipalities discontinued PB as soon as the project came to an end in 2004 (when the final disbursement of financial support had been made). In Hilden, a competition was held in 2009 that awarded prizes for three citizens’ proposals. In 2010 a consultant prepared proposals for budget consolidation. By then, ‘participatory budgeting’ meant only that citizens were able to click on the proposals in the Internet forum and write comments.  Whether or not all this still meets the aim of involving citizens in budget formulation, is a question that the reader can answer for him- or herself. Yet perhaps the Internet, and Web 2.0 in particular, would still make this possible.
A fresh start with Web 2.0
In 2009, the Hanseatic City of Hamburg and Freiburg  returned to the idea of participatory budgeting with some highly ambitious aims. Hamburg relied (almost) exclusively on the Internet, while Freiburg opted for a multi-channel process based on moderated public meetings, a citizen poll, and an online platform for participation. In Freiburg, the original objective of PB was revived: ‘Policymakers and administrators aimed to conduct a city-wide discourse on the services provided by the municipality, and to involve the citizens of Freiburg in the budget deliberation process.’ 
Citizens were granted the opportunity to set their own priorities (for expenditure). Not only were they able to make proposals for specific items of expenditure or projects; they were also able to suggest increases or reductions for specific areas of the municipal budget (e.g. schools, green spaces and thoroughfares, sports, public safety and order etc.), and thus prepare their own ‘personal’ budget plan. For the most part the budget areas were grouped on the basis of the division of tasks between the sections and departments of the municipality.
This model was designed either to produce an ‘expenditure-neutral’ budget or, if additional expenditure was necessary, to generate proposals for financing this expenditure. The formulation of personal budgets by citizens was made considerably easier by the use of relevant Web 2.0 tools. In Hamburg and Freiburg so-called ‘budget calculators’ were used, for instance: using a slide bar, each budget item (e.g. ‘the arts’) can be set along a scale e.g. from 0% to 200% (the ‘default’ setting of 100% corresponds to the current draft municipal budget). The screenshot below shows an example from Hamburg.
The Hamburg Internet tool offers citizens a range of opportunities to obtain information, prepare their ‘own budget’ and share ideas and views with other users. Yet it also shows where the problems lie. This is because in order to define the budget for citizens, for the constitutional court and for the audit office, it is necessary to have studied the material very closely. However, the inputs on the left suggest that this was not always the case, and it is questionable whether this can ever be the case for the majority of the population.
It therefore comes as no surprise that the ‘prospective benefits’ resulting from PB are more than vague: ‘In 2009 Hamburg will spend 9.2 billion euros on the various budget items (not including general financial management). In the short term there is little scope for amendment here, although in the medium term fundamental shifts in emphasis are possible. The participatory budget was therefore designed for the period up to 2020, and the changes that would be possible by that point.’  In 2009 some 0.04% of the electorate of the Hanseatic City of Hamburg participated in its PB process. Such rates of participation do not make it any more likely that citizens will seriously engage with the results of PB.
Moreover, in politics a planning horizon of more than 10 years is usually unrealistic. For instance, the school reforms planned in Hamburg in the year of the participatory budget have long since been consigned to history, due to a referendum. From the quote above we may conclude that there will be no further participatory budgeting (in this form) in Hamburg before 2020. Bearing in mind the low participation rate, a repetition of the procedure after 2020 is rather unlikely.
Nor does Freiburg plan to continue PB in the form practised in 2009, despite the fact that this was the original intention: ‘The ex-post analysis of other municipalities shows that in Germany the initiators of many of these procedures find it difficult to get citizens to make a long-term commitment to budgeting processes’. . Yet in Freiburg too, as in Hamburg, conclusions concerning the further (political) treatment of the results or participatory procedures are rather vague: ‘In the political discussions a clear convergence between the votes cast by citizens and the main priorities of budget investment clearly emerged’.  On the basis of the lessons learned with PB (see also next section), the municipal council decided first of all to continue only with the representative citizen poll every two years and, building on the results thereof, to organise an online discussion. Due to the low level of citizen responsiveness and the high costs, the idea of a city conference is no longer on the agenda. Instead, elements of a discourse are to be developed at district level. So, it would appear that citizen participation in budget formulation cannot be organised using Web 2.0 either.
Criticism number 2: Low participation – ‘preference’ for the middle class
In the case of Hamburg we already saw in the last section that only a fraction of the population took part in the PB process.
Low participation rates are the main problem
The participation rates in Freiburg were somewhat higher than in Hamburg: 0.13% for the city conference, 1.22% for the online platform and 1.68% for the poll (all figures as a percentage of the electorate). If we assume that the number of citizens participating in two or more of the above channels is not significant, then we have a total participation rate of just over 3% of the electorate. This is no doubt one of the key reasons why PB was not continued in Freiburg in the elaborate form seen in 2009.
This low level of participation becomes critical chiefly in connection with a further problem: the high costs. The total costs in Freiburg in 2009 were around EUR 680,000. The costs of the various components making up this figure varied widely. The representative poll generated costs of approximately EUR 58 per participant, the online discussion EUR 110. The intensive preparations in conjunction with the high personnel costs for the city conference generated costs of around EUR 1,500 per participant. The problem of costs and benefits arises once again – as it did in the initial ‘pilot phase’ of PB around 1990.
It is not only the ‘sophisticated’ participatory budgeting processes of the kind practised in Hamburg and Freiburg, which require participants to study and analyse the material in depth, that suffer from low rates of participation. PB processes that ‘merely’ collect citizens’ proposals (and usually also allow discussion of the proposals in online forums plus web-based evaluation of the proposals) do not attract higher numbers of participants either. In 2011, Trier for instance had a participation rate of 2.9% of the electorate, which was almost on a par with Freiburg in 2009. (We will return to the example of Trier later on.) Through its PB process ‘Solingen spart’ [‘Solingen is cutting costs’], Solingen reached around 2.8% of the electorate in 2010. However, this figure relates to all users who registered on Solingen’s participatory platform. It was just 1% of the electorate who actually submitted proposals themselves. In Cologne around 10,000 users – 1.4% of the electorate – registered in 2008 and 2009.
The most successful PB process to date is to be found in Potsdam. As in Freiburg, in Potsdam all channels are used in order to organise PB. Using this procedure, which combines public meetings, an online platform and a representative citizen poll, a participation rate of around 4.6% of the electorate was achieved in 2011 – the highest rate of participation in a German PB process to date. The increase in participation between 2009 and 2010 is particularly impressive.
Trend in participation in the PB process in Potsdam 2008-2011 by form of participation (as %, baseline [100%] = 2008)
However, the graphic also shows that the increase in participation figures slows down considerably from 2010 onward, and in the case of proposals and prioritisation even already shows a downward trend. We will return to this aspect below when we discuss criticism number three.
Middle-aged, well-educated men
One general point of criticism directed at all processes of citizen participation is that these processes provide individuals and influential interest groups who are in any case particularly politically active, with additional opportunities to get involved and influence things. At the level of individuals, this primarily means men from better-educated, higher-earning strata, who not infrequently dominate the participant groups. At the level of groups, this means well-organised interest groups and ‘grassroots elites’ that are usually both financially strong, and have human resources or time at their disposal. The criticism that additional participatory offerings enable these segments of society to gain disproportionate and basically undemocratic influence  also applies to numerous PB processes.
In Trier in 2010, only 37% of the participants were women. For Freiburg we discover that: ‘The participants at the conference largely comprised individuals representing the interests of various groups’.  The online procedure for Freiburg’s gender budgeting process also suffered from a ‘shortage of women’: 36% of the participants were women – almost the same percentage as in Trier. Unfortunately, most local authorities that conduct PB do not supply data on the social demographics of participants. At this point, though, we are able to draw on data collected in a survey of citizens on the topic ‘How can we finance the future of the city of Tübingen?’, which was conducted in 2010. Among other things, the survey looked at the evaluation of proposals for cutting costs or improving revenues in the municipality, a topic very similar to conventional PB. The survey was conducted using the Speyer panel of citizens method , i.e. two citizen surveys were in fact carried out. One of these was based on a representative random sample taken from the city’s register of residents, who were approached through a personal letter (409 respondents, approx. 43% of the sample). The other was freely accessible to all citizens online, and was advertised in local newspapers and on the Internet. The second of these two surveys resembles a PB process in which citizens participate online on their own initiative.
It is clearly evident that in the open survey, men and the age group 25 to 64 are significantly over-represented. In the group of those who self-recruit for participation, though, the ‘distortion’ is most clearly evident in the educational status: at 85.4%, people with a higher school-leaving certificate were patently overrepresented here.
These differences in the social structure of the participants will also, depending on the method of participation, inevitably affect its outcome. In Freiburg, all three forms of participation – poll, online forum and city conference – produced the result that allocations for children, youth and schools should be increased.  With online participation, public transport came second among the areas that should receive greater support. With the other methods, however, it did not appear at all among the areas to be promoted. Participants in the online procedure also called for greater support for individual ‘special interests’, such as sports and swimming baths, adult education etc. By contrast, those participating in the survey advocated reducing support for these very areas. With online participation, the interests of those groups particularly drawn to such procedures, i.e. better-educated groups and young people, are likely to have played a role. As was to be expected, the ‘key issues’ at the city conference were in some cases the precise opposite of the results obtained with the other forms of participation. Stakeholders rarely represent a cross-section of the population. According to the outcome of the other two procedures, urban development, construction and housing, and culture should receive less funding. The same phenomenon can also be observed in Potsdam. If we look in detail at the voting on proposals for the 2011 participatory budget, it becomes clear that preference is given to very different interests, depending on the medium of participation (poll, online or public meeting), The vast ‘representative’ majority of citizens polled were in favour of ‘finally building an animal sanctuary’. At the public meetings this topic played only a very minor role. On the online platform, ‘refurbishing the Brauhausberg swimming baths’ was an absolute priority. And at the public meetings too the swimming baths was a relatively high priority. The NowaWiese sports and leisure area, the introduction of a ‘vegetarian day’ and the Westkurve meeting place were popular issues at the public meetings, which was not the case at all in the representative poll. On the Internet and especially at public meetings, it is evidently the case that certain groups of citizens succeed in pushing through their ideas and wishes, either by recruiting a sufficient number of like-minded people to vote as they would themselves, or by themselves participating on the Internet multiple times by using various accounts registered under different names and identities. 
Criticism number 3: Is PB sustainable, or just a token flash in the pan?
The ‘fate’ of most of the first project municipalities to conduct participatory budgeting – Castrop-Rauxel, Hamm, Monheim am Rhein and Vlotho – already indicates that it would appear to be difficult to pursue participatory budgeting over a prolonged period. The two ambitious approaches pursued in Freiburg und Hamburg were also one-off activities. And where participatory budgeting has been conducted over a long period, as in Hilden, after a few years only rudiments of the original concept remain in place. In the case of Hilden this involves the award of prizes for proposals and the possibility of rating proposals made by experts. It is thus very difficult to claim that citizens have a significant influence on the municipal budget.
Dwindling activity in participatory budgeting
In the case of the comparatively successful PB process in Potsdam, it also became evident that the number of proposals and their prioritisation fell for the first time in 2012. The data for the PB process in Trier are similar:
Trend in participation in the PB process in Trier 2009-2012 by form of participation (as %, baseline [100%] = 2008)
It is important to note that the Trier process began by permitting proposals on expenditure and cost-cutting in 2009 and 2010, whereas in 2011 only proposals for cutting costs could be made. In 2012 it was then once again possible to make proposals on expenditure. It is striking that the number of participants (as measured by number of activated accounts) has been rising continuously since the beginning in 2009 (including 2011), yet the activity of participants, especially the number of proposals and the number of ratings, has not been rising to the same extent, but in fact has been declining. Whereas in 2009 0.27 proposals were still being submitted per participant, this figure fell in 2010 (to 0.19 proposals per participant) and in 2011 (to 0.10 proposals per participant). It was not until 2012 that this figure rose once again, this time to 0.13 proposals per participant – thus still falling quite a way short of the figures for the first two years. We also see a similar trend in the number of ratings per participant: in 2009 each participant was still rating an average of 26 proposals; a temporary low point was then reached in 2010 with just 16 proposals and in 2011 with 8 proposals rated per participant. Only in 2012 did this parameter rise again slightly to 10 proposals rated per participant. The case of Trier also demonstrates that citizens evidently find it considerably more attractive to make proposals on how to spend money than on how to save it.
The example of Hilden that has already been mentioned several times shows a clear decline in participation, the longer PB was continued. Whereas in 2003 685 individuals took part in the Hilden PB process, only one year later that this figure had fallen to just 220. In 2005 the number of participants rose again to 333, before falling continuously through 2006 (220 individuals) and 2007 (130 individuals) to reach the low mark of just 70 participants in 2008. This pattern was also repeated in Cologne: in 2008 around 5,000 proposals were received by the city of Cologne – one year later the figure had fallen to just 1,254.
Conclusion: The future still seems open
So far we have collected the key points of criticism of PB and the way it is currently being implemented. Now we will take stock, and also acknowledge the positive aspects of PB:
1. With the collection of citizens’ proposals, followed by their prioritisation and final voting on them, a procedure has been found that can possibly be practised over a prolonged period. It is true that the original intention of integrating citizens closely into the budget formulation process has not been realised. Furthermore, the influence that citizens wield in this procedure is limited, because only a few proposals made by a few citizens have a chance of being realised, even though each and every citizen does have an opportunity to influence things by prioritising and voting on the proposals. Ultimately, a single vote in local elections is likely to have the same degree of influence. The fact that the influence of citizens is not all that significant might also benefit PB, in that members of local councils might then be far more willing to embrace this tool. On the other hand, were their room for manoeuvre to be curtailed even further, the level of resistance from honorary council members would presumably be higher.
2. There remains a trend toward PB. The Fifth Status Report on ‘Participatory Budgeting in Germany’  published in March 2012 identifies 21 newly active municipalities that are conducting PB. This makes a total of 115 active local authorities that are either conducting PB or have at least done so within the last two years.
Of the aforementioned 115 local authorities, however, 60% are either preparing or introducing a participatory budget for the first time, or considering doing so. The Fifth Status Report also now identifies 22 municipalities that are no longer conducting PB (and have not being so for the last two years) – a figure that is rising. This would corroborate the ‘fatigue hypothesis’ put forward above, according to which after a few years participation in PB at best stagnates, and proposals on discussions – as is so often the case on the Internet – quickly become repetitive.
It would certainly be desirable that PB, in whatever form, becomes a fixed element of civic participation in Germany. At the moment there is evidence to suggest that this can succeed, as well as evidence to suggest that it may not. Municipalities such as Potsdam and Trier have developed a form of PB that might at least motivate relatively large sections of the population to participate regularly. This is certainly not the rule in the field of civic participation. No doubt we will know more in a few years.
Translation: John Cochrane
 A detailed survey and discussion of participatory budgeting will shortly be published by the publishing house VF-Verlag. Working title [German only]: K. Masser, A. Pistoia, P. Nitzsche, Bürgerbeteiligung und Web 2.0. Potentiale und Risiken webgestützter Bürgerhaushalte?
 Cf. Bertelsmann-Stiftung und Innenministerium Nordrhein-Westfalen, Kommunaler Bürgerhaushalt: Ein Leitfaden für die Praxis, (German only, published by the Bertelsmann Foundation and the Ministry of the Interior of the German federal state of North-Rhine Westphalia, Bielefeld 2004 [Internet: http://www.google.de/search?sourceid=navclient&hl=de&ie=UTF-8&rlz=1T4DBD...]).
 Ein Leitfaden für die Praxis, op. cit., p. 4.
 Then: http://www.hilden.de/ (click on ‘Finanzen’, then on ‘Bürgerhaushalt’).
 The following remarks on the PB process of the city of Freiburg are based on the essay by Schubert, A., Geschlechtersensibler Beteiligungshaushalt Freiburg 2009/2010, in: Hill, H. (ed.), Bürgerbeteiligung. Analyse und Praxisbeispiele, Baden-Baden 2010, p. 163-171 and on a PowerPoint presentation (slides) of the city of Freiburg presented at the symposium ‘The role of citizens in the municipality of the future’, staged in honour of the 80th birthday of Helmut Klages from 14-16 April 2010 in Berlin. (Downloaded from the homepage of Prof. Dr. H. Hill in June 2010), and the report on the results of the online debate on ‘Participatory gender budgeting in Freiburg 2009/2010’, (TuTech Innovation GmbH), Hamburg 2008.
 Schubert, A., op. cit., p. 163.
 Report on the results of the online debate on ‘Participatory budgeting in Hamburg 2009’, TuTech Innovation GmbH, Hamburg 2009, p. 5.
 Schubert, A., op. cit., p. 164.
 Schubert, A., op. cit., p. 168.
 Cf. Geißel, B., Wozu Demokratisierung der Demokratie?, in: Vetter, A. (ed.), Erfolgsbedingungen lokaler Bürgerbeteiligung, Wiesbaden 2008, p. 31.
 Schubert, A., op. cit., p. 167.
 More detailed information on the Speyer citizen panel can be obtained from the German Research Institute for Public Administration Speyer (FöV) (also available as a download) at: http://www.foev-speyer.de/publikationen/pubdb.asp?reihen_id=1. A brief description can be downloaded from the website of the Konrad Adenauer Foundation: http://www.kas.de/wf/de/33.15642/ (accessed on 07.01.2012). A good general overview that can be recommended is: Klages, H., Daramus, C. and K. Masser, Bürgerbeteiligung durch lokale Bürgerpanels. Theorie und Praxis eines Instruments breitenwirksamer kommunaler Partizipation, Berlin 2008.
 Schubert, A., op. cit., p. 168.
 Landeshauptstadt Potsdam, Bürgerhaushalt in Potsdam 2012, http://www.potsdam.de/cms/dokumente/10079486_1254717/f144ebf9/111028_Top....
 The status reports are published regularly on the Bürgerhaushalt.org website.