Program Budgeting: Program vs. Organizational Structures

[Marc Robinson]

One of the thorniest problems for program budgeting is the relationship between programs and organizational structure. If ministries have major organizational units which straddle several programs, significant practical problems arise in linking organizational unit budgets and program budgets. Read more at: http://blog.pfmresults.com/wordpress/?p=115

[STONE]

"Parking the thorns" is indeed a good pragmatic approach to the question raised by MoF staff on "how to match programme structures with administrative structures". 

In Albania, the government, acting on TA advice, introduced a programme structure some 8-10 years ago.  Initially this was set centrally and faced resistance from line ministries as it was seen to add a classification burden on to the way they continued to budget and manage things - by administrative organisation (directorates).  Things got better when the management arrangements for medium term budget programme (MTBP means MTEF, if MTEF has meaning) preparation were changed with the introduction of "programme management teams". 

The Head of these PMTs is the General Director who has most responsibility for the programme and the team brings in directors of administrative structures whose directorate is involved in the delivery of the programme's service delivery outputs.  It cuts across directorates.  As this institutional arrangement gained experience there was increasing demand from line ministries to adapt the programme structure, mostly to improve it.  There were some cases where the changes were an attempt to collapse back to the administrative arrangements  and some tendency towards the mish-mash that Robert Clifton mentions on your blog.  The MoF resisted this, often on grounds of preserving the logic of a programme approach and to avoid the creation of sub-programmes, but the resistance was also driven by a desire to minimise the number of programmes (to reduce work load demands on the challenge function of the MoF).  The advisors view was that a programme is a meaningful and manageable group of service delivery activities producing outputs that contribute towards the delivery of the programme objectives and so was  a policy and management issue for line ministries to decide with the MoF keeping the principles in mind.  That is, it would take time and learning from experience to develop a good programme structure. The tailor designed one suit but customers come in different shapes and sizes.

In Albania, the thorns are parked in the Planning Management and Administration programme that is common to all line ministries.  LM are free to move expenditures out of PMA into other programmes as they feel they can.  The principle is that whilst uniformity of approach to classification has some virtue, the goal of the reform is to improve programme service delivery and that should take precedence over uniformity.  It allows some PMTs to move faster and so provide experience and examples for other programmes.

The lesson I draw was that progress in programme budgeting is to ensure that the management arrangements support it and the transition from the administrative approach.  I recall a Budget Director in MoF saying that LMs only really understood the original approach to programme classification when the management arrangements were put in.

The  ideal is for the administrative structure  and management to collapse into the programme structure and management arrangements and this may come in time. Heads of Programme Management Teams are accountable for output delivery and that could work as an incentive to change day to day management systems to match the strategy and budget planning management arrangements.  Robert Clifton is right to say that constant and consistent attention is needed to avoid mish-mash.  The Budget Directorate in Albania has been successful in guarding against this.

[marybetley]

This is a fundamentally important issue in programme-based budgeting.

The appropriate definition of the programme structure is critically important and will likely involve both organisational and institutional change.  The effectiveness of programme-based budgeting (PBB) depends crucially on its ability to provide incentives for change and accountability for results.  This means that policies and trade-offs must be clear, and that there should be a single manager accountable for each programme.  Hence, a line ministry’s organisational structure and its policies reflected through its programmes should be inextricably linked.  Given that the process of introducing programme-based budgeting will involve significant stakeholder consultation and is likely to involve both institutional and organisational change, this alignment of programmes and organisational structures will take considerable time.  Lessons from where PBB has not worked support this point and suggest that changes to the budget process that these reforms imply are quite fundamental, and require changes in staffing level, professional capacities and organisational structure.

Whilst including a programme classification in the Chart of Accounts is often part of implementing a programme-based budget, it is rarely the first step.  Indeed, lessons from experience in reforming expenditure classifications indicate that caution is required in expanding the different types of classifications to be used.  According to these analyses, such an expansion may result in unreliable information due to the complexity of the budget nomenclature. It also requires more capacity and resources to generate the required information and maintain the system.

Supporting these two points, guidance on programme classification based on practical experience is clear: when establishing a programme classification, it is important to ensure that: (i) clear responsibility for managing the programme, and accountability for its results, is allocated to a specific unit and program manager within the ministry or department concerned; and (ii) the requirements for data collection and analysis are kept within reasonable bounds.

This conclusion is supported by other research which indicates that administrative capacity determines the pace of implementing more results-orientated budgeting (such as programme-based budgeting).  The underlying financial management basics need to be in place to create a suitable governance platform for realising the potential of PBB.  Where underlying internal control systems are weak, there is poor financial management, and a lack of skills in planning and financial management, should proceed with caution in implementing PBB.

[petagny]

While I agree that introducing the programme classification into the CoA might not necessarily be the very first step, for how long can programme budgeting stay interesting without the ability to report actual expenditures by programmes? This must surely be something to think about fairly early on and build into a structured implementation plan, rather than leaving it as an unresolved issue?

[LizMuggeridge]

Introducing programme classfication in itself is fairly straightforward.  A programme defines the function of an organisation and most organisations are already set up along functional lines, e.g. the Ministry of Education will have a primary education department and a primary education programme.  Where the programme structure can help is to then include within the programme other parts of the Ministry of Education which also focus on primary education including District Offices.   Where this can become confusing is for government ministries whose organisation and programme structures are essentially the same, so they ask, what is the point of the programme structure? 

I have noticed a recent confusion about programme budgeting and performance budgeting.  In some countries where I have assisted governments to introduce a three year budget process that includes programmes and performance indicators (we didn't give it a name) this is now being changed by the next round of consultants, to what is being called programme budgeting.  The rationale for this is that the previous process was too cumbesome and time consuming.  Performance based budgeting can become very cumbersome if ministries define their outputs and activities at a very detailed level, i.e. buying stationery is an activity.  Thus the focus needs to be on appropriate definitions of outcomes, outputs and activities so that the process is concerned with a strategic overview of performance rather than detailed activity planning.

[petagny]

The previous post seems to assume away the issue initially raised by Marc Robinson:

'One of the thorniest problems for program budgeting is the relationship between programs and organizational structure. If ministries have major organizational units which straddle several programs, significant practical problems arise in linking organizational unit budgets and program budgets.'

Do governments organise themselves neatly along functional lines? Perhaps more often in education, but even here there is not everywhere a clear demarcation between secondary and primary education, and bodies with responsibility for curriculum development, teacher training and inspection can serve both primary and secondary.

[Ronb]

From my experience, I think the issue of programs vs organizational structure really comes down to where programs are defined.  If they are created at the national level, there is a much greater probability that the programs will cross over more than one organizational unit.  It has been my experience that initially, it makes sense to focus programs at the Ministry level and have them define what it is they do and the strategy to achieve it.  Here too there can be sub-organizational cross over, but it will be much easier to develop coordination techniques such as a program coordination committee, a reorganization to align organization with the program or parceling individual activities to differing organizational units (each with their own budget and management structure). 

I have found that once there is comfort with this level of program definition moving to broader programs that transcend individual ministries becomes easier as the participants already have knowledge of what programs are and how they create control and accountability within the organization.

[Martin Johnson]

I have just been browsing the last few posts and it seems to me that whilst RonB’s main point is undeniably correct (i.e. that linking budgets with policies through an appropriate administrative and accountability structure is easier to achieve within ministries than it is between ministries), it still seems to by-pass the thorny issue of, well, dealing with the thorns. As Stone notes, when Programme Management Teams in Albania began to look critically at the existing (centrally-imposed) programme structure, there were instances where the tendency was to gravitate programmes to the existing administrative arrangements (irrespective of corresponding overlap between organisational units and otherwise well-defined programme areas). In this case (several years after programmes had initially been defined), it was the MoF that was the conscience of good programme specification rather than the line ministries. Putting aside for the moment the issue of whether programmes should be specified exclusively within ministries (which depends very much, of course, on how the concept of ‘programme’ is defined and also on how the concept is to be used in practice), the main and thorniest issue remains that of either reconciling administrative arrangements to well-defined programmes and/or specifying management arrangements to enable an administrative structure not otherwise suited to programmes to deliver those programmes well (policy design, planning, budgeting, delivery, accounting, reporting). This may be easier in some ministries than others, but as Petagny notes, it should not be taken for granted even in ministries where key functions are reasonably self-evident, such as in education.

Focusing attention for the moment on the scope of programmes, and particularly whether or not they should be defined across ministries, from a planning, management and accountability viewpoint the common sense position would be to avoid this at all costs (unless absolutely necessary). This does not mean to say that groups of programmes (within and/or between ministries) cannot be oriented to a higher and more strategic level of policy achievement than that defined for and linked specifically to individual programmes. If one goes back to basics in terms of MTEF concepts, in fact, and if one is in a programme budgeting world, this is what one has – a given sector strategy (e.g. for education or, indeed, for a major subset of education, such as tertiary education) that is to be delivered through a variety of different education programmes, whether they come exclusively under a Ministry of Education or not.

[petagny]

There's an interesting post on the IMF PFM-Blog on redesigning the administrative/organisation in Tajikistan:

http://blog-pfm.imf.org/pfmblog/2011/02/budget-classification-design-and-implementation-tajikistans-administrative-segment.html#more

It finishes with this:

'As stated in the IMF’s fiscal transparency code and manual, a well-designed administrative segment should be aligned accurately with the policy and management roles within the government sector, the structure and functions of government, and the responsibilities of different levels of government and the relationships between them. Thus, the administrative segment is the mechanism to ensure that fiscal management and government management institutions are aligned (i.e., that they have common accountabilities) so that their effectiveness and efficiency are not undermined by contradictions between them.'

So, based on this statement, if you get the administration/organisation classification right, do you also need a program classification?

[Martin Johnson]

“ … a well-designed administrative segment should be aligned accurately with the policy and management roles … [etc.] … ” (my underlining).

Interesting point this. It appears to suggest that admin arrangements should be flexible in response to policy, management and accountability requirements. It doesn’t, however, appear to advise on what to do when admin arrangements are inconsistent with policy, management requirements etc. where admin arrangements are prohibitively difficult or costly to change. It does, however, appear to assume a starting point where policy and management roles, structure and functions of government etc. lend themselves to good and appropriate budget-policy links, so that the first step is to verify admin arrangements and change these if required (and otherwise leave them as they are).

In my experience and, I suspect, that of Petagny and most of our colleagues, policy and management roles with regard to budgets, planning and policy are often not that clear clear. So, is it not the case in such situations that creating a sensible programme approach is, in fact, about sorting out policy and management roles (and related arrangements) where these are unclear and/or where do not lend themselves well to good and appropriate budget-policy links? If so, then whilst a new programme approach may not always be required, as Petagny suggests, adopting a sensible programme approach does have its place.

[STONE]

I had a quick look at the article on the Tajik administrative structure and found it interesting.  Tajikistan appropriated the budget using functions and then had various complex management arrangements for spending the money but did not appropriate by administrative unit!  'Programmes' were identified as a fourth level in the functional classification and it seems that the Tajik challenge was the reverse of the issue we are discussing here - that of coming up with an administrative classification to fit an existing programme classification!

(The article is very good – download through the IMF blog link – it has one of its purposes as preserving the institutional memory of the reform process – a lesson to all!  I do wonder though if once upon a time there wasn’t an administrative classification of the budget that changed to the functional one…, but that institutional memory might have been lost….)

Petagny’s question …  When advising on introducing a programme structure the implicit view is that the admin classification is no longer (if it ever was) fit for purpose of efficient and effective management of service delivery. Because it is hard to change admin structures we put a programme structure in on top or alongside, focusing on cost of service delivery and, if we get all excited about performance as well, come up with new institutional arrangements for management responsibility and accountability for programme results.  In Albania, Petagny’s question was asked the other way round.  The answer for me is as Daniel Tommasi puts it (and so much more nicely than my earlier post saying ‘the admin should collapse into the programme’).

“Idéalement, il serait souhaitable de procéder conjointement à la structuration du budget en programmes et à une réorganisation administrative de manière à faire correspondre l’organisation administrative et la structure programmatique. Toutefois, les réformes administratives demandent beaucoup de temps. De plus, cette correspondance est en général difficile à établir au niveau régional, les échelons régionaux étant souvent responsables de la mise en œuvre de plusieurs programmes. À défaut de réforme administrative, il convient de définir des procédures adéquates de dialogue entre le responsable de programme et les responsables des services mettant en œuvre plusieurs programmes.”

Tommasi p.63 (see section of the board on have you seen this? for the link).


(Ideally, it would be desirable to combine the introduction of a programme structure with an administrative reorganization in order to match the administrative and programmme structure. However, administrative reforms take a long time to implement. Moreover, this correspondence (match) is generally difficult to establish at the regional level, regional level management structures are often responsible for the implementation of several programmes. In the absence of administrative reform, it is necessary to define adequate procedures for dialogue between the programme manager and heads of services (administrative units, departments directorates) involved in implementing the programme.)


Why are administrative structures no longer fit for purpose?  I suppose it is the natural bureaucratic tendency to respond to new developments and initiatives (regardless of whether they will have a long and useful life) by creating and new (and permanent) structures with new staff. 

I, of course, share Martin Johnson's view.  For me, the approach of putting a programme classification on top or alongside the administrative classification has always been a Trojan Horse used to tackle this tendency.  But it looks like Odysseus is having trouble opening the trap door.  (Yes Bernard, I know Paris wasn’t a bureaucrat and Helen was not a Homerian reference to the beauty of administrative classifications).