reduction. Institutional reform involves the emergence of new patterns of economic and
political control which fundamentally change the operation of the economic system. An
important component of this, governance reform - now a central part of developing
countries’ policy packages - is in part aimed at improving service delivery. Success in this
area could change the relationship between public spending on health and education and
outputs in these sectors. As countries reach certain stages of development, it is said that the
decentralisation of some policy levers to regional governments can be good for government
efficiency and service delivery. This is an example of a qualitative policy shift which gives rise
to new instruments and relationships between instruments and objectives. Other important
types of institutional change in developing countries are the transfer or sharper delineation
of property rights and the formalisation of informal sectors of the economy, both of which
may activate new instruments and bring new mappings between instruments and objectives.
The distinction between qualitative and quantitative policy shows just how complex the
policy problem is for developing countries, especially in attaining a broad set of targets such
as the MDGs. Structural changes often take place during times of political and social
instability, common features of developing countries. Policy analysis therefore needs to take
into account of political and social factors and likely conflict and contestation as new
structures emerge, triggering new relationships between instruments and objectives.
Uncertainty and learning
Points (7), (9) and (10) are prominent in the MDG policy problem because of the importance
of structural change in developing economies. When uncertainty exists over the parameters
of the model describing the economy, policy decisions are dependent on the statistical
distribution of these parameters. In this case we need to consider learning effects as
instruments and objectives move over time; policy making then becomes a drawn out
process of discovery.
In a situation of passive learning, policy makers’ estimates of the economy’s structural
parameters change as new information emerges. Under active learning, this updating takes
place as a direct result of policy makers’ manipulation of instruments which allows them to
discover more about the behaviour of the system (Kendrick, 2002; Petit 1990). Policy actions
then have a dual purpose of bringing the economy closer to the desired path and reducing
uncertainty about the operation of the system. Under active learning there may be a trade
off between system performance and learning: certain policy actions may lead to a worse
system performance at a point in time compared to others but yield better information
about the operation of the economy, helping to give rise to better outcomes in the long run.
Active learning is critical if there is large uncertainty about the parameters of the economy
or when the economy is going through structural change. Both come into play in the MDGs:
there is clearly much uncertainty about the causal mechanisms of developing economies
while development itself is a process through which the economy undergoes structural
evolution equivalent to changes in underlying parameters. Important questions therefore
surround the ability of governments to use new information to refine their policy making so
as to move closer towards the MDGs. Critical parameters, new information on which may