In the last several months, all eyes were on Kenya’s presidential election. This election was significant not just for the decisions that voters made about their leadership, however, but also for the policy reforms that it set in motion.

The first election after approval of the 2010 Constitution was a key milestone in the transition to a novel system of devolved government in Kenya.

Many analysts have noted that Kenya’s devolution is of the “big bang” variety. What does this mean? Mainly, it refers to the fact that Kenya is creating a new level of government and giving it political, financial and administrative autonomy all at once.

In other countries, devolution has often been piecemeal. For example, new levels of government are created, but perhaps their leaders are not elected and/or they have limited control over their finances.

Later, they may be given control over finances but still no elections. Local leaders may be appointed for many years before a new round of reforms introduces local democracy.

The Kenyan approach is radical in that all of these things are happening at once. Counties have fully elected leaders, control over their finances, and administrative control in their areas of responsibility as laid out by the Constitution.

True, the Constitution foresees three years of transition, and during this period, the financial and administrative control is partial, while the political control is total.

Across the world, central governments often try to undermine local governments and maintain their grip on finances and administration. Their ability to do this depends in part on the sequence of decentralisation.

Data from Latin America suggests that in countries where political decentralisation comes later in the sequence of reforms, local governments are weakened.

Therefore, one encouraging aspect of the Kenyan sequence of reforms is that the governors and assemblies are in place to start demanding their fair share. This will make it much harder for the national government to claw back financial and administrative control.

It should be noted, however, that the Constitution gives national government control over “policy” in many areas, such as health and agriculture, and this administrative power may be used to reduce county financial control by forcing counties to spend money in particular ways.

At the same time, the share of total revenues constitutionally guaranteed to county governments (15 per cent “equitable share” plus what they raise in local taxes, which has been about an additional 2.8 percentage points using local authority estimates from 2010/11) is relatively low by international comparison.

In the Philippines, for example, the unconditional grant share for lower levels of government is 40 per cent of revenues, plus a share in taxes.

South Africa gave provinces and local government about 36 per cent of revenues through its equitable sharing formula in 2012.

Based on Commission on Revenue Allocation recommendations, Kenya’s guaranteed share is too low for the functions they are to perform (CRA has suggested counties need over 30 per cent, which may be slightly high). The fact that counties are guaranteed less than what they need suggests that their independence can be undermined.

A glance at Uganda’s local authorities after the elimination of their main source of own revenue, the graduated tax 2005, indicates that this has eroded their autonomy.

A 2010 study of Tanzania’s devolved units found that they have similarly been forced to implement a series of national policies through transfers with little fiscal room to respond to local concerns.

Overdependence on transfers from the centre in the Philippines has led to debate over whether local governments should get greater taxing powers. A similar debate is likely to start before long in Kenya.

Note too that Kenyan counties also cannot borrow freely, but must get approval from national government, which must guarantee any loans they receive.

This is logical from the perspective of avoiding costly bailouts of counties by national government (a problem that has plagued other decentralisers), but it erodes county autonomy further.

Of course, county autonomy should not be reified. Local governments are not always efficient users of public money. They may spend heavily on administrative costs, as they have in Philippines and as Kenya’s local authorities tended to do.

National control of borrowing is one response; funding counties through conditional grants rather than augmenting the unconditional “equitable share” is another way to ensure that counties spend on priority areas.

If greater citizen access to key social services is a goal of devolution, then unmitigated county autonomy is not always the answer.

The theory of devolution rests on the assumption that citizens can and will hold local officials to account for services more effectively than national officials. This assumption has not always proved correct.

One key consideration is whether voters and officials understand their roles under the new system. If they do not, it is not clear that they will know what local leaders are accountable for. Citizen preferences must also align with improved service delivery.

Where citizens see officials as a source of private benefits, such as burial costs or school fees, they may not hold them accountable for policy.

Moreover, Kenya’s devolution has created powerful county assemblies to check county governors, yet almost no attention has been paid to these assembly members.

More than a week after they were elected, official results for county representatives had still not been reported. There appears to be little awareness of what county assembly members are supposed to do.

This lack of understanding threatens to undermine formal checks on the governor and leave this responsibility in the hands of citizens who may not yet be fully prepared to exercise it.

For voters to know what leaders are supposed to deliver, there should be a clear division of responsibilities (functions) between levels of government.

It is common in decentralised systems for various levels of government to shift the blame for service delivery to other levels, which is facilitated by insufficient clarity of roles.

Unfortunately, Kenya is off to a slow start in this regard as there have been serious delays in the process of assigning functions to each level based on the Fourth Schedule of the Constitution.

Not only does this mean devolution is starting without this basic clarity, but money is also being shared without knowing exactly what it is to be spent on.

A final concern is what will happen to inequality within counties. In some countries, when resources are devolved, they do not reach everyone.

Those residents and ethnic groups concentrated close to the capital of the local government are able to monopolise resources, further marginalising others and stirring conflict.

The best antidote to this is good data on intra-county inequalities and the use of conditional grants that target the most marginalised.

Kenya needs to have a broader discussion about how all resources flowing to counties—conditional and unconditional alike—should be organised and shared to ensure inequalities are reduced, not exacerbated, by devolution.