The movement toward greater neutrality has two dimensions: the lowering of the average level of protection and the reduction in the average variance of protection. If the dispersion of tariffs is not reduced as the tariff average is reduced, the tariff structure may not become more neutral. Indeed, a reform that reduces tariffs on intermediate and capital goods but leaves intact those on final outputs could increase the level of protection afforded to domestic value added, even though it reduced the average level of tariffs. It is possible to reduce at the same time both the average level of tariffs and their dispersion.
Governments have approached the task in several ways: an equiproportional cut in all tariffs, an equiproportional reduction of the excess of each tariff over some target level, higher proportional reductions of higher tariffs, or some combination of these and other methods. As a rule, simple schemes widely applied work better than case-by-case and fine-tuning methods. Some tariff reforms have attempted to target the effective, rather than the nominal, rate of protection (the Philippine reforms of 1981-85 are one example). This is complicated and may misfire because of measurement problems. Many economists favour the so-called concertina approach to tariff cutting. First, all tariffs above a certain ceiling are lowered to that ceiling; next, all tariffs above a new, lower ceiling are lowered to that ceiling; and so on. This should yield the lowest adjustment costs without leading to inadvertent increases in effective protection. Chile's tariff reductions in the 1970s more or less followed this scheme.
Lessons about the amount of time necessary to eliminate quantitative restrictions and tariffs are difficult to draw. Some reforms have taken a long time - Korea and the countries of southern Europe, for instance, have still not completed their reforms after at least two decades. Fewer have been completed within the medium term - the process lasted five years in Chile, for example. But none have been fully implemented over the short-term. There is no obvious relationship between the length of the period of policy reform and its chances of success. But the apparently low adjustment costs in most trade reforms, together with the danger that more lengthy reforms will be less credible, are arguments for faster reform.
Some tariff reforms have used institutions, typically tariff commissions, either to set tariffs on a case-by-case basis or to hear appeals for exceptions to the reforms that have been scheduled. Tariff commissions such as those in Australia, New Zealand, the Philippines, and Sri Lanka have often approached their task with too many objectives so that their work has probably not contributed to increasing neutrality.