While groundbreaking, the studies on tool classifications were fairly preliminary and in some cases somewhat rudimentary. However, they provided clear directions for future studies to take. Hood’s taxonomy of policy instruments or design elements, for example, failed to distinguish between procedural and substantive uses, or between different levels of policy means, but was a valuable start to consideration and development of more comprehensive and accurate taxonomies of these instruments in later years.10 And Doern and his colleagues did not deal with the subject of policy mixes, or the selection and design of not just a single instrument, but rather a ‘package’ or ‘bundle’ of different types. However, what Doern and his colleagues had actually discovered was a link between the basic tool types set out in Hood’s NATO model and the willingness of governments to use these different resources against specific target groups (Woodside 1986; Baxter-Moore 1987). That is, the Doern model centred on the relationship existing between policy tool choice and a specific kind of governance arrangement which predetermined, among other things, an ‘appropriate’ mode of co-ordinating state and societal actors. And, third, both sets of studies tended to ignore the ‘micro-level’ of the specific reasons why some permutation within a general tool category – such as the setting of a specific level of tariff or subsidy, or the use of a loan rather than a grant, would occur.
Fortunately, however, earlier generations of implementation scholars had not completely neglected procedural instruments, and other works, like those of Gunningham and his colleages (Gunnigham et al. 1998) had begun to deal with the problem of the design of policy mixes. And others, like those of Schneider and Ingram, developed more sophisticated notions of the inter-relationships existing between governance modes and target group behaviour (Schneider and Ingram 1997). Meanwhile other scholars such as Salamon had also turned their attention to the issue of micro-calculations and calibrations, and developed ideas about the sets of factors policy formulators took into account in fine-tuning their instrument choices (Salamon 1989; 2002).