Gradual transformative policy change has attracted increasing attention in recent times. However, existing explanations for the evolvement of the various modes of such change do not account for the direction and content of change and also have difficulty with more complex policy change processes in which several modes are employed. Ideational analysis can fill this gap, albeit most ideational research does not address gradual policy and ideational changes. Based on a process tracing of two cases from Israeli immigration policy, this article argues that a low level of discretion and a strong status quo bias are conditions that enable institutional conversion through layering. Ideational change in policy solution and problem definition ideas, and the interaction between the two, can explain this pattern and its direction. The article not only expands our knowledge of gradual transformative change, but also moves ideational research a step beyond punctuated equilibrium.

How do institutions transform? To answer that question, this article introduces a dynamic theoretical framework of gradual institutional changes. Instead of looking at each mode of gradual change—like layering or drift—as a stand‐alone process, we examine how the application of one mode of change affects the opportunities of change agents to induce additional modes of gradual transformation. We first point to the fact that any single mode of change produces a real but limited transformation. Nevertheless, since the application of a gradual mode of change alters the institutional context, it opens new change opportunities by affecting the support in the targeted institution and/or its internal coherence. Consequently, change agents who aspire to comprehensive transformation will be able to use these new opportunities to implement additional modes of gradual transformation. Two case studies of gradual social policy transformations in Israel exemplify these theoretical assertions.

Facing barriers to change, policy entrepreneurs often shift their activities to a more favourable venue. In the new venue they either pressure the system from the outside, leading to policy punctuation, or make incremental changes within the existing system which over time will accumulate to a significant change in the status quo. This article aims to expand our understanding of policy entrepreneurs' role in the policy process by examining the strategies they use in the incremental path. Specifically, it focuses on one gradual change strategy, conversion – redirection of existing institutions to new purposes. Based on the case of policy entrepreneurs in the Holy Basin of Jerusalem, the findings indicate that when the entrepreneurs shift the venue to the bureaucracy they can establish cooperative relations with the government that will provide them with the needed capacities to exploit the gaps in the existing rules and redirect them to serve their aim.

Policy drift occurs when actors block attempts to adjust policies to changing realities, thereby changing policy impact. We know much about the conditions for policy drifting, but lack theorization of the conditions for reversing the drift by updating the policy. This article examines when and how actors favouring drift decide to reverse it. These actors have different characteristics from those favouring policy update. Based on the case of prescription drug coverage in the US, the article argues that actors promote drift reversal when maintaining the current drift becomes politically risky. Even then, they will act only when able to limit the scope of the reversal. In addition, in order to implement the policy update, actors will use blame avoidance strategies to gain the support of their constituency and soften opposition.

One of the most important phases of case-oriented research is identifying the relevant population of cases. The best practice scheme suggests first identifying the positive and then the negative cases, based on the assumption that the population is small and known. This scheme is inapplicable when the population is unknown, as is the case with many public administration and public policy outcomes, which are often less visible. However, most research either ignores this difficulty or recommends a single case study. This article suggests that when the population is unknown scholars can apply the possibility principle to identify the relevant population of cases. In order to do so, a large dataset is required. In policy studies, the Comparative Agendas Project (CAP) can serve this aim. This article highlights CAP's applicability for case selection based on the possibility principle in two less visible policy outcomes: conversion and drift. In that, it not only addresses an overlooked challenge but also promotes both CAP as a useful tool for qualitative researchers and the expansion of comparative work on conversion and drift.