Before turning to different perspectives on the link between government ideology and retrenchment policies, it is necessary to address a question that must logically be addressed before any assessment of causal relationships and is still regarded as the most pressing question in welfare research by some scholars: How much change is there to explain (e.g., Pierson 2011: 12)? This intricate question, in turn, prompts a range of four more concrete questions that this chapter answers one by one: What is to be understood as retrenchment? How can we operationalize retrenchment? Which empirical patterns exist? And what are the criteria to assess the significance of the changes we find?

What Is Retrenchment? Searching for a Definition

Scholars have emphasized the need for a clear specification of what we can precisely understand by the term retrenchment (Clasen 2005; Sainsbury 2001); and yet, no consensus has been reached so far. Those attempting to pin down the term readily concede its “relativity” (Wintermann 2005: 43), and even in literature reviews on retrenchment, the term is only implicitly defined (Green-Pedersen and Haverland 2002; Starke 2006; Wintermann 2005). Its etymological roots can be traced to the word retrancher from sixteenth-century French – a combination of re (expressing reversal) and trancher (to cut, to slice). Lexically, retrenchment simply means cutting expenses or (cost) curtailment. While political scientists use the term in a broadly similar manner to refer to nonexpansionary changes in the welfare state, the precise meanings may vary considerably. The remainder of this section reviews, discusses, and balances the pros and cons of different conceptualizations in order to extract a working definition that is applicable in a large number of countries across the course of four decades.

Systemic versus Programmatic Retrenchment

The obvious starting point for such a discussion is Paul Pierson's seminal book Dismantling the Welfare State – certainly the focal point of the retrenchment debate – in which Pierson builds on Titmuss’ distinction of “institutional” and “residual” welfare states, viewing retrenchment as (ideologically motivated) residualization of the welfare state (Pierson 1994: 14-15).