I. WHAT IS CHANGING? It appears that any overt expression of interest in methodology considered a clear sign of weak-mindedness or premature senility. --L. Boland, Methodology entry in the New Palgrave Dictionary of Economics As time passes, one should expect the techniques, and consequently the underlying methodology also, to evolve. This article will discuss some of the major changes in methodology since the middle of the last century to now. Among the major driving forces for this change the growth in the amount of data available in most parts of economics together with the tremendous growth in the speed and storage ability of computers, as well as the interactive aspects of the Internet. Techniques or methodologies that were unthinkable or certainly unavailable 25 years ago are now standard and common. A new mind-set required. To limit dissension I will consider economics to be a decision science, describing the decisions made by economic agents such as consumers, employers, investors, and policy makers. It follows that the purpose of any economic analysis will be to help actual decision makers make better decisions, on average, in a stochastic environment. Such an analysis can be thought of as a good; some analysis will be an intermediate good, which helps another economist produce a better final good. By taking these attitudes, a pragmatic viewpoint can be taken in many situations and provide solutions to some questions, as will be seen. The many books and articles on economic methodology and philosophy in the past century consider many topics that were considered of considerable interest then but have since lost their importance. For example, a topic earlier viewed as being of some importance was is economics a science?, although I believe that it no longer a hot topic around economics departments. The answer to the question that no one really cares! However, that may not be the correct response because if economics was classified as a science, the discipline may be eligible for much larger research grants. Another major topic with a lasting impact was about whether a model needs an input of economic theory or not; or equally, can a satisfactory model be achieved just by introspection without any use of data? I can recommend a very interesting short article on this topic titled Methodenstreit (or battle of the models) by Daniel Fusfield in the New Palgrave (1987). The battle was between Carl Menger and Gustav Schmoller, starting in the 1880s, with Menger in favor of pure theory and Schmoller wanting to use just an empirical approach. Unfortunately their initial debate became confused in a more political one with Menger pushing laissez-faire policies, whereas Schmoller supported interventionist government policy approaches--essentially right versus left in today's terminology. Fusfield states that on the basic issue of the place of theory and empirical studies in economics, Menger and Schmoller agreed that both were necessary. They disagreed on the emphasis to be placed on each and their role in the development of conclusions. Menger correctly pointed out that a model based just on data limited in generality, because the data cover only some of the possible space; it needs (correct) theory to go beyond the sample space. He also said that data acted as a bridge between the principles of pure economics and the policy problems of applied economics; in fact, he warned against policy analysis without an empirical study. Schmoller was less generous to theorists, finding the work too abstract, irrelevant, and based on assumptions that are unrealistic. A later, similar discussion took place between Koopmans (1947) and Burns and Mitchell (1947) that influenced modeling into the 1960s and 1970s. Where do we stand 115 years later in this discussion? In many ways it seems quite modern, certainly compared to some of the remarks made around real business cycle theory a few years back. …