18 Methodology of Empirical Econometric Modeling
17001750180018501900 1950 20002,5005,0007,50010,00012,500
Numbers of bankruptcies17001750180018501900 1950 200010,00020,00030,00040,000 Numbers of patents17001750180018501900 1950 20001020304050 Industrial output per capita17001750180018501900 1950 2000510152025 Real equity pricesFigure 1.1 Historical time series for the UK
1.4.1.2 Long-run change
“I see you’re admiring my little box,” the Knight said in a friendly tone.
“It’s my own invention – to keep clothes and sandwiches in. You see I
carry it upside-down, so that the rain can’t get in.”
“But the things can get out,” Alice gently remarked. “Do you know the
lid’s open?” (Lewis Carroll, 1899)Figure 1.1 records some historical time series for the UK over the period from about
1700 to 1991 (the dates differ for the various variables). Many other variables man-
ifesting dramatic non-stationarities are shown in Hendry (2001a, 2001b, 2005) and
Clements and Hendry (2001), where the first and last examine UK industrial output
in more detail. Here we focus on numbers of bankruptcies and patents, industrial
output per capita, and real equity prices (deflated by a cost of living price index)
(see Feinstein, 1972; Mitchell, 1988; Crafts and Mills, 1994,inter alia).
These four variables were selected from a range of alternatives as being related
to advances in technology and medicine, their implementation, incentives for
progress through intellectual property, and one source of financing (see Siegel and
Wright, 2007, for a recent review and bibliographic perspective). Technological
change is sometimes modeled as an “exogenous” random walk. While that is an
improvement over a deterministic trend, it is hardly a convincing representation
of a process which requires substantial inputs of human and physical capital, as
highlighted by endogenous growth models (see, e.g., Crafts, 1997). At the very