Monday, August 22, 2011

Understanding the context for innovation policy

Across many years of research on science, technology and innovation policy and indicators, I have often felt that that there is a lack of holistic approach to mapping policy and metrics. I have addressed this in various bits of writing including the conference paper written with Susan Cozzens a few years ago called ‘knowledge ecologies’. I haven’t developed it for publication because I am not sure that I yet have the language to express the ideas embedded in that paper.
Most of the work on policy appears to assume particular ecomic contexts and studies do not directly link policy mixes with the economy.

As I have written elsewhere in this blog, the idea for my econscape graphs came to me in the early 1990s but they are somewhat problematic entities. They are brilliant for communicating to a certain audience that every country has its own economic structure fingerprint. This would make a lovely colour picture book someday but the graphs are difficult in and of themselves to do much with because:

1.       The charts themselves do not help to address a particular research questions for academic publication – that is better suited to the standard I-O economic analysis methods;

2.       They are better suited for policy development – but the question is how to use them.

Then about a year ago or so an idea occurred to me (in Church funnily enough) on how to push this further.

Section 1: Explaining the Matrix

In the paper I wrote with colleagues from the OECD (Yamano and Webb) we included the following diagram. It is a standard input-output framework diagram.
The Basic Structure of an Input-Output Table

Wixted 2006.

If we focus just on the domestic intermediate matrix, then we could divide the economy into 4 main components which turns into 16 (4X4) interacting blocks of activity.
1.       Resource base activities (mining and agriculture)
2.       Manufacturing
3.       Services
4.       Government and non-profit sector

Technical note discussion – please read.

In the intermediate matrix above the interaction of the first [demand] column (agriculture) and the first [supply] row agriculture appears in the top left. If you turn this into a graph visually much detail is lost because the eye is looking down the most important section of data – the diagonal as the largest interactions have always been intra-industry.

The econscape images are therefore rotated 90o so that the interaction just noted appears in the bottom left.

If the matrix is going to be used in conjunction with the econscape image then I recommend rotating the matrix as well, otherwise it is probably best to align it with the traditional I-O configuration.

As a first cut of a policy and metrics we can cut up the domestic intermediate matrix into 16 significant quadrants.

Stage 1:                Filling in the matrix

Stage 2: Name the matrix sectors

Stage 3: Overlay the matrix on an econscape (this time Australia)
This can be done for any economy where a symmetric domestic industry X industry transactions matrix is available. Develop the econscape, this time for Australia.

Then overlay the matrix – suitably rotated for the orientation chosen for the econscape.

From this we can observe the Australian economy is heavily concentrated in the services and resource based segments of the economy

Section 2: Implications for innovation metrics

Typically STI indicators and policy have emphasised the segment F (mfg * mfg). However, in this example we can see it is important that indicators be concentrated in the resource and services segments.  But this is not straightforward.

Section 3: Implications for innovation policy

In recent decades there has been increasingly emphasis on innovation in the services sector but resource sector based innovation is a real problem for many economies, policy makers and academics. For too long it has been ignorned as old economy and not ‘knowledge based’, but this is not the case Australia and Canada. We need to start looking at economies in all their complexities.

Section 4: Cautions
The one caution that needs to be made regarding this approach is that despite the fact that the data is based on industry interdependencies the data can not convey where the economy’s true value lies.

Resource sectors for example earn vast revenues from exports typically which then re-enters the economy in the service sector. This is what I refer to worm-holes in the economy. The data doesn’t necessarily highlight connections but case studies do make them clear.

 © Brian Wixted.