Degree Name

Doctor of Philosophy


Department of Economics


This thesis provides an exploratory analysis as to whether input-output analysis can be extended to measure information spillovers. A modified input-output framework is developed which links regional context, global markets, an information sector and innovation indicator matrices. Regional context is described as the cultural and economic qualities inherent in a regional economy and the information sector is derived in a manner described by Karunaratne (1988). Innovation indicator matrices are developed as an extension of the DeBresson (1991) innovative interactive matrices. They have added value by calculating the export related innovation spillover share for each export performance influencing variable, export strategy or innovation information source (defined by the TVH typology) at both the sector and firm level on a comparative and consistent basis. The innovation indicator matrices are used to investigate whether the input-output approach will provide additional insights into issues raised in the regional development literature regarding the measurement and importance of information spillovers to small, regional exporting firms. A level of detail and analysis not possible using alternative more aggregated econometric methods is achieved.

The method derives information innovation multipliers using National and State Input-Output tables constructed from ABS data, modified with the TVH typology data sourced from a survey of exporting strategies used by NSW regional firms. Innovation indicator matrices consist of two types as follows: those used for calculations which contain the TVH strategy use frequencies, and the storage matrices which contain their corresponding spillover values (the results of these calculations). Strategy spillover values are derived by proportioning the export portion of the innovation information value-added industry support multiplier across the strategy use frequencies. Specifically, the indicator matrices are used (1) to measure the spillover portion that adds value to the developments or improvements in export related products or production processes, (2) determining from where a how these spillovers are sourced or generated according to the TVH typology variables, (3) determining the TVH typology characteristics of the leading export enhancing sectors, multiplier groups, regions and firms, and (4) determining suitable export enhancing policies by modelling various network generated or government assisted information flows.

This unique method found that:

  • innovation information was flowing away from export production towards non-exports production (sector-regional dimensions are provided;
  • strategy set components were more significant at the network and sector than regional levels;

The unique method identified:

  • significant variations in strategy and spillover use for networks across the regions;
  • significant MAR and Jacobs spillover flows at the sector and regional levels;
  • that supply and demand networks were independent of strategy use and spillover values at the region and sector levels;
  • that firm‟s choice of region and information spillover sources were independent of TVH strategy choice;
  • cooperative networking arrangements using overseas agencies and partnerships and electronic communications such as e-mail and internet, self sourced information on new technologies, attending business meetings and capital city conferences, offering after sales service, and obtaining external market development information using visits to or from service providers, as the main strategies and sources used by leading exporting firms (also significant for the networks, regions and sectors;
  • successful exporting policies based on the spatial use of leading strategies.

Innovation indicator matrices successfully measured information spillovers. Survey amendments are suggested to obtain the TVH data for more than one time period thus allowing strategy changes to be assessed and some weighting of strategy relative importance would be useful for improving spillover measure accuracy. Data disaggregation in the IO approach illustrated the coexistence of different innovation and development paths among firms and regions and broad networks across many regions. This revealed the simultaneous influence of many theories, the absence of one dominant approach, and the need to utilize existing spatial networks in policy design. Policies generated to encourage the use of successful strategies delivered results consistent with the innovative milieu theory in that assistance to promote related scientific research delivered the best export, associated employment outcomes and related flow-on effects.



Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.