A knowledge-based economy: new directions of macromodelling.(Report)

Abstract The development of new economies, leading to economies mostly based on knowledge, implies the construction of new, long-term macroeconometric models. They should incorporate the impacts of new technologies being endogenized, as well as human capital. The paper discusses several issues related to extending the notion of a production function. It covers the measurement and explanation of total factor productivity (TFP), the role of domestic and foreign research and development expenditures, as well as educational expenditures. The discussion is extended to include proposals to construct new submodels explaining the sections of research, education and the Information and Communication Technology (ITC) industries.

Keywords Polish economy. Knowledge based economy. Macromodelling. Knowledge capital. R&D. Human capital

JEL O10.C32.C57.E22

Introduction

The theory of economic growth has enjoyed its renaissance for over ten years, particularly the theory of endogenous and sustained growth. Various aspects of the growth theory have been elaborated recently, in the Polish literature mainly by Tokarski (2001). The subject is still vivid and continues to attract theoretical research (Nahuis 2003).

Empirical investigations following these developments have their macroeconomic and microeconomic dimensions (for instance, the innovation process analysis). Regarding the macroeconomic dimension, we need to mention empirical economic growth models conceptualized by Welfe (2000) The implementation of the concept contributed to the construction of the long-term macroeconometric model W8D of

In some sense, research activities expanding along the outlined directions triggered responses to challenges that economic sciences have to face because of the contemporary market economy transforming into a "new economy" and prospectively into a knowledge-based economy. The knowledge-based economy concept was formulated to contrast the new system with the industrial economy prevailing in the last centuries. Even though economic systems existing in the past did take advantage of knowledge that determined technological progresses, the knowledge capital started to dominate as late as in the turn of the twentieth and twenty-first centuries. This applies to the functioning of economies (relevant examples are the automation of manufacturing processes, fast transmission of management information, recently via the Internet, etc.) and economic growth (endogenization of technological progress mainly due to the development of Information and Communication Technology (ICT).)

It is not easy to find a measure indicating, whether the level of knowledge capital absorbed by an economy is sufficient to view it as a knowledge-based system (Smith 2002) However, in the economic growth analyses, a working assumption can be taken that research and development (R&D) and educational expenditures growing faster than investment outlays on fixed capital signify that such a system of economic relationships predominates. A more demanding criterion would aim to verify, whether a threshold value representing an adequately high share (e.g. 50-60%) of the Total Factor Productivity (TFP) in GDP growth has been exceeded. At the same time, the importance of fixed capital and employment growth as factors of production would be largely reduced in the system. (1)

Despite Poland's unquestionable technological backwardness, the knowledge-based economy concept was reflected in a work initiated by Kuklinski (2001) and then in empirical research. First, empirical studies were undertaken by a team headed by Zienkowski (2003). However, extensive analyses provided in the studies did not go beyond the formulation of simple regressions between GDP growth and single growth factors in the Polish economy.

The authors did not attempt to treat them within the framework of multi-regression analysis and consider with reference to the system of linkages characterizing an entire economy, including a knowledge-based economy. This gap has to be filled.

The paper analyzes the use of macroeconometric models as a major analytical tool for the knowledge-based economies. Next, the key issues in the generation and use of knowledge are discussed, i.e. the TFP dynamics and the factors behind its growth - the R&D and educational educational expenditures, investment in human capital. The paper

Towards Macroeconometric Models for the Knowledge Based Economies

Mechanisms stimulating the growth of a knowledge-based economy can be quantified by means of adequately expanded macroeconometric models. The models should enable the generation of endogenous technological/organizational progresses and explicitly react to changes in knowledge capital, while allowing for relevant feedbacks. They should make it possible to run simulation exercises useful in composing long-term growth scenarios that take into account the dynamics of knowledge capital.

One proposal that we intend to put forward to attain the presented goal involves the macroeconometric long-term model W8D of the Polish economy (see Welfe Ed. 2004a). However, the model needs to be adequately extended, modified, and transformed into a new structure-the model of a knowledge-based economy. Let us add that the model is the only operational model of this class in Poland, and one of few analogous models in Europe. (2)

The extension of the long-term econometric model W8D of the Polish economy would mainly require:

(a) The introduction of an extensive account of the consequences of increased domestic and foreign expenditures on innovations to the model. These expenditures would enlarge knowledge capital embodied in fixed capital and boost the expansion effects of broadly understood human capital

(b) Incorporation of satellite submodels into the model, which describe the functioning of the sectors of research and education (including higher education), as well as the sector of high-tech industries (with ICT industry)

(c) Regional and sectoral disaggregation of the model, which implies application of the cross-section-time series data.

Problems in Measuring TFP Dynamics

The Total Factor Productivity Concept

In the economic growth process, the effects of broadly understood technological progress are typically described using the concept of total factor productivity - TFP (Florczak and Welfe 2000, and Welfe 2001)-that refers to the Solow residual concept. The use of this variable is quite common also in international analyses. The methodology is not perfect though, which is evidenced by the results of our most recent research (Welfe 2002; see also Cornwall and Cornwall 2002). Significant reservations can be raised against treating potential production generated by the

Doubts are provoked by methods used to calibrate the production function parameters that are applied to identify the TFP dynamics. Therefore, it becomes necessary to find some empirically attainable alternatives. It seems appropriate to extend the scope of research by reaching for sectors/branches (distinguishing sectors with high absorption of technical knowledge) and for regional dimension. Similar research can be undertaken in the case of macromodels with appropriate disaggregation (e.g. input-output models). Let us consider the indicated problems more in detail.

It is known that TFP is not directly observable. A general assumption is that its identification starts with so-called "Solow residual" (Solow 1957), i.e. a residual derived from a production function determined for the case, when only two production factors are used as explanatory variables: fixed capital and employment.

Extended Production Function and Solow Residual

Let us have a Cobb-Douglas production function with constant returns to scale:

[X.sub.t.sup.p] = B[A.sub.t][K.sub.t.sup.[alpha]][N.sub.t.sup.(1-[alpha])] [e.sub.n.sup.[[epsilon].sub.1]] (1)

where

[A.sub.t] TEP describing effects of technological progress

[K.sub.t] fixed capital (fixed prices)

[N.sub.t] employment

[X.sub.t.sup.p] potential output (fixed prices)

[[epsilon].sub.t] a disturbance term

TFP dynamics is usually defined by means of the residual's dynamics. It is found by deducting the rate of potential output growth obtained assuming that only the fixed capital rate [K.sub.t] and employment rate [N.sub.t] are taken into account - from the rate of growth of effective output [X.sub.t]:

[RES.sub.t] =[X.sub.t] - [[[alpha].sub.t] [K.sub.t] + (1 - [alpha]) [N.sub.t]], (2)

alternatively, switching to logarithms:

[DELTA]lnRE[S.sub.t] = [DELTA]ln[X.sub.t] - [[alpha][DELTA]ln[K.sub.t] + (1 - [alpha])[DELTA]ln[N.sub.t]],

where [X.sub.t] - effective output; RES - residual and (o) - the rate of growth.

However, this broad notion of TFP is not generally applied. Several authors define inputs of the primary production factors in terms of their quality to make them more homogenous. Consequently, fixed capital is decomposed into either generations or, more recently, into the ICT and non-ICT components (3). Specific quality

In the first case, the residual represents technology impacts except for changes in the quality of fixed capital (especially productivity of ICT equipment and labour quality), whereas in the second case - the impact of technical progress embodied in fixed capital only.

Assuming, however, that the quality indicators are separable, then their impacts can be combined with the narrowly defined TFP impact and so a return to the broad TFP notion described above becomes possible. For that reason, we will use this concept in our further analysis.

Output Indicators

This concept of TFP gives rise to many problems. The literature of the subject handles them in different ways (see Welfc 2002). Firstly, the choice of an output measure is an issue. Regarding the macro scale, a typical assumption is that production should be represented by value added (GDP in terms of an entire economy), which is equivalent to introducing only the primary production factors (fixed capital and employment) to the production function. For the micro scale and frequently also the mezzo scale (sectors, branches), the gross (or sold) output is used quite often, where it is more easily available and calculated more precisely. This approach, however, involves the modification of the production function l by adding intermediate inputs (energy and materials) as the explanatory variables, i.e. the KLEM production function is used. The last requirement cannot be easily met, because of the shortage of information about intermediate inputs.

Print Get the Mag Weekly Updates