Economic Impact of Smart Specialization and Research in Advanced Adaptive Systems in a Monetary Union

Abstract

Tensions in the eurozone have brought back to the surface issues/criteria related to its solidity and optimality. We will suggest that technology and research can help buttress a monetary union, through their impact on trade and specialization and production structures. This applies to advanced adaptive systems, such as robotic systems that can be used in wide arrays of sectors to enhance performance and competitiveness. It can open the way for European economies to compete internationally by emphasizing their strong card, technology, as opposed to competing on labor cost. It can mitigate the tensions caused by divergent performances between members of a monetary union, through promoting convergence toward higher technological plateaus. This emphasis on technoeconomic convergence has a counterpart, in terms of technology and innovation policies, which can hardly compete globally on the basis of labor costs. The sovereign debt crisis has increased pressure on countries to redress structural problems in their economies, while preserving the margin for public investment in knowledge-based capital, in a “smart way,” contributing to productivity growth and competitiveness. Smart specialization, the concept and the associated policy process, aims to do just that, fostering growth potential in a context of rapid technological change and globalization. The rationale for smart specialization stresses the role of policymakers, knowledge-based institutions, and entrepreneurs in shaping specialization and competitiveness. Horizontal key enabling technologies, such as robotic technologies, play a particularly essential role in boosting existing strengths, as well as revealing new economic opportunities in sectors at various levels of technological sophistication.

Annex: A Formal Model

We have a set of countries X = {x(l), x(2), x(3), … x(n)} and a set of industries I = {I(l), I(2), …. I(m)}. We assume perfect competition among identical firms within each industry, and we represent them by a single firm F(j), where j is an indicator for the industry to whom this firm belongs. The firm’s revenue equals payments to labor LB and capital KP and at complete specialization constitutes the income Y of the host country.

$$ Y = p\;D = {\text{wg}}\,{\text{LB}} + {\text{ rt}}\,{\text{KPd}}Y/{\text{d}}T = p\,{\text{d}}D/{\text{d}}T $$

(4.1)

where p is the exogenously given world price for the country’s product, wg is wage paid to labor, rt is rent to capital, D is demand for the product, and T is a measure of a demand shock. The income lost from an adverse shock is A = p dD. Since the interest rate i and the exchange rate are in effect set by the European Central Bank (ECB), the country can borrow the sum A only if expected income E[Y] = E[p] E[D] > A(l + i) over the horizon of the loan. Furthermore, if the expected return in the country’s industry E[r] > i, then there will be capital inflow. What interest rate should the home country be aiming for during monetary policy deliberations? Assume K flows into the country, of which A is consumed to redress the lost income effect of the adverse shock and K−A is invested in the home country’s industry. We must have

$$ \left( {K - A) \, \left( { 1+ E[r]} \right.} \right) > K({\text{l}} + i) \Rightarrow i < E\left[ r \right]\left( { 1- A/K} \right) - \left( {A/K} \right) = i^{\prime} $$

(4.2)

Hence, the home country will strive for an i less than or equal to i′ and the tension in the ECB, and the concomitant loss to utility in the particular country and the EU as a whole will increase with the distance between i′ and the interest rate i* eventually chosen by the ECB.

In the intra-industry case, many industries are represented in each country and the country’s income equals the sum over all j’s of p(j)D(j), where j is an industry indicator and D(j) indicates demand for that particular country’s variety of products of industry j (as before all revenue is paid to factors of production). For each country, Y(j) is income from industry j and Y is the sum total income for the country from all industries. A negative demand shock T will lead to loss of income dY(j) = p dD(j) with the elasticity of demand to that shock assumed to be the same in all countries. The pressure from each country will increase, and the desired i′ will decrease the share of that industry’s revenue with its country’s income increases. As before, countries can borrow the lost income A if the dot product of the vector of expected future prices times the vector of expected future demands exceeds the dot product of the present prices times present demands by at least the amount borrowed plus interest, plus the current deficit in consumption.

Depending on shares of Y(j) in a country’s total Y, the i′ a country will pursue will vary from country to country. When it comes to deciding on monetary policy, countries will be faced with a trade-off: A low i* will please the countries hit the hardest by the shock and will disappoint those who aimed for a higher i* because the lower i in their cases would bring about inflationary pressures.

In the case of an emphasis on inter-industry trade and concomitant specializations, one country (the one extensively hit by the shock) promotes a low interest rate and all the others prefer a high one. If the high one is adopted, the loss for the country hit by the shock may be devastating and may even lead to withdrawal from the monetary union. If the low one is adopted, then all the other countries will suffer increased inflation.

In the case of intra-industry-trade emphasis, and concomitant specializations, the juxtaposition is not so extreme––some countries’ target i’ will be very close to the i* chosen and others countries’ i′ will be more distant, on either side. It seems easier in the intra-industry case to achieve near-satisfaction of more countries without driving any single one to desperation.

In order to examine these claims, we will propose an elementary loss function L which measures loss due to high i* for countries which aimed for a lower i as well as loss due to expected inflation for countries aiming for a higher interest rate than the i* eventually adopted. The loss for the former set of countries increases with the share w of the industry hit by the shock in their income. The loss for the latter set of countries is mitigated by that share w in the former country’s income because a high w indicates a high risk for dissolution of the community, seen as a loss by all, including the inflation-fearing countries.

More concretely, one can suggest that bargains are struck which reimburse inflation-fearing countries for their accepting a lower i*––the higher the w is, the higher the reimbursement the inflation-fearing partner can extract. We could add terms for other kinds of losses, such as inefficiency due to lack of specialization––it would not change the flavor of the analysis.

The issue at hand involves promotion, or not, of inter-industry-trade specialization within the new context of economic and monetary union. Are there strong new reasons, reasons which are emerging in the context of economic and monetary integration and were inoperative before? We will limit the size of the group of countries as well as the number of industries to two, in order to simplify the analysis. Country 1 aimed for a lower interest rate i′(l) than the i* eventually decided, whereas country 2 aimed for a higher one, i′(2). The constants h and a are positive exponential coefficients of utility loss, and the exponents are overall negative because the bases are between 0 and 1. The exponent a is weighed by (l − w), 0 < w < l, whose mitigating effect increases with w, for the reasons mentioned above. Each country can specialize in one of the two industries, or opt for producing and trading in both (given the standard assumptions about similar country size, tastes, technology).

Our loss function L is

$$ L\left( w \right) = \left( {i^{ * } - i^{\prime}(l)} \right){ \exp }\left( {- {\text{h}}w} \right) + \left( {i^{\prime}(2} \right) - i^{ * } ){ \exp }\left( { - a\left( {{\text{l}} - w} \right)} \right) $$

(4.3)

We will examine how L varies with w. If L is minimized for an extreme value of w (i.e., 0 or 1, or a value <0 or >1), this would favor complete specialization and inter-industry trade. If a value of w between 0 and 1 minimizes L, this would favor incomplete specialization and intra-industry trade. The analysis is symmetric in the sense that country 2 can play tomorrow the role that country 1 is playing today, if an adverse shock hits country 2.

For an extremum, dL/dw = 0 =>

$$ - h \, y - {\text{h}}w{ \ln }y + {\text{ az}} - a\left( {{\textit{l}} - w} \right){ \ln }z = 0\;{\text{where}}\;y = i^{ * } - i^{\prime}(l)\;{\text{and}}\;z = i^{\prime}\left( 2 \right) - i^{ * } $$

$$ \Rightarrow {\text{aln}}z/{\text{hln}}y = y - {\text{h}}wz - {\textit{a}}\left( {{\textit{l}} - w} \right) $$

(4.4)

$$ {\text{set}}\left( {{\text{aln}}z/{\text{hln}}y} \right) = s $$

(4.5)

$\Rightarrow { \ln }s = - {\text{h}}w{ \ln }y + {\text{a}}\left( {{\textit{l}} - w} \right){ \ln }z$ after taking natural logarithms on both sides

$$ \Rightarrow { \ln }s = - w + s\left( {{\text{l}} - w} \right)\;{\text{after}}\;{\text{dividing}}\;{\text{through}}\;{\text{by}}\;{\text{hln}}y $$

$$ {\text{ }} \Rightarrow w{\text{ }} = \left( {{\text{ }}\!s{\text{ }} - {\text{ }}\left( {{\text{ln}}s/{\text{hln}}y} \right)} \right)/\left( {{\text{l}} + s} \right) $$

(4.6)

Let us examine this ratio. We note that 0 < y, z < l => lny < 0 and lnz < 0. Furthermore, since both lnz and lny are negative and both h and a are positive, we have s > 0 => the denominator l + s > 1.

If s < 1 then lns < 0 and lns/hlny > 0, and hence, the numerator is less than 1 since we are subtracting a positive number from s, which we have assumed is less than l. Therefore, if s < 1 the numerator < denominator => w < l.

If the numerator is >0, then 0 < w < 1, and intra-industry trade is favored. Is the numerator greater than 0? Yes, if alnz/hlny > lns/hlny <=> alnz < ln(alnz/hlny) <=> h < alnz/(zalny). Since lnz and lny would usually not be substantially different and since 0 < z < 1 => za must be very small, this condition should often be fulfilled. From s < l, we have that h > alnz/lny => for a/za > h > (alnz/lny) ~ a we have the optimal w between 0 and 1, and intra-industry trade is promoted.

The conditions above have the following significance: s < l means that the loss for #2 is less than the loss for #1 before weighing losses with the exponents which are functions of w. In this case, w is less than 1 because a higher value of w (i.e., w = l) would increase exponentially the already high loss term for #l and would not offer much in terms of reducing the loss for #2, which is already low. The condition h < a/za, 0 < z < l, indicates that the larger a is––namely, the larger the exponent of the loss for country 2 is––the wider the range in which h satisfies the inequality.

The above applies when s is assumed less than 1. Note that if s = l, then w = ½––the second-order condition for a minimum is satisfied since L″(w) > 0

Now, let us see what happens in case s > 1:

$$ {\text{if}}\;s > {\text{l}}\;{\text{then}}\;{ \ln }s/{\text{hln}}y < 0 \Rightarrow {\text{in}}\; w = \left( {s - \left( {{ \ln }s/{\text{hln}}y} \right)} \right)/\left( {{\text{l}} + s} \right)\;{\text{the}}\;{\text{numerator}} > 0 \Rightarrow w > 0 $$

$$ {\text{For}}\;w < {\text{l}} \Leftrightarrow s - { \ln }s/{\text{hln}}y < {\text{l}} + s \Leftrightarrow - { \ln }s/{\text{hln}}y < 1\Leftrightarrow - { \ln }s/{ \ln }y < {\text{h}} $$

${\text{From}}\;s > {\text{l}} \Rightarrow {\text{h}} < {\text{aln}}z/{ \ln }y,\;{\text{hence}}\; - { \ln }s/{ \ln }y < {\text{h}} < \left( {{\text{aln}}z/{ \ln }y} \right) \sim {\text{a}}$.

The conditions above have the following significance: s > l means that the loss for #l is less than the loss for #2, before taking the exponent into account.

w is greater than 0 because a smaller w (i.e., w = 0) would increase #2’s already high loss exponentially, whereas it would have a smaller beneficial effect in terms of reducing #l’s small loss. The condition −lns/lny < h < (alnz/lny) ~ a indicates that the range of h which satisfies the condition is reduced as a grows larger––which means that in order for incomplete specialization to be optimal, it should not be the case that a ≫ h, namely the loss for #2 should not be much greater than the loss for #1.

In other words, when the loss for #1 is greater than the loss for #2, it should not be so overwhelming that any w > 0 would lead to exorbitant losses for #1 that #2’s benefit from a higher w could never outweigh it (and similarly for loss of #2 > loss of #1 and w < 1).