%Paper: ewp-dev/9707001
%From: wkeller@ssc.wisc.edu
%Date: Thu, 3 Jul 97 15:25:05 CDT

%% This document created by Scientific Word (R) Version 2.0


\documentstyle[12pt]{article}
%%%%%%%%%%%%%%%%%%%%%%%%%%%%
%TCIDATA{TCIstyle=Article/art4.lat,jart,sw20jart}

%\input tcilatex
\QQQ{Language}{
American English
}

\begin{document}

\title{Trade and the Transmission of Technology\thanks{$\,\,$An earlier version of
this paper circulated under the title: ''International R\&D Spillover and
Intersectoral Trade Flows: Do they Match?''. Thanks are due to seminar
audiences at Colorado-Boulder, the 1995 NBER Summer Institute, Purdue,
Wisconsin-Madison, and Yale, especially Robert Evenson, Zvi Griliches, James
Harrigan, Keith Maskus, F.M. Scherer, T.N. Srinivasan, Marie Thursby, and an
anonymous referee. Thanks also to Christian Langer for providing the data on
trade flows, to Michael Craw for the US import input-output data, and to
Jonathan Putnam for his help in obtaining the technology flow matrix used in
the paper.}}
\author{Wolfgang Keller \\
%EndAName
University of Wisconsin and NBER\thanks{$\,\,\,\,$Department of Economics,
1180 Observatory Drive, Madison, WI 53706. Email: wkeller@ssc.wisc.edu.}}
\date{First Version: September 1996\\
This Version: July 1997}
\maketitle

\newpage\setcounter{page}{1}

\section{Introduction}

The question of whether goods trade contributes to the transmission of new
technologies is both old and new. It is an old question in the context of
closed-economy studies that have examined whether investments in research
and development (R\&D) in one industry affect productivity in other
industries, where these other industries are often identified through
input-output relations. In these studies, trade is viewed as a channel of
transmission for newly created technologies (see Griliches 1984, and, more
recently, Nadiri 1993). In the open-economy context, the idea that trade
might be contributing to the international transmission of technology has
been emphasized only in the recent growth theory literature (Rivera-Batiz
and Romer 1991, Grossman and Helpman 1991).

For a long time, there was no consensus on the question of why, on average,
outward-oriented economies tend to grow more rapidly (see, e.g., World Bank
1987, Rodrik 1995). The crux of the debate centered around two important
developments in the literature. First, models built within the framework of
static trade with competitive markets predicted gains from trade which were
very small, in comparison to real-world differences in productivity growth
between the average open country and the average protectionist country. Once
models capable of generating positive productivity growth rates even in the
long-run (e.g., Romer 1990), however, were placed into an open-economy
context, changes in the trade regime could have both long-run and large
growth effects, and this possibility seemed to allow for a reconciliation
between theory and empirics.

Secondly, there was the dissension of views on the mechanism by which trade
contributes to productivity growth. According to one view, trade affects a
country's growth rate through its effects on domestic resource allocation.
Equating the growth effects of trade to those which are generated by changes
in the domestic resource allocation, however, met with the troubling
implication that a purely domestic tax-and-subsidy policy can achieve
exactly what trade achieves. An alternative view, arising from the growth
literature, suggested that international trade directly affects productivity
growth because trade is a mechanism through which technological knowledge is
transmitted internationally. In the models of growth in an open-economy,
R\&D investments create new technology in the form of construction designs
for new intermediate products (Romer 1990; see also Ethier 1982). These
products, traded internationally, benefit the recipient country that employs
the new intermediate goods because that country need not first invent the
construction designs. Importing a foreign intermediate good, therefore,
allows a country to capture the R\&D-, or 'technology-content' of the good.
For a given primary resources, productivity is increasing in the range of
different intermediate goods which are employed, due to the assumption that
they are imperfect substitutes for each other. The model predicts that total
factor productivity (TFP) is positively affected by the country's own R\&D,
as well as by R\&D investments made by trade partners.

This framework is clearly also relevant to the process of technology
transmission between different sectors within one economy. In this sense,
the open-economy model could well serve as a theoretical underpinning of the
empirical analyses in the closed-economy studies referred to above. This
paper provides a unified approach of analyzing the importance of trade for
the transmission of technology, both internationally as well as domestically.

Of course, technology need not flow exclusively from the seller to the
buyer. Mansfield (1984), e.g., points to evidence that technology is also
transferred from the downstream to the upstream industry. Technology flows
can also be disproportional to goods trade, as in the case of
reverse-engineering, or not at all related to goods trade, as in the case of
attending a conference where state-of-the-art technology is being
demonstrated. Hence, while the model where all technology diffusion is based
on market transactions is a useful benchmark, the answer to the question of
how technology is transmitted from one sector to another is more complex
than simply given by the input-output structure of an economy. This point
has been recognized in both the trade and growth as well as the R\&D-TFP
growth literature,\footnote{%
See Grossman and Helpman (1991), Ch.6.5, and Scherer (1984), respectively.\ }
and the analysis below accounts for that.

The importance of input-output relations in the transmission of technology
in a domestic context has been emphasized by Terleckyj (1974) and, among the
more recent papers, by Wolff and Nadiri (1993).\footnote{%
See also Griliches (1979), in particular for a discussion of the older
literature in the closed-economy context, and the papers in Griliches (1984).%
} The analysis presented below builds on these contributions. Another strand
of literature, including Scherer (1984), Jaffe (1986), Evenson et al.
(1991), and Branstetter (1996), goes beyond modeling technology diffusion
based on input-output market-transactions. In the contributions by Jaffe
(1986) and Branstetter (1996), the authors identify the direction of
technology flows between firms by differences in their patent portfolios:
the more similar these are, the more likely it is that R\&D in one firm
affects the productivity of another. Scherer (1984) and Evenson et al.
(1991) have modeled the structure of inter-industry technology diffusion by
creating technology flow matrices which predict the R\&D originating from as
well as the R\&D used by a given sector. Below we will employ a technology
flow matrix based on the work of Evenson et al. (1991).

Coe and Helpman (1995), Coe et al. (1995), Englander et al. (1995), and Park
(1995) have recently presented evidence on how trade contributes to the
international diffusion of R\&D spillovers.\footnote{%
See also Eaton and Kortum (1996).} Although motivated by recent theories of
trade and long-run growth, their empirical analysis is very much in the
spirit of the papers by Terleckyj, Griliches, and Scherer. Coe and Helpman
use country-level data to show that countries' productivity levels are
positively affected by domestic as well as bilateral import-share weighted
foreign R\&D stocks; hence, the import shares take the place of input
coefficients in the domestic TFP-R\&D literature.

The present paper departs from earlier work in several respects: First, we
use industry-level data on international transactions, as opposed to the
country-level data employed by Coe and Helpman (1995), Coe et al. (1995),
and Park (1995). It has been argued that studies employing aggregate data
are likely to miss much of the technology flows, because the diversity of
sectoral characteristics as well as the very nature of technology diffusion
tends to confound any inferences which can be made from that data (e.g.,
Branstetter 1996). The use of two- or three-digit industry level data should
reduce this problem. Secondly, this paper integrates the recent emphasis on
the open-economy relations with earlier work, in particular by Terleckyj
(1974), Scherer (1984), and Evenson et al. (1991), which stressed the
dimension of domestic intersectoral technology transmission. The advantage
of this approach is that all transactions predicted by the theoretical
model, both domestic as well as international are considered, in contrast to
earlier studies which have focused on only a subset of those.

Third, in the model of technology diffusion presented below, we can derive
the specific form of the model's prediction on how outside R\&D should
matter for productivity. Therefore, we are able to test whether new
technology created in the industry of a particular country has the same
productivity effects as new technology from other domestic sectors, from
abroad, or from other sectors in other countries. This should improve our
understanding of the way and extent to which trade is related to the
transmission of technology.

The remainder of the paper is as follows. In the next section, we describe
the R\&D-driven growth and intermediate inputs model. Section 3 gives an
account of the data which will be used. Section 4 gives the estimation
results, and section 5 concludes.

\section{R\&D-Driven Growth and Intermediate Inputs Trade}

We consider a typical model of the Ethier-Romer, Grossman and Helpman
variety, in which long-run growth is endogenously driven by R\&D
investments, and technology is being transmitted via trade in intermediate
inputs.

\subsection{Domestic Intersectoral Trade}

Assume that good $z_j$ in sector $j,\,j=1,...,J,$ is produced according to 
\begin{equation}
z_j=A_j\,l_j^\alpha d_j^{1-\alpha },  \label{outp}
\end{equation}
where $A_j$ is a constant, $l_j$ are labor services used in final output
production, and $d_j$ is a composite input consisting of horizontally
differentiated goods $x$ of variety $s.$ Specifically, $d_j$ is given by 
\begin{equation}
d_j=\left( \int_0^{n_j^{de}}x_j(s)^{1-\alpha }\,ds\right) ^{\frac 1{1-\alpha
}}.  \label{int}
\end{equation}
The variable $n_j^{de}$ denotes the range of intermediate inputs which are
employed in this sector (ignoring integer constraints). Here, the
superscript $d$ stands for domestic, and $e$ stands for employed. We
distinguish $n_j^{de}$ from $n_j^p,$ the range of intermediate inputs
produced in sector $j$; the latter is increased by entrepreneurs devoting
resources to R\&D. Denote with $\phi _j^p$ the flow of R\&D expenditures in
sector $j$. Blueprints of new inputs are simply created according to $\dot
n_j^p=\,\phi _j^p.$ These resources could be in form of labor services which
have an alternative use in the output sector.\footnote{%
This presentation does not account for those; see, e.g., Romer (1990) for a
complete description.} If designs never become obsolete, the stock of
intermediate inputs produced in sector $j$ at time $T$ is equal to $%
n_j^p(T)=\int_{-\infty }^T\,\dot n_j^p(t)dt=\int_{-\infty }^T\phi
_j^p(t)\,dt $, that is, equal to the cumulative R\&D resources at time $T$;
we define $n_j^p(T)\equiv b_j(T).$

We assume that to produce one unit of any intermediate good requires one
unit of sectoral output. Then, if capital $k_j$ is defined as cumulative
foregone sectoral output, this will be equal to $k_j=\int_0^{n_j^p}\,x_j(s)%
\,ds.$ In a symmetric equilibrium, all intermediates $x$ are produced at the
same level, so that $k_j=n_j^px_j.$ Rearranging for $x$, and substituting
into (\ref{int}) leads to the following expression for output\footnote{%
Here, $A_j^{^{\prime }}=A_j\left( n_j^{de}/n_j^p\right) ^{1-\alpha }.$ In
the following, the term $\left( n_j^{de}/n_j^p\right) ^{1-\alpha }$ will be
ignored, expecting that this will not crucially affect the estimation below.}
\[
z_j=A_j^{^{\prime }}\,\left( n_j^{de}\right) ^\alpha l_j^\alpha
\,\,k_j^{1-\alpha }, 
\]
Defining an index of TFP, $f_j^{*},$ as $f_j^{*}=\frac{z_j}{l_j^\alpha
\,k_j^{1-\alpha }},$ and taking logs results in 
\begin{equation}
\log f_j^{*}=\log A_j^{^{\prime }}+\alpha \log n_j^{de}.  \label{lnfs}
\end{equation}
Note that in equation (\ref{lnfs}), $f^{*}$ is positively related not to the
range of intermediates which have been invented in sector $j$ ($n_j^p$), but
to those which are employed there ($n_j^{de}$). We model the range of
intermediates employed as the weighted sum of the ranges of intermediates of
all sectors, where the weights are given by the input-output relations of
the sectors\footnote{%
See Terleckyj (1974); the origin of the idea to model R\&D spillovers as the
weighted sum of other sector's R\&D is Griliches (1979).} 
\[
n_j^{de}=n_j^p+\sum_{v\neq j}^J\omega _{jv}n_v^p,\,\,\forall j. 
\]
Let ${\bf \Omega }$ be the matrix of observed input-output coefficients,
with a typical element $\omega _{jv}$%
\begin{equation}
{\bf \Omega }=\left[ 
\begin{array}{cccc}
0 & \omega _{jv} & \omega _{jw} & \cdots \\ 
\omega _{vj} & 0 & \vdots & \vdots \\ 
\vdots & \vdots & \ddots & \vdots \\ 
\cdots & \cdots & \cdots & \ddots
\end{array}
\right]  \label{sig}
\end{equation}
In terms of observables, this means that the effective domestic R\&D stock
which affects TFP in sector $j$ is 
\[
b_j^{de}=b_j+{\bf \Omega }_j\,{\bf b}_v,\text{ }v\neq j\text{, }\forall j. 
\]
Here, ${\bf \Omega }_j,$ of dimension ($1\times J$), is the $j$th row of $%
{\bf \Omega ,}$ and ${\bf b}_v$ is of dimension ($J\times 1$). Hence, $%
b_j^{de}$ is an input-output weighted sum of the cumulative R\&D stocks of
all sectors $v$. For a given sector $j$, denote the R\&D effect from other
sectors by $b_j^{io}$. With this notation, we have that 
\begin{equation}
\log f_j^{*}=\log A_j^{^{\prime }}+\alpha \log b_j^{de}=\log A_j^{^{\prime
}}+\alpha \log \left( b_j+b_j^{io}\right) ,\forall j.  \label{lnf2}
\end{equation}

\subsection{International Intra-Industry and Inter-Industry Trade}

When there is more than one country, output producers in country $i$'s
sector $j$ can employ intermediates from other countries $h$ in addition to
those from other domestic sectors, $v\neq j$.\footnote{%
This will happen in equilibrium. Analytic results for the symmetric
two-country case are derived, e.g., in Keller (1996).} The effects of
intermediate goods from the same sectors in other countries and from other
sectors in other countries can be treated analogously. First, we consider
international intra-industry trade. Coe and Helpman (1995) have recently
proposed to model the effect of foreign R\&D on domestic productivity by
utilizing bilateral import shares as weights. These are conceptually
identical to the input-output relations capturing domestic trade
transactions. Let $m_{ihj}$ be the bilateral import share of country $i$
from country $h$ for industry $j$. For a given country and sector $ij,$ the
effect from foreign intermediates produced in the same industry, $b_{ij}^f,$
is then\footnote{%
So far, we have considered the contemporaneous relation between R\&D stocks
and TFP. It is clear, however, that R\&D outlays must precede productivity
effects because of the presence of various lags in the invention and
commercialization process, see, e.g., Pakes and Schankerman (1984). These
authors argue that the mean lag should be in the range of 1.2 to 2.5 years,
p.84. Of particular interest is here whether foreign R\&D influences
domestic productivity with the same mean lag as domestic R\&D, or whether
this lag is perhaps longer. While data availability precludes estimation of
the precise lag structures for domestic and foreign R\&D, we present some
results in Table A.7 of the appendix for the case where the relation between
TFP and domestic R\&D\ is contemporaneous, whereas with foreign R\&D, there
is a one-year lag. These results are discussed below.} 
\begin{equation}
b_{ij}^f=\sum_{h\neq i}^Im_{ihj}\,b_{hj}\,,\forall ij.  \label{bf}
\end{equation}
The domestic sector $j$ can also employ intermediate inputs from foreign
sectors $v\neq j,$ and therefore benefit from technology created there. The
matrix which captures those market relations is the input-output matrix for
imports. Let $\gamma _{ijv}$ denote the share of country $i$'s imports of
the $j$ intermediate which go to the $v$ industry, where $i=1,...,I$, and $%
\,j,v=1,...,J.$ Then, define $b_{ij}^{f,io}$as the R\&D effect from foreign
intermediates in industries other than $j,$%
\begin{equation}
b_{ij}^{f,io}=\sum_{v\neq j}^J\gamma _{ijv}b_{iv}^f,\forall i,j.
\label{bfio}
\end{equation}
If we allow the productivity effects through intermediate inputs from
different sources to differ, then a TFP index analogous to (\ref{lnf2}) can
be written as 
\begin{equation}
\log f_{ij}^{^{\prime }}=\mu _{ij}+\beta _1\log \left( b_{ij}+\beta
_2b_{ij}^{io}+\beta _3b_{ij}^f+\beta _4b_{ij}^{f,io}\right) \,+\varepsilon
_{ij},\forall i,j,  \label{full}
\end{equation}
where the error term $\varepsilon _{ij}$ is assumed to be mean zero and
independently and normally distributed, capturing all influences to the
relation of R\&D and TFP which are not modeled. If trade plays no role for
the transmission of technology---embodied in intermediate inputs---then only
own-industry R\&D will be significantly correlated with TFP, and $\beta
_2,\beta _3$, and $\beta _4$ will be equal to zero. Another interesting
hypothesis is that domestic, other industry as well as foreign, and foreign,
other industry R\&D have the same effect as own R\&D expenditures, in which
case $\beta _2,\beta _3$, and $\beta _4$ will equal one.

\subsection{Capturing Intersectoral Technology Flows Beyond Input-Output
Relations}

We have given above several reasons why intersectoral trade relations might
only account for a part of technology diffusion between sectors. Scherer
(1984) and Evenson et al. (1991) in particular are among those who have gone
beyond relating intersectoral technology flows to intersectoral trade alone.
Both studies estimate a technology flow matrix, denoted with ${\bf TM},$ 
\begin{equation}
{\bf TM}=\left[ 
\begin{array}{cccc}
\pi _{jj} & \pi _{jv} & \pi _{jw} & \cdots \\ 
\pi _{vj} & \pi _{vv} & \vdots & \vdots \\ 
\vdots & \vdots & \ddots & \vdots \\ 
\cdots & \cdots & \cdots & \ddots
\end{array}
\right] .  \label{tm}
\end{equation}
A typical element $\pi _{jv}$ indicates to what extent the R\&D conducted in
sector $j$ is used in sector $v$. In this approach, the entries for any row $%
j$ sum to one, i.e., $\sum_j\pi _{vj}=1,\forall v;$ this includes the R\&D
which is both conducted and used in sector $v,$ that is, the elements on the
main diagonal of {\bf (}\ref{tm}{\bf )}. We are primarily interested in
using the technology flow matrix as an alternative to the input-output
matrix ${\bf \Omega }$ in capturing intersectoral technology diffusion
between sectors, and therefore set the main diagonal in ${\bf TM}$ to zero.
The magnitude of the off-diagonal elements in the matrix ${\bf TM}$ then
indicates to what extent the R\&D results of a row industry $v$ are
beneficial to another column industry $j$.\footnote{%
The fact that any row in ${\bf TM}$ sums to one implies a private
goods-notion of R\&D, because any R\&D dollar spent in industry $j$ can only
allocated to one particular industry (including own industry); this differs
from the public goods nature of the production designs in, e.g., a Romer
(1990)-type model. However, this issue is well-recognized; see Scherer
(1984), pp. 432-5 for a discussion. For our empirical purposes, the issue is
not central; what matters is that the off-diagonal elements in a given row
capture the relative extent to which column industries benefit from the row
industry's R\&D.} Let $b_{ij}^{tm}$ and $b_{ij}^{f,tm}$ denote the domestic,
other industry, and the foreign, other industry R\&D variable, which are
computed in analogy to $b_{ij}^{io}$ and $b_{ij}^{f,io},$ respectively, by
replacing the input-output matrix, and the import input-output matrix,
respectively, with the technology flow matrix ${\bf TM}$. The technology
flow-based specification is then given by 
\begin{equation}
\log f_{ij}=\mu _{ij}+\beta _1\log \left( b_{ij}+\beta _2b_{ij}^{tm}+\beta
_3b_{ij}^f+\beta _4b_{ij}^{f,tm}\right) \,+\varepsilon _{ij},\forall i,j.
\label{full2}
\end{equation}
With equation (\ref{full2}), it is implicitly assumed that the general form
of how outside industry R\&D matters for productivity is the same as in the
model of technology diffusion through intermediate inputs trade, and that
technology diffusion beyond intermediate inputs trade is appropriately
captured by substituting $b^{tm}$ and $b^{f,tm}$ for $b^{io}$ and $b^{f,io},$
respectively.

We now turn to describing the basic characteristics of the data.

\section{Data}

This paper uses data for eight OECD countries for the years 1970-1991 (for
more details on data sources and construction, see the appendix). The
countries are Canada, France, Germany, Italy, Japan, Sweden, the United
Kingdom, and the United States, hence, the G-7 group plus Sweden. We use an
industry classification with thirteen two- to three-digit manufacturing
industries according to the UN\ International Standard Industrial
Classification (ISIC).\footnote{%
These are: (1) ISIC (adjusted revision 2) 31 Food, beverages, and tobacco;
(2) ISIC 32 Textiles, apparel, and leather; (3) ISIC 33 Wood products and
furniture; (4) ISIC 34 Paper, paper products and printing; (5) ISIC 351+352
Chemicals and drugs; (6) ISIC 353+354 Petroleum refineries and products; (7)
ISIC 355+356 Rubber and plastic products; (8) ISIC 36 Non-metallic mineral
products; (9) ISIC 37 Basic metal industries; (10) ISIC 381 Metal products;
(11) ISIC 382+385 Non-electrical machinery, office and computing equipment,
and professional goods; (12) ISIC 383 Electrical machines and communication
equipment; and (13) ISIC 384 Transportation equipment.} A TFP (index) is
constructed using the Structural Analysis industrial (STAN) database of the
OECD (1994) by first calculating the growth of TFP as the difference between
output and factor-cost share weighted input growth. Then, the level of TFP
is normalized to 100 in 1970 for each of the 8 x 13 time series. In Table
A.1, summary statistics on the TFP data are shown.

As defined above, the unobservable technology stock variable $n$ is
identified with the sectoral cumulative R\&D stocks, derived from OECD
(1991) data on private R\&D expenditures.\footnote{%
We assume a rate of knowledge depreciation, $\delta ,$ of 10\%, which is
typical in this literature (Nadiri and Prucha 1993). This is contrary to the
model above, where $\delta =0$. In Table A.7, we show results based on
alternative assumptions on $\delta ,$ which are discussed below.} This data
covers all intramural business enterprise expenditures. Summary statistics
on this data are given in Table A.2. The R\&D stocks are derived from the
R\&D expenditure series using the perpetual inventory method.

Constructing the import-weighted foreign R\&D capital stocks as described
above requires data on bilateral import flows. These are obtained from the
World Trade Data Base of the Hamburg Institute of Economic Research (HWWA).
It is clear from the construction of the $b^f$ variables that the origin of
a given country's imports (together with the R\&D efforts there) determines
the size of the foreign R\&D variable of the importing domestic industry. In
Tables A.3-1 and A.3-2, a subset of these bilateral import shares by sector
are shown for sectors ISIC 31 and ISIC 384.

We employ the input-output matrix of the U.S. economy for all countries in
the sample. It is derived from the benchmark input-output Table 2 published
in U.S. Department of Commerce (1991). The $13\times 13$ matrix of
input-output coefficients can be found in Table A.4. The input-output matrix
for imports is also derived from U.S. data, and assumed to be the same for
all countries. It is based on unpublished material of the U.S. Department of
Commerce (1996) on the use of commodities by industry in the import sector.%
\footnote{%
This data was collected in conjunction with the 1987 benchmark survey.} The $%
525\times 505$ matrix is aggregated up to the $13\times 13$ industry
classification used in this paper, and is shown in Table A.5. As can be seen
from tables A.4 and A.5, to avoid double counting of own-industry R\&D, the
own-industry effect has been eliminated by setting the main diagonals equal
to zero. The technology flow matrix ${\bf TM}$ is based on the work by
Evenson et al. (1991). These authors have created a concordance between the
industry-of-origin and the industry-of-use of an invention, using Canadian
patent data at a 4-digit SIC level.\footnote{%
The work by Evenson et al. (1991) is similar in spirit to the project by
Scherer (1984) and his associates.} Here, this matrix, called the ''Yale
Technology Concordance'', has been aggregated up to the $13\times 13$
industry classification used in this paper. The technology flow matrix is
shown in Table A.6.\footnote{%
The bilateral trade shares matrices are averaged over time (1972-91); the
input-output matrix reflects the relations in the U.S. economy in the
benchmark year of 1980; the import input-output matrix is for the benchmark
year of 1987, and the patent classification data for the Yale Technology
Concordance is from the years 1978-87.}

It is important to note that the practice of data collection and preparation
by the statistical agencies is such that part of the estimated spillover
effect due to input-output relations is spurious, primarily because of the
unavailability of deflators which fully adjust for increases in the quality
of a product, or because these deflators are not continuously, but only
discretely adjusted. Therefore, measured TFP in sector $j$ is affected by
productivity improvements in industry $v\neq j$ to the extent that it
purchases from $v$ and that these improvements have not been incorporated in
the official input price indices of industry $j$.\footnote{%
Another important determinant of the magnitude of these measured TFP effects
is the extent to which the supplying industry can appropriate the
improvements in the quality of its product; see Griliches (1979) for more
discussion on this.} While we do not have a way of estimating how large
these effects are, and whether they are invariant across sectors and
countries, the presence of these effects generally suggests that the
spillovers estimates presented below should be viewed as an upper bound on
the true spillover effects.

\section{Estimation Results}

There are eight countries with thirteen industries each, for the time period
of 1970-91, making a total of 2288 annual observations. We allow for a
generalized time trend by including a fixed effect for each year; further, a
full set of 104 (8 x 13) country-industry fixed effects are employed in all
regressions. Rather than including these sets of dummy variables in the
regression equation, the log of any variable $q,$ $q=f,b,b^{io},b^{tm},b^f,$ 
$b^{f,io},$ and $b^{f,tm}$ is obtained by first subtracting the year-, and
then the country-industry-means from the data.

\subsection{Technology Diffusion through Goods Trade at Arm's Length}

In this section, we present results based R\&D measures which capture trade
in goods alone: through the input-output matrix (leading to $b^{io}$), the
international bilateral trade share matrix (giving the variable $b^f$), and
the bilateral import share matrix together with the import input-output
matrix (resulting in the variable $b^{f,io}$). The specification used in the
regression is given by 
\begin{equation}
\log f_{ijt}=\eta _0+\beta _1\log \left( b_{ijt}+\beta _2b_{ijt}^{io}+\beta
_3b_{ijt}^f+\beta _4b_{ijt}^{f,io}\right) +\varepsilon _{ijt},\forall i,j,t.
\label{basic}
\end{equation}
Table 1 shows the results of these estimations. The method is non-linear
least squares, which simplifies to ordinary least squares (OLS) in
regression (T1.1). The elasticity of TFP with respect to R\&D, $\beta _1,$
is estimated to be 7.4\% in regression (T1.1), significantly different from
zero at a 1\% level.\footnote{%
In the tables, a parameter which is significantly different from zero at a
5\% (10\%) level is denoted with $^{**}(^{*}).$} When, with the $b^{io}$
variable, the effect of R\&D via domestic intermediate inputs is included,
the R\&D elasticity is raised to 19.4\%, and the estimated coefficient on $%
b^{io}$ is, with $1.639$, larger than one (see T1.2). Taken at face value,
this implies that domestic R\&D external to an industry has larger TFP
effects than the R\&D undertaken in the industry itself. When also the
foreign, same-industry R\&D variable is included, in (T1.3), we obtain an
estimate on $b^f$ which is only slightly lower than one, suggesting that
foreign same-industry R\&D substitutes almost perfectly for domestic R\&D in
this sample of industries. Finally, regression (T1.4) gives the full
specification of (\ref{basic}), including the effect from foreign,
other-sector R\&D ($b^{f,io}$). The effect from R\&D in foreign industries
outside the receiving industry is estimated to have no significant effect on
domestic TFP, with a parameter estimate of -0.885, and an estimated standard
error of 0.773. According to the full specification, the elasticity with
respect to own-industry R\&D is 21.2\%, the effect from domestic,
different-sector R\&D is more than double that ($\beta _2=2.316$), and the
effect from foreign, same-sector R\&D through input-output channels is
estimated to be very similar to domestic, same-sector R\&D ($\beta _3=1.01$).

It is surprising to see that the estimate of the productivity effects of
domestic outside-industry R\&D, $\beta _2,$ exceeds one. This implies a very
strong, and, it seems, too strong form of R\&D spillovers. According to
Figure 1, the R\&D intensity--defined as R\&D\ expenditures divided by the
value of production--is smaller than three percent in all but four
industries. While it is therefore possible that outside-industry
productivity effects are quite strong relative to own R\&D effects, it is
nevertheless implausible to estimate a stronger outside-industry
productivity effect than the effect from own R\&D. For one, this finding
stands in contrast to most of the comparable earlier studies; see the survey
by Griliches (1995). For another, the fact that our sample constitutes
exclusively of manufacturing industries makes the finding of $\beta _2\gg 1$
even less plausible: Non-manufacturing industries might benefit to a large
extent from technology spillovers provided by manufacturing industries, but
within a sample of manufacturing industries, the own-R\&D effect is almost
certainly larger than the productivity effect from outside-industry R\&D.%
\footnote{%
A similar point is made in Griliches and Lichtenberg (1984b), footnote 5.}
We have tried to address this issue by making various changes in the
econometric specification of (\ref{basic}), by analyzing results by
individual industry, and by redefining the input-output weighing matrix,
without being able to fully resolve it.\footnote{%
These results are available upon request.} Thus we conclude that the 'pure
transactions view' of intersectoral technology diffusion, as captured by the
input-output matries{\bf ,} does not fully capture the process of
intersectoral technology diffusion. In the following, therefore, a
technology flow matrix is considered as an alternative to the input-output
matrices.

\subsection{Intersectoral Diffusion Captured by a Technology Flow Matrix}

In the following, we substitute the variables $b^{tm}$ and $b^{f,tm}$ for $%
b^{io}$ and $b^{f,tio}$, respectively, as shown in equation (\ref{basic}) 
\[
\log f_{ijt}=\eta _0+\beta _1\log \left( b_{ijt}+\beta _2b_{ijt}^{tm}+\beta
_3b_{ijt}^f+\beta _4b_{ijt}^{f,tm}\right) \,+\varepsilon _{ijt},\forall
i,j,t, 
\]
while keeping the foreign, same sector R\&D variable $b^f$ unchanged. The
technology flow matrix on which basis the variables are constructed is
derived from Evenson et al. (1991).\footnote{%
In the spirit of Scherer (1984) and Evenson et al. (1991), the following
results are based on allocating the R\&D stock of a particular sector (in a
certain row of the ${\bf TM}$ matrix) fully to all using industries,
including the row industry itself. That means that the own-industry R\&D
variable $b_j$ is now scaled by the diagonal element of the technology
matrix shown in Table A.6, i.e., by $\pi _{jj}$. The procedure corresponds
to a private goods notion of these R\&D stock, see the discussion above. The
estimation results are very similar, whether $b_j$ is scaled by $\pi _{jj}$
or not.} Table 2 shows the results of this estimation.

The OLS regression of (T2.1) is identical to (T1.1). The specification which
includes the $b^{tm}$ variable generates an estimate of $\beta _1=0.107$ for
the domestic, own-industry R\&D variable, and a coefficient of $0.472$ for
domestic, outside-industry R\&D; the latter is not precisely estimated,
though, and not statistically significant at standard levels (p-value of
0.22)$.$ When the variable $b^f,$ based on international trade in the same
industry, is included, the domestic, own-industry elasticity estimate rises
to $\beta _1=0.171,$ the domestic, outside-industry effect remains at about
50\% of that ($\beta _2=0.511$, again not significant at standard levels),
and the effect from foreign R\&D through trade in the same industry, $b^f,$
is estimated to be $0.952,$ statistically significant at a 1\% level. In
regression (T2.4), also the foreign, other-industry variable $b^{f,tm},$ is
included. Now, the domestic, own-industry effect is estimated to be $0.103$,
and the foreign, same-industry effect ($\beta _3$) is equal to $0.848$. The
domestic, other-industry effect is now estimated to be statistically
significantly larger than zero, but it is also larger than one, with $\beta
_2=1.24$. At the same time, the foreign, other-industry effect is estimated
to be large and significantly negative, with $\beta _4=-1.808$. It is,
however, rather difficult to think of a reason why foreign R\&D in other
sectors should have a large negative effect on domestic productivity; more
likely, the estimate of $\beta _4$ picks up something else, and given that
it statistically significant and large, one should in general discount the
results of regression (T2.4).

It appears that, with the given approach and the data available, we cannot
adequately trace the productivity effects from foreign R\&D in other sectors.%
\footnote{%
This might be not surprising, given that the variable $b^{f,tm}$ combines
elements of the 'transactions' approach to technology diffusion, as captured
in the bilateral import share matrix used to compute $b^f,$ with a more
general view of technology flows (computing $b^{f,tm}$ with the matrix ${\bf %
TM}$); the two weighing matrices might not be compatible. We have
experimented with an alternative foreign, other-industry variable, $\hat
b^{f,tm}$, constructed as follows: Let $\hat b_{ijt}^f\equiv \sum_{h\neq
i}^Ib_{hjt},\forall i,j,t$. Then, the variable $\hat b^{f,tm}$ is computed
by channeling $\hat b^f$ through the matrix ${\bf TM}$. That is, we simply
sum up the R\&D stocks in a given industry and year, except a given
country's own R\&D, and then feed these variables through the technology
flow matrix to compute the foreign, other-industry variable. However, also
this variable has a large and negative coefficient.} On the plus side,
employing the technology flow matrix leads to plausible estimates for the
domestic, same-industry as well as the foreign, same-industry effect, and in
particular for the productivity effect from domestic, other-industry R\&D,
even though the latter is not precisely estimated. Overall, from this sample
of manufacturing industries, we estimate an elasticity with respect to
own-industry R\&D of 10-17\%, a foreign, own-industry effect which is almost
as strong as the domestic, own-industry effect, and a domestic,
other-industry effect which seems to be about half the size of the
own-industry effect, although it is estimated rather imprecisely,

We can formally test some interesting hypotheses. In regression (T2.3),
which does not include the foreign, other-industry R\&D stock ($b^{f,tm}$),
we cannot reject the null hypotheses that $\beta _2=\beta _3=1$ at a 5\%
level: a likelihood ratio test gives a statistic of 4.54, whereas the
critical value $\chi ^2\left( 2,5\%\right) $ is equal to 5.99. The same null
hypothesis can, however, be rejected at a significance level somewhat larger
than 10\% (critical value $\chi ^2\left( 2,10\%\right) =4.61$). Contrary to
that, the null hypothesis that domestic, other-industry R\&D has a
productivity effect which is half the size of own-industry, and foreign,
own-industry R\&D ($H_0:\beta _2=0.5,\beta _3=1$) is, with a likelihood
ratio test statistic of 2.08, very hard to reject.\footnote{%
It is clear that we could not reject the null that $\beta _2=0.5$ and $\beta
_3=0.9,$ or, $\beta _2=0.5$ and $\beta _3=1.1,$ either.} In that restricted
regression, the own-industry R\&D coefficient $\beta _1$ is estimated to be $%
0.169$. The hypothesis that domestic, other-industry R\&D has the same
productivity effects as own-industry R\&D ($H_0:\beta _2=1$) is, with a
likelihood ratio test statistic of $3.52$, not rejected at a 5\% level
(critical value $\chi ^2\left( 1,5\%\right) =3.84$), but at a 10\% level
(critical value $\chi ^2\left( 1,10\%\right) =2.71$). Lastly, the hypothesis
that foreign R\&D in the same industry has the same productivity effects as
domestic R\&D ($H_0:\beta _3=1$) cannot be rejected even at a 10\% level of
significance.\footnote{%
In this footnote we discuss the sensitivity of the results with respect to
the assumptions on the lag structure and obsolescence of technological
knowledge; see Tables A.7-1 and A.7-2 in the appendix. The former includes
the regressor $b^{f,tm},$ whereas the latter does not. On the lag structure
issue, note that the estimation results change very little if we assume that
foreign R\&D has its productivity effect with a one-year lag; $\beta _3$
equals $0.812$ with lag ($0.848$ without) in A.7-1, and $0.959$ with lag
(versus $0.952$ without) in A.7-2. Assuming different knowledge depreciation
rates makes more of a difference. Assuming $\delta =0$ ($\delta =0.2$) leads
to a higher (lower, generally) coefficient on own R\&D, relative to assuming 
$\delta =0.1$. In A.7-2, the coefficient on $b^{tm}$ is higher (lower) for $%
\delta =0$ ($\delta =0.2$), relative to assuming $\delta =0.1$; with $\delta
=0,$ the coefficient on $b^{tm}$ is significantly positive at a 10\% level,
and approximately equal to the foreign, same-industry effect. For $\delta
=0.2,$ $\beta _2$ is essentially zero. Lastly, assuming $\delta =0$ leads to
a lower estimate for the productivity effect from foreign, same-industry
R\&D in A.7-2, compared to assuming $\delta =0.1$. Overall, it is clear that
different assumptions on $\delta $ have, in particular for $b^{tm}$, some,
but no overwhelming effect on the estimates. Below we will attempt to obtain
a more precisely estimate of $\beta _2$ by considering only a subsample of
the data.}

We might be able to infer more by restricting the sample to R\&D intensive
industries; naturally, the point estimates might differ, but because
measurement and imputation errors are likely to be relatively smaller,
focusing on R\&D intensive industries perhaps allows to obtain more precise
estimates. As seen from Figure 1, there are four industries which have a
considerably higher R\&D intensity--defined as R\&D\ expenditures divided by
the value of production-- than the other sectors: Chemicals and Drugs (ISIC
351/2), Non-electrical Machinery and Precision/Optical Instruments\ (ISIC
382/5), Electrical Machinery (ISIC 383), and Transportation (ISIC 384). In
the following section, we focus on those industries. This reduces the sample
size to 704 observations. Table 3 gives the results of specification (\ref
{full2}) in the restricted sample.

As regression (T3.1) shows, the own-industry R\&D elasticity is estimated to
be $0.442,$ about six times as large as in the full sample. Including the
domestic, other-industry variable $b^{tm},$ we estimate a coefficient of $%
0.192,$ significant at a 10\% level. Note that this implies an elasticity
with respect to domestic, other-industry R\&D of $9.8\%$ ($0.512\times 0.192$%
), which is about double of our estimate in the full sample ($0.107\times
0.472=5.1\%,$ regression T2.2). In regression (T3.3), the effect from the
import-channeled foreign, same-industry R\&D, $b^f$, is included: we
estimate $\beta _1=0.730,$ $\beta _2=0.381,$ and $\beta _3=0.577,$ where all
coefficients are significantly different from zero at a 1\% level. Lastly,
including the foreign, other-industry variable $b^{f,tm}$ results, as in the
full sample, in a significantly negative coefficient for $\beta _4,$ with $%
-0.882$ (see T3.4), which is hard to rationalize on a priori grounds.
Although the own-industry R\&D effect drops from $0.73$ to $0.443,$ we note
that the coefficients $\beta _2$ and $\beta _3$ remain at about $0.5$ of the
own-industry effect.

Also in the subsample of R\&D intensive industries, we can test several
hypotheses; consider regression (T3.3) which omits the $b^{f,tm}$ regressor.
First, the null hypothesis $H_0:\beta _2=\beta _3=1$ is, with a likelihood
ratio test statistic of 15.4, rejected at a 1\% level (critical value $\chi
^2\left( 2,1\%\right) =10.60$). Contrary to that, the null that $\beta
_2=\beta _3=0.5$ is not rejected even at a 10\% level (test statistic of
4.14). Second, the null hypotheses that either the domestic, other-industry
effect or the foreign, same-industry effect is equally strong as the
own-industry effect are both rejected at a 1\% level. Contrary to that, the
hypotheses that $\beta _2=0.5$ or $\beta _3=0.5$ are not rejected at a 10\%
level.

Comparing the results for all industries and the R\&D intenstive subsample
using the results of (T2.3) and (T3.3), we find, first, that the
own-industry elasticity is about four times larger for R\&D intensive
industries than for all thirteen manufacturing industries ($0.73$ versus $%
0.17$). The finding of larger R\&D elasticities for R\&D intensive
industries is typical in this literature\ (e.g., Englander et al. 1988).
Second, the benefit from other sector's R\&D is estimated to be relatively
smaller for R\&D intensive industries. This holds both for domestic,
other-industry R\&D as well as for foreign, same-industry R\&D ($0.38$
versus $0.51$ for the former, and $0.57$ versus $0.95$ for the latter). This
is consistent with the notion that less R\&D intensive industries tend to be
to a larger extent technology 'users' than industries which conduct a lot of
R\&D themselves.

Third, the finding that R\&D intensive industries benefit to a lesser extent
from outside R\&D is more pronounced for foreign, same-industry R\&D than
for domestic, other-industry R\&D. In the former, the benefit of R\&D
intensive industries from outside R\&D falls to 60\% of that in the full
sample, whereas it falls only to 75\% of that in the full sample in the case
of domestic, outside-industry R\&D. While there are many possible
explanations for this finding, one might be that R\&D intensive firms tend
to operate in a monopolistically competitive environment, where a relatively
high degree of appropriatation by the inventors leads to lower
intra-industry technology spillovers.

Summarizing, for the sample of all manufacturing industries (R\&D intensive
manufacturing industries), we find own-R\&D elasticities in the range of
7-17\% (44-73\%). The elasticity for domestic R\&D\ from other industries is
estimated to be in the range of 5-9\% (10-27\%), and we calculate an
elasticity for foreign R\&D in the same industry of about 16\% (42\%).%
\footnote{%
For the input-output specification, we have estimated a $\beta _3$ of about
23\%, see (T1.3).} Note that the magnitude of the effects of foreign R\&D is
much larger than estimated, e.g., in Coe and Helpman (1995), who estimate a
foreign R\&D elasticity of only 6\%.\footnote{%
Coe and Helpman (1995), p.869, Table 3, (ii)). This comparison is even more
striking if one takes into account that Coe and Helpman's specification (a)
does not include year fixed-effects, and (b) their analysis uses
country-level data. The former implies that their estimate is an upper bound
of the 'true' effect, because without year effects, the regression is likely
to pick up time trends. On point (b), the use of country-level data should
lead to strong spillover effects, because at a country level of aggregation,
productivity effects are internalized whereas they might be external effects
at an industry level (see Griliches 1995). The result has to do with Coe and
Helpman's specification, which is log-linear: $\log f=\alpha +\kappa _1\log
b+\kappa _2\log b^f+\varepsilon $. Using this specification in our context,
we estimate an elasticity with respect to foreign R\&D in the same industry (%
$\kappa _2$) of about 4\%, as opposed to 16\%. Given the model laid out
above, however, the non-linear specification is preferred.}

\section{Conclusions}

A model of R\&D-driven growth has been presented which predicts that
technology in form of product designs, created through R\&D\ investments, is
transmitted to other sectors by being embodied in differentiated
intermediate goods demanded by these sectors. This occurs both domestically
as well as internationally. We derive the predictions of the model on how
other sectors, as well as other countries R\&D investments should affect
domestic productivity. Empirical results are presented employing data from
thirteen manufacturing industries in eight OECD countries over the period of
1970 to 1991.

The empirical analysis, first, confirms earlier findings that cumulative
R\&D expenditures investments are positively related to productivity levels,
and we estimate an elasticity of TFP with respect to own-industry R\&D\
between 7\% and 17\%. Secondly, the receiving industry's productivity level
benefits, as predicted by the model, also from other industries' technology
investments, an effect which is at least in part due to trade in embodied
technology. We find that the benefit derived from foreign R\&D\ in the same
industry is in the order of 50-95\% of the productivity effect of own R\&D.
These results are consistent with international trade being an important
transmittent of foreign technology in the same industry.

For domestic interindustry technology flows, the results strongly suggest
that trade in goods is not all what matters for technology transmission, as
the results based on a technology flow matrix are preferred to results based
only on the input-output structure of the economy. We estimate that
domestic, outside-industry R\&D is one-fifth to one-half as effective in
raising productivity as own-industry R\&D. These results suggest that at the
two- to three-digit industry level, industries benefit generally more from
foreign technology creation in the same industry than from domestic
technology creation in other industries. Our attempts to trace technology
flows from other foreign sectors are largely unsuccessful: the estimates
obtained in that respect are never very plausible. This question will have
to be addressed by future work.

Some additional support for the notion that technology is transmitted in
part through being embodied in goods which are traded at arm's length comes
from considering different types of industries within the sample
individually. First, we find that outside R\&D is more important for sectors
which are themselves not conducting much R\&D on their own--as it should be,
according to the embodied technology transmission hypothesis. Second, the
finding that in R\&D intensive industries, there is only a relatively small
gain from foreign, same-industry R\&D is consistent with the notion that
market conduct in these industries tends to be monopolistically competitive,
with internalization of the return to the R\&D investment being a primary
concern, and where there is therefore little trade of innovative products
among competing firms.

Finally, it is important to note that this study has not addressed the
important question of the extent to which technology transmission is
embodied, relative to being disembodied. This is evidenced by the fact that
we have not tried to estimate the technology flow matrix, computing the
'effective' R\&D for a given sector, jointly with the relation between that
'effective' R\&D and productivity. Estimating the channels of technology
transmission, and quantifying the extent to which technology transmission is
embodied (and disembodied) is therefore an important topic for future
research.\newpage\ \pagestyle{empty}\setlength{\oddsidemargin}{0.0in} 
\[
\fbox{$
\begin{tabular}{lcccc}
\multicolumn{5}{c}{Table 1: Intersectoral Input-Output and International
Trade Specification} \\ 
\multicolumn{5}{c}{Dependent Variable: Log of TFP index; 2288 observations}
\\ \hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{(T1.1)} & \multicolumn{1}{c|}{
(T1.2)} & \multicolumn{1}{c|}{(T1.3)} & (T1.4) \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _1\text{: Same Sector, } \\ 
\text{Domestic R\&D (}b\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.074^{**} \\ 
(0.010)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.194^{**} \\ 
(0.029)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.252^{**} \\ 
(0.032)
\end{array}
$} & $%
\begin{array}{c}
0.212^{**} \\ 
(0.058)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _2\text{: Other Sector,} \\ 
\text{Domestic R\&D (}b^{io}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{$%
\begin{array}{c}
1.639^{**} \\ 
(0.471)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
2.021^{**} \\ 
(0.567)
\end{array}
$} & $%
\begin{array}{c}
2.316^{**} \\ 
(0.633)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _3\text{: Same Sector,} \\ 
\text{Foreign R\&D (}b^f\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.911^{**} \\ 
(0.276)
\end{array}
$} & $%
\begin{array}{c}
1.010^{**} \\ 
(0.304)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _4\text{: Other Sector,} \\ 
\text{Foreign R\&D (}b^{f,io}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & 
$%
\begin{array}{c}
-0.885 \\ 
(0.773)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\text{F-statistic} \\ 
(\text{Degr. of freedom)}
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
52.67 \\ 
(1,2286)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
37.14 \\ 
(2,2285)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
30.99 \\ 
(3,2284)
\end{array}
$} & $%
\begin{array}{c}
23.79 \\ 
(4,2283)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$\bar R^2$} & \multicolumn{1}{c|}{$0.75$} & 
\multicolumn{1}{c|}{$0.759$} & \multicolumn{1}{c|}{$0.765$} & $0.766$
\end{tabular}
$} 
\]
\newpage
\[
\fbox{$
\begin{tabular}{lcccc}
\multicolumn{5}{c}{Table 2: Technology Flow and International Trade
Specification} \\ 
\multicolumn{5}{c}{Dependent Variable: Log of TFP index; 2288 observations}
\\ \hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{(T2.1)} & \multicolumn{1}{c|}{
(T2.2)} & \multicolumn{1}{c|}{(T2.3)} & (T2.4) \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _1\text{: Same Sector, } \\ 
\text{Domestic R\&D (}b\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.074^{**} \\ 
(0.010)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.107^{**} \\ 
(0.028)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.171^{**} \\ 
(0.035)
\end{array}
$} & $%
\begin{array}{c}
0.103^{**} \\ 
(0.043)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _2\text{: Other Sector,} \\ 
\text{Domestic R\&D (}b^{tm}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.472 \\ 
(0.383)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.511 \\ 
(0.417)
\end{array}
$} & $%
\begin{array}{c}
1.24^{**} \\ 
(0.386)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _3\text{: Same Sector,} \\ 
\text{Foreign R\&D (}b^f\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.952^{**} \\ 
(0.259)
\end{array}
$} & $%
\begin{array}{c}
0.848^{**} \\ 
(0.216)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _4\text{: Other Sector,} \\ 
\text{Foreign R\&D (}b^{f,tm}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & 
$%
\begin{array}{c}
-1.808^{**} \\ 
(0.498)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\text{F-statistic} \\ 
(\text{Degr. of freedom)}
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
52.67 \\ 
(1,2286)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
26.02 \\ 
(2,2285)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
24.33 \\ 
(3,2284)
\end{array}
$} & $%
\begin{array}{c}
24.62 \\ 
(4,2283)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$\bar R^2$} & \multicolumn{1}{c|}{$0.75$} & 
\multicolumn{1}{c|}{$0.75$} & \multicolumn{1}{c|}{$0.758$} & $0.768$
\end{tabular}
$} 
\]
\newpage
\[
\fbox{$
\begin{tabular}{lcccc}
\multicolumn{5}{c}{Table 3: Technology Flow and International Trade
Specification} \\ 
\multicolumn{5}{c}{R\&D Intensive Industries} \\ 
\multicolumn{5}{c}{Dependent Variable: Log of TFP index; 2288 observations}
\\ \hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{(T3.1)} & \multicolumn{1}{c|}{
(T3.2)} & \multicolumn{1}{c|}{(T3.3)} & (T3.4) \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _1\text{: Same Sector, } \\ 
\text{Domestic R\&D (}b\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.442^{**} \\ 
(0.040)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.513^{**} \\ 
(0.054)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.730^{**} \\ 
(0.068)
\end{array}
$} & $%
\begin{array}{c}
0.443^{**} \\ 
(0.093)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _2\text{: Other Sector,} \\ 
\text{Domestic R\&D (}b^{tm}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.192^{*} \\ 
(0.103)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.381^{**} \\ 
(0.131)
\end{array}
$} & $%
\begin{array}{c}
0.556^{**} \\ 
(0.138)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _3\text{: Same Sector,} \\ 
\text{Foreign R\&D (}b^f\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{$%
\begin{array}{c}
0.577^{**} \\ 
(0.144)
\end{array}
$} & $%
\begin{array}{c}
0.499^{**} \\ 
(0.131)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\beta _4\text{: Other Sector,} \\ 
\text{Foreign R\&D (}b^{f,tm}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{} & 
$%
\begin{array}{c}
-0.882^{**} \\ 
(0.192)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$%
\begin{array}{c}
\text{F-statistic} \\ 
\text{(Degr. of freedom)}
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
122.64 \\ 
(1,702)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
63.50 \\ 
(2,701)
\end{array}
$} & \multicolumn{1}{c|}{$%
\begin{array}{c}
52.31 \\ 
(3,700)
\end{array}
$} & $%
\begin{array}{c}
53.14 \\ 
(4,699)
\end{array}
$
\end{tabular}
$} 
\]

\setlength{\oddsidemargin}{0.5in}\newpage\ 

\begin{thebibliography}{99}
\bibitem{}  Branstetter, L. (1996), ''Are Knowledge Spillovers International
or Intranational in Scope? Microeconometric Evidence from the U.S. and
Japan'', mimeo, Harvard University, August.

\bibitem{ch}  Coe, D.T., and E. Helpman (1995), ''International R\&D
Spillovers'', {\it European Economic Review} 39: 859-887.

\bibitem{chh}  Coe, D.T., E. Helpman, and A.W. Hoffmaister (1995),
''North-South R\&D Spillovers'', CEPR Working Paper No.1133, London.

\bibitem{eatkor}  Eaton, J., and S. Kortum (1996), ''Trade in Ideas:
Patenting and Productivity in the OECD'', {\it Journal of International
Economics} 40: 251-278.

\bibitem{engl}  Englander, A., R. Evenson, and M. Hanazaki (1988), ''R\&D,
Innovation and the Total Factor Productivity Slowdown'', {\it OECD\ Economic
Studies} 11: 7-42.

\bibitem{et}  Ethier, W.J. (1982), ''National and International Returns to
Scale in the Modern Theory of International Trade'', {\it American Economic
Review} 72: 389-405.

\bibitem{Fullt}  Fullerton, D., and M. Karayannis (1993), ''United States'',
Chapter 10 in Jorgenson and Landau (1993a), pp.333-367.

\bibitem{gril}  Griliches, Z. (1979), ''Issues in Assessing the Contribution
of Research and Development to Productivity Growth'', {\it Bell Journal of
Economics} 10: 92-116.

\bibitem{gril}  Griliches, Z. (1984), {\it R\&D, Patents, and Productivity},
(ed.), Chicago: The University of Chicago Press.

\bibitem{gril}  Griliches, Z. (1992), ''The Search for R\&D Spillovers'', 
{\it Scandinavian Journal of Economics} 94: 29-47.

\bibitem{gril}  Griliches, Z. (1995), ''R\&D and Productivity: Econometric
Results and Measurement Issues'', in P.Stoneman (ed.), {\it Handbook of the
Economics of Innovation and Technological Change}, Blackwell, Oxford, UK,
pp.52-89.

\bibitem{grilich}  Griliches, Z., and F. Lichtenberg (1984a), ''R\&D and
Productivity Growth at the Industry Level: Is there Still a Relationship?'',
in Griliches (1984), pp.465-496.

\bibitem{grill}  Griliches, Z., and F. Lichtenberg (1984b), ''Interindustry
Technology Flows and Productivity Growth: A Reexamination'', {\it Review of
Economics and Statistics} 66: 324-329.

\bibitem{grhe}  Grossman, G., and E. Helpman (1991), {\it Innovation and
Growth in the World Economy}, Cambridge, MA.: MIT Press.

\bibitem{hall}  Hall, R.E. (1991), ''Invariance Properties of Solow's
Productivity Residual'', in Diamond (ed.), {\it %
Growth/Productivity/Employment}, MIT Press, Cambridge, Ma., pp.71-112.

\bibitem{jaffe}  Jaffe, A.B. (1986), ''Technological Opportunity and
Spillovers of R\&D: Evidence from Firms' Patents, Profits, and Market
Value'', {\it American Economic Review} 76: 984-1001.

\bibitem{john}  Johnson, D., R. E. Evenson, and C. DeBresson (1997),
''Innovation and Invention in Canada'', mimeographed, Yale University,
February.

\bibitem{jorg}  Jorgenson, D.W. (1993), ''Introduction and Summary'',
Chapter 1 in Jorgenson and Landau (1993a), pp.1-56.

\bibitem{jorgl}  Jorgenson, D.W., and R. Landau (1993a), {\it Tax Reform and
the Cost of Capital}, (eds.), The Brookings Institution, Washington, D.C.

\bibitem{jl}  Jorgenson, D.W., and R. Landau (1993b), ''Appendix'' in
Jorgenson and Landau (1993a), pp.369-406.

\bibitem{wo}  Keller, W. (1996), ''Absorptive Capacity: On the Creation and
Acquisition of Technology in Development'', {\it Journal of Development
Economics} 49: 199-227.

\bibitem{kort}  Kortum, S., and J. Putnam (1997), ''Assigning Patents to
Industries: Tests of the Yale Technology Concordance'', mimeographed, Boston
University, January.

\bibitem{lang}  Langer, C. (1986), {\it World Trade and Production.
Methodology and Tables in ISIC}, Hamburg: Verlag Weltarchiv.

\bibitem{mans}  Mansfield, E. (1984), ''Comment'' on the paper by Scherer,
in Griliches (1984), pp. 461-464.

\bibitem{nadi}  Nadiri, M. I. (1993), ''Innovations and Technological
Spillovers'', NBER Working Paper No. 4423, Cambridge, MA.

\bibitem{nadpr}  Nadiri, M.I., and I.R. Prucha (1993), ''Estimation of the
Depreciation Rate of Physical and R\&D Capital in the U.S. Total
Manufacturing Sector'', NBER Working Paper No. 4591, Cambridge, MA.

\bibitem{basic}  OECD (1991), {\it Basic Science and Technology Statistics},
various years, OECD, Paris.

\bibitem{stan}  OECD (1994), {\it STAN Database for Industrial Analysis},
OECD, Paris.

\bibitem{Pak}  Pakes, A., and M. Schankerman\ (1984), ''The Rate of
Obsolescence of Patents, Research Gestation Lags, and the Private Rate of
Return to Research Resources'', in Griliches (1984), pp.73-88.

\bibitem{park}  Park, W. (1995), ''International R\&D Spillovers and OECD\
Economic Growth'', {\it Economic Inquiry} 33: 571-591.

\bibitem{rivrom}  Rivera-Batiz, L., and P.M. Romer (1991), ''Economic
Integration and Endogenous Growth'', {\it Quarterly Journal of Economics}
CVI: 531-555.

\bibitem{rod}  Rodrik, D. (1995), ''Trade and Industrial Policy Reform'', in
J. Behrman and T.N. Srinivasan (eds.) {\it Handbook of Development Economics}%
, Vol.III, Amsterdam: North Holland.

\bibitem{rom90}  Romer, P.M. (1990), ''Endogenous Technological Change'', 
{\it Journal of Political Economy} 98: S71-S102.

\bibitem{schere}  Scherer, F.M. (1984), ''Using Linked Patent and R\&D Data
to Measure Interindustry Technology Flows'', in Griliches (1984), pp.
417-461.

\bibitem{terl}  Terleckyj, N.E. (1974), ''Effects of R\&D on the
Productivity Growth of Industries: An Exploratory Study'', Washington, D.C.:
National Planning Association.

\bibitem{usdpt}  U.S. Department of Commerce (1991), ''Benchmark
Input-Output Accounts for the U.S. Economy, 1982'', {\it Survey of Current
Business}, July: 30-71.

\bibitem{wonad}  Wolff, E.N., and M.I. Nadiri (1993), ''Spillover Effects,
Linkage Structure, and Research and Development'', {\it Structural Change
and Economic Dynamics} 4: 315-331.

\bibitem{wb}  World Bank (1987), {\it World Development Report 1987}, New
York: Oxford University Press for the World Bank.\pagestyle{empty}%
{\scriptsize \newpage}
\end{thebibliography}

\appendix

\section{Data on labor inputs, physical capital, and gross production}

Data on these variables comes from the OECD (1994) STAN database. It
provides internationally comparable data on industrial activity by sectors,
primarily for OECD countries. The STAN figures are not submitted by the OECD
member countries, but based on estimates by the OECD, which tries to ensure
greater international comparability. See OECD (1994) for the details on
adjustments of national data.

In constructing the TFP variable, we consider only inputs of labor and
physical capital. Data on labor inputs $l$ can be extracted directly from
the STAN database--the number of workers engaged. This includes employees as
well as the self-employed, owner proprietors and unpaid family workers. The
physical capital stock data is not available in that data base, but gross
fixed capital formation in current prices is. We first convert the sectoral
investment flows into constant 1985 prices. The deflators used for that are
output deflators, because sectoral investment goods deflators were
unavailable for this study. The capital stocks are then estimated using the
perpetual inventory method, with--suppressing the industry subscripts-- 
\begin{equation}
\begin{array}{ccc}
k_{it} & = & (1-\delta _i)\,k_{it-1}+inv_{t-1},\,\,\text{for }%
t=2,...,22,\,\,i=1,...,8 \\ 
& \text{and} &  \\ 
k_{i1} & = & \frac{inv_{i1}}{(g_i+\delta _i)}\text{ \thinspace , }i=1,...,8,
\end{array}
\label{kstock}
\end{equation}
where $inv$ is gross fixed capital formation in constant prices (land,
buildings, machinery and equipment), $g$ is the average annual growth rate
of $inv$ over the period 1970-1991, and $\delta $ is the rate of
depreciation. Country-specific depreciation rates are used, taken from
Jorgenson and Landau (1993b), Table A-3: 
\[
\frame{$
\begin{array}{lccccl}
\text{Canada: 8.51\%} &  &  &  &  & \text{Japan: 6.6\%} \\ 
\text{France: 17.39\%} &  &  &  &  & \text{Sweden: 7.7\%} \\ 
\text{Germany: 17.4\%} &  &  &  &  & \text{United Kingdom: 8.19\%} \\ 
\text{Italy: 11.9\%} &  &  &  &  & \text{United States: 13.31\%}
\end{array}
$ } 
\]
The numbers, which are used throughout, are estimates for machinery \&
manufacturing in the year 1980.

The parameter $\alpha _{ijt}$ is the share of the labor in total production
costs. Following the approach by Hall (1991), the $\alpha _{ijt}$'s are not
calculated as the ratio of total labor compensation to value added (the
revenue-based factor shares), but as cost-based factor shares, which are
robust in the presence of imperfect competition. For this we use the
framework of the integrated capital taxation model of King and Fullerton
(see Jorgenson 1993 and Fullerton and Karayannis 1993) and data provided in
Jorgenson and Landau (1993b). The effective marginal corporate tax rate $%
\tau $ is given by the wedge between before-tax ($p$) and after-tax rate of
return ($\rho $), relative to the former 
\begin{equation}
\tau =\frac{p-\rho }p.  \label{met}
\end{equation}
Here, the variable of interest is $p$, the user cost of capital. It will be
a function of the statutory marginal tax rate on corporate income, available
investment tax credits, the rates of depreciation, and other influences.

In the case of equity financing, the after-tax rate of return will be $\rho
=\chi +\varsigma ,$ where $\chi $ is the real interest rate, and $\varsigma $
is the rate of inflation. Jorgenson (1993) tabulates the values for the
marginal effective corporate tax rate, $\tau ,$ in Table 1-1. According to
the ''fixed-r'' strategy, one gives as an input a real interest rate $\chi $
and deduces $\tau .$ In this case, we use a value of $\chi =0.1$, which,
together with the actual values of $\varsigma $ allows, using equation (\ref
{met}) to infer the user cost of capital, $p$. From Jorgenson's Table 1-1 on 
$\tau $, we use the values on ''manufacturing'' (the 1980 values given are
used for 1970-1982 in the sample, the 1985 values for 1983-1986, and
Jorgenson's 1990 values are used for 1987-1991). This certainly introduces
an error; in addition, the Jorgenson Table 1-1 is derived from a ''fixed-p''
approach, as opposed to the ''fixed-r'' approach employed here. Further, the
results depend on the chosen real interest rate, $\tau $ varies by asset
type, and $\rho $ is a function of the way of financing (equity versus debt
primarily). Hence, there are, due to unavailability of more detailed data,
several shortcomings in the construction of the cost-based factor shares.
However, the chapter by Fullerton and Karayannis (1993) presents a
sensitivity analysis in certain dimensions. Second, limited experiments with
different values for $\chi ,$ the real interest rate, indicate that the
basic results presented above do not depend on a particular value for $\chi $%
. Finally, the cost-based approach has the advantage of using all data on
the user cost of capital compiled in Jorgenson and Landau (1993a) while at
the same time being robust to deviations from perfect competition.

Having obtained the series on the user cost of capital and capital stock
data, $\alpha $ is given by 
\begin{equation}
\alpha =\frac{w\,l}{w\,l+p\,k},  \label{ws}
\end{equation}
where $wl$ are the constant price labor costs. Labor and capital inputs
together with the factor shares allow to construct a Thornqvist index of
total inputs $I_t$%
\begin{equation}
\begin{array}{ccc}
\ln \left( \frac{I_{ijt}}{I_{ijt-1}}\right) & = & \frac 12*\left[ \alpha
_{ijt}+\alpha _{ijt-1}\right] \ln \left( \frac{l_{ijt}}{l_{ijt-1}}\right) \\ 
& + & \frac 12*\left[ (1-\alpha _{ijt})+(1-\alpha _{ijt-1})\right] \ln
\left( \frac{k_{ijt}}{k_{ijt-1}}\right) .
\end{array}
\label{torn}
\end{equation}
This gives a series of growth of total factor input. The TFP growth series
is obtained by subtracting total input from output growth. A value of 100 in
1970 is chosen for each of the $8\times 13$ time series, for all industries $%
j$ and countries $i.$

\section{Data on R\&D}

The raw data on R\&D expenditures comes from OECD (1991). It is more patchy
than the series on output, investment, and employment. This is not so much a
problem of the sectoral breakdown, because the national statistical offices
do collect their R\&D data along the lines of the two- or three-digit ISIC
classification. But R\&D surveys were not conducted annually in all
countries over the entire sample period. In the United Kingdom, for
instance, they were held only every third year until well into the 1980s. In
Germany, R\&D data is collected only bi-annually. Estimates of about 25\% of
all the R\&D expenditure data were obtained by interpolation.

The construction of the technology stock variables $n$ is based on data on
total business enterprise intramural expenditure on R\&D ($\phi $), in
constant 1985 US \$, and it uses the OECD purchasing power parity rates for
conversion. The OECD code for this series is BERD, given in Table 9B of OECD
(1991). Also here, we use the perpetual inventory method to construct
technology stocks, assuming that (industry and country subscripts are
suppressed) 
\begin{equation}
\begin{array}{ccc}
n_t & = & (1-\delta )\,n_{t-1}+\phi _{t-1},\,\,\text{for }t=2,...,22 \\ 
& \text{and} &  \\ 
n_1 & = & \frac{\phi _1}{(\lambda +\delta +0.1)}\text{ \thinspace .}
\end{array}
\label{techst}
\end{equation}
The rate of depreciation of the knowledge stock, $\delta ,$ is set at $0.1,$
and $\lambda $ is the average annual growth rate of $n$ over the period of
1970-1991.\footnote{%
The denominator in the calculation of $n_1$ is increased by $0.1$ in order
to obtain positive estimates of $n_1$ throughout.} A higher (lower) choice
of $\delta $ reduces (increases) the rate of growth of the knowledge stock
over the period of observation. Nadiri and Prucha (1993) estimate $\delta ,$
and obtain a value of $0.12;$ earlier estimates by Pakes and Schankerman
(1984) implied a $\delta $ of 0.25 (see the survey in Nadiri and Prucha
1993). Given that the data available for this study does not permit a good
estimate of $\delta ,$ we present a limited sensitivity analysis for
alternative values of $\delta $ in Table A.7.

\section{Data on Import Flows, Input-Output Relations, and the Technology
Flow Matrix}

Data on import flows for 1972, and 1975-91, in ISIC format, comes from the
HWWA institute in Hamburg. The HWWA data base, in turn, relies on UN
Commodity Trade Statistics, up to 1981, and the OECD Foreign Trade by
Commodities Statistics, from 1982 onward. Because this data originally is in
the Standard International Trade Classification (SITC) of the United
Nations, a conversion to the ISIC scheme is necessary. We implicitly use the
HWWA conversion scheme; it takes account of the three revisions the SITC
classification has been undergoing during the period of observation. Langer
(1986) describes the HWWA conversion from SITC to ISIC in some detail, and
also shows that the differences to the OECD's conversion scheme are minor.
Bilateral import shares are formed for all thirteen industries and eight
countries, and a subset of them are reported in Table A.3-1 and A.3-2. We
employ time-invariant bilateral import shares in the construction of the
variables $b^f$, $b^{f,io}$, and $b^{f,tm}$; these import shares are
calculated as the average over the year 1972 and 1975-91.

The sources of the economy-wide input-output relations are for the year
1980, from U.S. Department of Commerce (1991). The input-output table for
imports was not available for 1980; therefore, it has been derived from 1987
data, and draws on unpublished data as described in the text. The
input-output matrix is given in Table A.4, and import input-output matrix is
given in Table A.5. The technology flow matrix employed in this paper is
based on the Yale Technology Concordance, the construction of which is
described in detail in Evenson et al. (1991). The Yale Technology
Concordance is based on more than 180,000 individual patent classifications
at a 4-digit SIC level by the Canadian Intellectual Property Rights Office
over the years of 1978-87. The concordance has been updated in the meantime;
see Kortum and Putnam (1997) for tests of the patent assignments of the Yale
Technology Concordance, and Johnson, Evenson, and DeBresson (1997) for a
comparison of patents by industry as assigned by the Canadian Intellectual
Property Office and industry-level data from innovation surveys. The $%
13\times 13$ industry technology flow matrix ${\bf TM}$ is shown in Table
A.6.

{\scriptsize {\normalsize \newpage}\setlength{\oddsidemargin}{-0.5in}%
{\footnotesize $\fbox{$%
\begin{tabular}{lcccccccccc}
\multicolumn{11}{c}{Table A.1} \\ 
\multicolumn{11}{c}{Total Factor Productivity Index Summary Statistics} \\ 
\multicolumn{11}{c}{By Industry and By Country} \\ \hline\hline
\multicolumn{1}{l|}{ISIC} & \multicolumn{1}{c|}{1970} & \multicolumn{1}{c|}{
1980} & \multicolumn{1}{c|}{1991} & \multicolumn{1}{c||}{$
\begin{array}{c}
\text{Average } \\ 
\text{Annual } \\ 
\text{Growth} \\ 
\text{1970-91 (\%)}
\end{array}
$} &  & \multicolumn{1}{c|}{} & \multicolumn{1}{c|}{1970} & 
\multicolumn{1}{c|}{1980} & \multicolumn{1}{c|}{1991} & $
\begin{array}{c}
\text{Average } \\ 
\text{Annual } \\ 
\text{Growth} \\ 
\text{1970-91 (\%)}
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{31} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1056.6} & \multicolumn{1}{c|}{982.0} & \multicolumn{1}{c||}{1.0} &  & 
\multicolumn{1}{c|}{CAN} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
1837.4} & \multicolumn{1}{c|}{1990.3} & 2.0 \\ \hline
\multicolumn{1}{l|}{32} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1070.6} & \multicolumn{1}{c|}{1290.0} & \multicolumn{1}{c||}{2.3} &  & 
\multicolumn{1}{c|}{FRA} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
2405.1} & \multicolumn{1}{c|}{2871.7} & 3.8 \\ \hline
\multicolumn{1}{l|}{33} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1044.7} & \multicolumn{1}{c|}{1206.8} & \multicolumn{1}{c||}{2.0} &  & 
\multicolumn{1}{c|}{GER} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
1927.8} & \multicolumn{1}{c|}{2221.5} & 2.6 \\ \hline
\multicolumn{1}{l|}{34} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1004.0} & \multicolumn{1}{c|}{1138.0} & \multicolumn{1}{c||}{1.7} &  & 
\multicolumn{1}{c|}{IT} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
2058.0} & \multicolumn{1}{c|}{3372.0} & 4.5 \\ \hline
\multicolumn{1}{l|}{351/2} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1376.6} & \multicolumn{1}{c|}{1768.5} & \multicolumn{1}{c||}{3.8} &  & 
\multicolumn{1}{c|}{JAP} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
2398.4} & \multicolumn{1}{c|}{2892.5} & 3.8 \\ \hline
\multicolumn{1}{l|}{353/4} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
2014.2} & \multicolumn{1}{c|}{1958.3} & \multicolumn{1}{c||}{4.3} &  & 
\multicolumn{1}{c|}{SWE} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
1936.5} & \multicolumn{1}{c|}{1836.9} & 1.6 \\ \hline
\multicolumn{1}{l|}{355/6} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1158.4} & \multicolumn{1}{c|}{1346.4} & \multicolumn{1}{c||}{2.5} &  & 
\multicolumn{1}{c|}{UK} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
1513.4} & \multicolumn{1}{c|}{2220.0} & 2.5 \\ \hline
\multicolumn{1}{l|}{36} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1084.6} & \multicolumn{1}{c|}{1356.2} & \multicolumn{1}{c||}{2.5} &  & 
\multicolumn{1}{c|}{USA} & \multicolumn{1}{c|}{1300} & \multicolumn{1}{c|}{
1641.6} & \multicolumn{1}{c|}{1978.2} & 2.0 \\ 
\cline{1-5}\cline{7-11}\cline{6-7}
\multicolumn{1}{l|}{37} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1150.5} & \multicolumn{1}{c|}{1516.7} & \multicolumn{1}{c||}{3.0} &  & 
\multicolumn{5}{c}{} \\ \cline{1-5}
\multicolumn{1}{l|}{381} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1046.8} & \multicolumn{1}{c|}{1182.8} & \multicolumn{1}{c||}{1.9} &  & 
\multicolumn{5}{c}{} \\ \cline{1-5}
\multicolumn{1}{l|}{382/5} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1261.4} & \multicolumn{1}{c|}{1967.0} & \multicolumn{1}{c||}{4.3} &  & 
\multicolumn{5}{c}{} \\ \cline{1-5}
\multicolumn{1}{l|}{383} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1323.1} & \multicolumn{1}{c|}{2093.4} & \multicolumn{1}{c||}{4.6} &  & 
\multicolumn{5}{c}{} \\ \cline{1-5}
\multicolumn{1}{l|}{384} & \multicolumn{1}{c|}{800} & \multicolumn{1}{c|}{
1126.7} & \multicolumn{1}{c|}{1577.1} & \multicolumn{1}{c||}{3.2} &  & 
\multicolumn{5}{c}{}
\end{tabular}
$}$ } }

{\scriptsize {\normalsize \bigskip\bigskip\bigskip}{\tiny $\fbox{$%
\begin{tabular}{lccc|ccc|cccc}
\multicolumn{11}{c}{Table A.2} \\ 
\multicolumn{11}{c}{R\&D Stock Summary Statistics} \\ 
\multicolumn{11}{c}{By Industry and By Country; 1985 US \$} \\ \hline\hline
\multicolumn{1}{l|}{ISIC} & \multicolumn{1}{c|}{1970} & \multicolumn{1}{c|}{
1980} & 1991 & \multicolumn{1}{c||}{$
\begin{array}{c}
\text{Average } \\ 
\text{Annual } \\ 
\text{Growth} \\ 
\text{1970-91 (\%)}
\end{array}
$} &  &  & \multicolumn{1}{c|}{1970} & \multicolumn{1}{c|}{1980} & 
\multicolumn{1}{c|}{1991} & $
\begin{array}{c}
\text{Average } \\ 
\text{Annual } \\ 
\text{Growth} \\ 
\text{1970-91 (\%)}
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{31} & \multicolumn{1}{c|}{6390.7} & \multicolumn{1}{c|}{
15424.7} & 30092.5 & \multicolumn{1}{|c||}{7.4} &  & CAN & 
\multicolumn{1}{c|}{4930.9} & \multicolumn{1}{c|}{10435.5} & 
\multicolumn{1}{c|}{22820.6} & 7.3 \\ \hline
\multicolumn{1}{l|}{32} & \multicolumn{1}{c|}{4744.6} & \multicolumn{1}{c|}{
7482.3} & 9816.1 & \multicolumn{1}{|c||}{3.5} &  & FRA & \multicolumn{1}{c|}{
25216.9} & \multicolumn{1}{c|}{60913.3} & \multicolumn{1}{c|}{112246.8} & 7.1
\\ \hline
\multicolumn{1}{l|}{33} & \multicolumn{1}{c|}{1794.4} & \multicolumn{1}{c|}{
3211.0} & 4798.0 & \multicolumn{1}{|c||}{4.7} &  & GER & \multicolumn{1}{c|}{
41545.6} & \multicolumn{1}{c|}{98871.5} & \multicolumn{1}{c|}{193959.4} & 7.3
\\ \hline
\multicolumn{1}{l|}{34} & \multicolumn{1}{c|}{4112.3} & \multicolumn{1}{c|}{
9058.7} & 14966.6 & \multicolumn{1}{|c||}{6.2} &  & IT & \multicolumn{1}{c|}{
7807.3} & \multicolumn{1}{c|}{19329.5} & \multicolumn{1}{c|}{45193.6} & 8.4
\\ \hline
\multicolumn{1}{l|}{351/2} & \multicolumn{1}{c|}{55698.7} & 
\multicolumn{1}{c|}{133493.4} & 259920.6 & \multicolumn{1}{|c||}{7.3} &  & 
JAP & \multicolumn{1}{c|}{37341.0} & \multicolumn{1}{c|}{106730.8} & 
\multicolumn{1}{c|}{284083.3} & 9.7 \\ \hline
\multicolumn{1}{l|}{353/4} & \multicolumn{1}{c|}{11104.7} & 
\multicolumn{1}{c|}{22640.8} & 37347.5 & \multicolumn{1}{|c||}{5.8} &  & SWE
& \multicolumn{1}{c|}{6674.0} & \multicolumn{1}{c|}{15234.3} & 
\multicolumn{1}{c|}{25765.8} & 6.4 \\ \hline
\multicolumn{1}{l|}{355/6} & \multicolumn{1}{c|}{6757.1} & 
\multicolumn{1}{c|}{16073.8} & 29553.5 & \multicolumn{1}{|c||}{7.0} &  & UK
& \multicolumn{1}{|c|}{39067.6} & \multicolumn{1}{c|}{76971.6} & 
\multicolumn{1}{c|}{121302.8} & 5.4 \\ \hline
\multicolumn{1}{l|}{36} & \multicolumn{1}{c|}{5080.6} & \multicolumn{1}{c|}{
11319.5} & 23585.0 & \multicolumn{1}{|c||}{7.3} &  & USA & 
\multicolumn{1}{|c|}{248541.0} & \multicolumn{1}{c|}{517898.8} & 
\multicolumn{1}{c|}{950958.3} & 6.4 \\ \cline{1-5}\cline{7-11}\cline{6-7}
\multicolumn{1}{l|}{37} & \multicolumn{1}{c|}{13472.5} & \multicolumn{1}{c|}{
26591.0} & 41960.1 & \multicolumn{1}{|c||}{5.4} &  & \multicolumn{5}{c}{} \\ 
\cline{1-5}
\multicolumn{1}{l|}{381} & \multicolumn{1}{c|}{5115.6} & \multicolumn{1}{c|}{
11939.9} & 23450.6 & \multicolumn{1}{|c||}{7.3} &  & \multicolumn{5}{c}{} \\ 
\cline{1-5}
\multicolumn{1}{l|}{382/5} & \multicolumn{1}{c|}{48366.7} & 
\multicolumn{1}{c|}{131.561.1} & 303919.1 & \multicolumn{1}{|c||}{8.8} &  & 
\multicolumn{5}{c}{} \\ \cline{1-5}
\multicolumn{1}{l|}{383} & \multicolumn{1}{c|}{104071.0} & 
\multicolumn{1}{c|}{221154.1} & 425524.1 & \multicolumn{1}{|c||}{6.7} &  & 
\multicolumn{5}{c}{} \\ \cline{1-5}
\multicolumn{1}{l|}{384} & \multicolumn{1}{c|}{14415.7} & 
\multicolumn{1}{c|}{296435.0} & 551396.9 & \multicolumn{1}{|c||}{6.4} &  & 
\multicolumn{5}{c}{}
\end{tabular}
$}$ \ } }

{\scriptsize {\normalsize \bigskip\ \bigskip\ } }

{\scriptsize {\normalsize $\fbox{$%
\begin{tabular}{lcccccccc}
\multicolumn{9}{c}{Table A.3-1} \\ 
\multicolumn{9}{c}{Bilateral Import Shares in Food , Beverages \& Tobacco
Manufacturing (ISIC 31)} \\ 
\multicolumn{9}{c}{Average over 1972-1991; in per cent} \\ \hline\hline
\multicolumn{1}{l|}{from / to} & \multicolumn{1}{c|}{CAN} & 
\multicolumn{1}{c|}{FRA} & \multicolumn{1}{c|}{GER} & \multicolumn{1}{c|}{IT}
& \multicolumn{1}{c|}{JAP} & \multicolumn{1}{c|}{SWE} & \multicolumn{1}{c|}{
UK} & USA \\ \hline
\multicolumn{1}{l|}{CAN} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{1.93}
& \multicolumn{1}{c|}{1.4} & \multicolumn{1}{c|}{0.67} & \multicolumn{1}{c|}{
14.46} & \multicolumn{1}{c|}{3.87} & \multicolumn{1}{c|}{7.07} & 35.2 \\ 
\hline
\multicolumn{1}{l|}{FRA} & \multicolumn{1}{c|}{8.33} & \multicolumn{1}{c|}{0}
& \multicolumn{1}{c|}{48.95} & \multicolumn{1}{c|}{39.4} & 
\multicolumn{1}{c|}{6.6} & \multicolumn{1}{c|}{17.13} & \multicolumn{1}{c|}{
35.78} & 17.5 \\ \hline
\multicolumn{1}{l|}{GER} & \multicolumn{1}{c|}{3.37} & \multicolumn{1}{c|}{
36.69} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{45.97} & 
\multicolumn{1}{c|}{3.64} & \multicolumn{1}{c|}{30.24} & \multicolumn{1}{c|}{
25.09} & 9.24 \\ \hline
\multicolumn{1}{l|}{IT} & \multicolumn{1}{c|}{3.6} & \multicolumn{1}{c|}{
24.99} & \multicolumn{1}{c|}{22.96} & \multicolumn{1}{c|}{0} & 
\multicolumn{1}{c|}{0.98} & \multicolumn{1}{c|}{7.78} & \multicolumn{1}{c|}{
15.03} & 12.26 \\ \hline
\multicolumn{1}{l|}{JAP} & \multicolumn{1}{c|}{2.6} & \multicolumn{1}{c|}{
0.53} & \multicolumn{1}{c|}{1.14} & \multicolumn{1}{c|}{0.14} & 
\multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.57} & \multicolumn{1}{c|}{2.16
} & 5.84 \\ \hline
\multicolumn{1}{l|}{SWE} & \multicolumn{1}{c|}{2.0} & \multicolumn{1}{c|}{
0.66} & \multicolumn{1}{c|}{1.35} & \multicolumn{1}{c|}{0.89} & 
\multicolumn{1}{c|}{2.23} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{1.5}
& 1.54 \\ \hline
\multicolumn{1}{l|}{UK} & \multicolumn{1}{c|}{7.8} & \multicolumn{1}{c|}{
21.74} & \multicolumn{1}{c|}{8.94} & \multicolumn{1}{c|}{6.07} & 
\multicolumn{1}{c|}{7.53} & \multicolumn{1}{c|}{20.46} & \multicolumn{1}{c|}{
0} & 18.43 \\ \hline
\multicolumn{1}{l|}{USA} & \multicolumn{1}{c|}{74.63} & \multicolumn{1}{c|}{
13.46} & \multicolumn{1}{c|}{15.26} & \multicolumn{1}{c|}{6.86} & 
\multicolumn{1}{c|}{65.46} & \multicolumn{1}{c|}{19.95} & 
\multicolumn{1}{c|}{13.37} & 0
\end{tabular}
$}$ } }

{\scriptsize {\normalsize \bigskip\ \bigskip\ } }

{\scriptsize {\normalsize \ $\fbox{$%
\begin{tabular}{lcccccccc}
\multicolumn{9}{c}{Table A.3-2} \\ 
\multicolumn{9}{c}{Bilateral Import Shares in Transportation Equipment (ISIC
384)} \\ 
\multicolumn{9}{c}{Average over 1972-1991; in per cent} \\ \hline\hline
\multicolumn{1}{l|}{from / to} & \multicolumn{1}{c|}{CAN} & 
\multicolumn{1}{c|}{FRA} & \multicolumn{1}{c|}{GER} & \multicolumn{1}{c|}{IT}
& \multicolumn{1}{c|}{JAP} & \multicolumn{1}{c|}{SWE} & \multicolumn{1}{c|}{
UK} & USA \\ \hline
\multicolumn{1}{l|}{CAN} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.5}
& \multicolumn{1}{c|}{0.38} & \multicolumn{1}{c|}{0.37} & 
\multicolumn{1}{c|}{0.79} & \multicolumn{1}{c|}{0.83} & \multicolumn{1}{c|}{
1.02} & 38.22 \\ \hline
\multicolumn{1}{l|}{FRA} & \multicolumn{1}{c|}{0.74} & \multicolumn{1}{c|}{0}
& \multicolumn{1}{c|}{43.57} & \multicolumn{1}{c|}{34.79} & 
\multicolumn{1}{c|}{3.51} & \multicolumn{1}{c|}{8.25} & \multicolumn{1}{c|}{
18.52} & 3.41 \\ \hline
\multicolumn{1}{l|}{GER} & \multicolumn{1}{c|}{2.06} & \multicolumn{1}{c|}{
43.93} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{44.37} & 
\multicolumn{1}{c|}{23.77} & \multicolumn{1}{c|}{47.77} & 
\multicolumn{1}{c|}{40.94} & 12.44 \\ \hline
\multicolumn{1}{l|}{IT} & \multicolumn{1}{c|}{0.23} & \multicolumn{1}{c|}{
17.38} & \multicolumn{1}{c|}{13.31} & \multicolumn{1}{c|}{0} & 
\multicolumn{1}{c|}{1.96} & \multicolumn{1}{c|}{3.49} & \multicolumn{1}{c|}{
6.64} & 1.69 \\ \hline
\multicolumn{1}{l|}{JAP} & \multicolumn{1}{c|}{7.26} & \multicolumn{1}{c|}{
7.05} & \multicolumn{1}{c|}{15.04} & \multicolumn{1}{c|}{2.09} & 
\multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{13.94} & \multicolumn{1}{c|}{
13.85} & 37.26 \\ \hline
\multicolumn{1}{l|}{SWE} & \multicolumn{1}{c|}{0.53} & \multicolumn{1}{c|}{
2.95} & \multicolumn{1}{c|}{2.11} & \multicolumn{1}{c|}{2.02} & 
\multicolumn{1}{c|}{4.24} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{6.11
} & 2.29 \\ \hline
\multicolumn{1}{l|}{UK} & \multicolumn{1}{c|}{1.27} & \multicolumn{1}{c|}{
9.72} & \multicolumn{1}{c|}{12.77} & \multicolumn{1}{c|}{7.61} & 
\multicolumn{1}{c|}{6.59} & \multicolumn{1}{c|}{12.84} & \multicolumn{1}{c|}{
0} & 4.7 \\ \hline
\multicolumn{1}{l|}{USA} & \multicolumn{1}{c|}{87.9} & \multicolumn{1}{c|}{
18.47} & \multicolumn{1}{c|}{12.82} & \multicolumn{1}{c|}{8.75} & 
\multicolumn{1}{c|}{61.63} & \multicolumn{1}{c|}{12.88} & 
\multicolumn{1}{c|}{12.91} & 0
\end{tabular}
$}$}\bigskip\ \bigskip\ \bigskip\ }

{\scriptsize $\fbox{$%
\begin{tabular}{lccccc|cccccccc}
\multicolumn{14}{c}{Table A.4} \\ 
\multicolumn{14}{c}{Economy-wide Input-Output Relations; U.S. economy 1980;
13 ISIC industries; in percent} \\ 
\multicolumn{14}{c}{Shares of column industry's intermediate inputs from 12
partner industries; by row industry} \\ \hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{31} & \multicolumn{1}{c|}{32} & 
\multicolumn{1}{c|}{33} & \multicolumn{1}{c|}{34} & \multicolumn{1}{c|}{351/2
} & \multicolumn{1}{c|}{353/4} & \multicolumn{1}{c|}{355/6} & 
\multicolumn{1}{c|}{36} & \multicolumn{1}{c|}{37} & \multicolumn{1}{c|}{381}
& \multicolumn{1}{c|}{382/5} & \multicolumn{1}{c|}{383} & 384 \\ \hline
\multicolumn{1}{l|}{31} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.81}
& \multicolumn{1}{c|}{0.07} & \multicolumn{1}{c|}{0.32} & 
\multicolumn{1}{c|}{0.84} & \multicolumn{1}{c|}{0.10} & \multicolumn{1}{c|}{
0.02} & \multicolumn{1}{c|}{0.07} & \multicolumn{1}{c|}{0.01} & 
\multicolumn{1}{c|}{0.01} & \multicolumn{1}{c|}{0.06} & \multicolumn{1}{c|}{
0.01} & 0.01 \\ \hline
\multicolumn{1}{l|}{32} & \multicolumn{1}{c|}{0.04} & \multicolumn{1}{c|}{0}
& \multicolumn{1}{c|}{3.45} & \multicolumn{1}{c|}{1.02} & 
\multicolumn{1}{c|}{0.14} & \multicolumn{1}{c|}{0.02} & \multicolumn{1}{c|}{
4.71} & \multicolumn{1}{c|}{0.40} & \multicolumn{1}{c|}{0.04} & 
\multicolumn{1}{c|}{0.13} & \multicolumn{1}{c|}{0.67} & \multicolumn{1}{c|}{
0.10} & 2.07 \\ \hline
\multicolumn{1}{l|}{33} & \multicolumn{1}{c|}{0.05} & \multicolumn{1}{c|}{
0.12} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{4.80} & 
\multicolumn{1}{c|}{0.08} & \multicolumn{1}{c|}{0.03} & \multicolumn{1}{c|}{
0.41} & \multicolumn{1}{c|}{1.13} & \multicolumn{1}{c|}{0.28} & 
\multicolumn{1}{c|}{0.49} & \multicolumn{1}{c|}{0.39} & \multicolumn{1}{c|}{
0.75} & 1.27 \\ \hline
\multicolumn{1}{l|}{34} & \multicolumn{1}{c|}{4.78} & \multicolumn{1}{c|}{
1.22} & \multicolumn{1}{c|}{1.23} & \multicolumn{1}{c|}{0} & 
\multicolumn{1}{c|}{3.29} & \multicolumn{1}{c|}{0.27} & \multicolumn{1}{c|}{
3.77} & \multicolumn{1}{c|}{4.87} & \multicolumn{1}{c|}{0.26} & 
\multicolumn{1}{c|}{1.45} & \multicolumn{1}{c|}{1.51} & \multicolumn{1}{c|}{
1.76} & 0.31 \\ \hline
\multicolumn{1}{l|}{351/2} & \multicolumn{1}{c|}{1.74} & \multicolumn{1}{c|}{
14.96} & \multicolumn{1}{c|}{2.25} & \multicolumn{1}{c|}{7.52} & 
\multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{3.23} & \multicolumn{1}{c|}{
42.31} & \multicolumn{1}{c|}{9.91} & \multicolumn{1}{c|}{5.19} & 
\multicolumn{1}{c|}{2.04} & \multicolumn{1}{c|}{1.07} & \multicolumn{1}{c|}{
3.90} & 0.68 \\ \hline
\multicolumn{1}{l|}{353/4} & \multicolumn{1}{c|}{0.63} & \multicolumn{1}{c|}{
1.36} & \multicolumn{1}{c|}{3.58} & \multicolumn{1}{c|}{3.73} & 
\multicolumn{1}{c|}{4.10} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{1.57
} & \multicolumn{1}{c|}{2.86} & \multicolumn{1}{c|}{2.04} & 
\multicolumn{1}{c|}{2.04} & \multicolumn{1}{c|}{1.49} & \multicolumn{1}{c|}{
1.21} & 1.50 \\ \hline
\multicolumn{1}{l|}{355/6} & \multicolumn{1}{c|}{1.92} & \multicolumn{1}{c|}{
1.98} & \multicolumn{1}{c|}{2.54} & \multicolumn{1}{c|}{2.39} & 
\multicolumn{1}{c|}{3.35} & \multicolumn{1}{c|}{0.18} & \multicolumn{1}{c|}{0
} & \multicolumn{1}{c|}{1.03} & \multicolumn{1}{c|}{0.70} & 
\multicolumn{1}{c|}{1.49} & \multicolumn{1}{c|}{3.63} & \multicolumn{1}{c|}{
5.81} & 4.22 \\ \hline
\multicolumn{1}{l|}{36} & \multicolumn{1}{c|}{1.96} & \multicolumn{1}{c|}{
0.27} & \multicolumn{1}{c|}{0.82} & \multicolumn{1}{c|}{0.12} & 
\multicolumn{1}{c|}{0.83} & \multicolumn{1}{|c|}{0.24} & \multicolumn{1}{c|}{
1.28} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.74} & 
\multicolumn{1}{c|}{1.01} & \multicolumn{1}{c|}{1.26} & \multicolumn{1}{c|}{
1.76} & 1.23 \\ \hline
\multicolumn{1}{l|}{37} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.01}
& \multicolumn{1}{c|}{3.68} & \multicolumn{1}{c|}{0.31} & 
\multicolumn{1}{c|}{0.62} & \multicolumn{1}{|c|}{0.08} & \multicolumn{1}{c|}{
0.80} & \multicolumn{1}{c|}{1.00} & \multicolumn{1}{c|}{0} & 
\multicolumn{1}{c|}{43.16} & \multicolumn{1}{c|}{16.44} & 
\multicolumn{1}{c|}{11.76} & 10.51 \\ \hline
\multicolumn{1}{l|}{381} & \multicolumn{1}{c|}{4.68} & \multicolumn{1}{c|}{
0.12} & \multicolumn{1}{c|}{6.29} & \multicolumn{1}{c|}{0.83} & 
\multicolumn{1}{c|}{2.51} & \multicolumn{1}{|c|}{0.56} & \multicolumn{1}{c|}{
2.53} & \multicolumn{1}{c|}{1.74} & \multicolumn{1}{c|}{1.62} & 
\multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{8.20} & \multicolumn{1}{c|}{6.39
} & 10.67 \\ \hline
\multicolumn{1}{l|}{382/5} & \multicolumn{1}{c|}{0.26} & \multicolumn{1}{c|}{
0.92} & \multicolumn{1}{c|}{1.41} & \multicolumn{1}{c|}{1.18} & 
\multicolumn{1}{c|}{1.23} & \multicolumn{1}{|c|}{0.05} & \multicolumn{1}{c|}{
1.34} & \multicolumn{1}{c|}{1.63} & \multicolumn{1}{c|}{3.31} & 
\multicolumn{1}{c|}{4.51} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{2.78
} & 7.68 \\ \hline
\multicolumn{1}{l|}{383} & \multicolumn{1}{c|}{0.01} & \multicolumn{1}{c|}{
0.04} & \multicolumn{1}{c|}{0.18} & \multicolumn{1}{c|}{0.03} & 
\multicolumn{1}{c|}{0.05} & \multicolumn{1}{|c|}{0.01} & \multicolumn{1}{c|}{
0.34} & \multicolumn{1}{c|}{0.28} & \multicolumn{1}{c|}{0.91} & 
\multicolumn{1}{c|}{1.48} & \multicolumn{1}{c|}{10.58} & \multicolumn{1}{c|}{
0} & 4.67 \\ \hline
\multicolumn{1}{l|}{384} & \multicolumn{1}{c|}{0.01} & \multicolumn{1}{c|}{
0.03} & \multicolumn{1}{c|}{0.07} & \multicolumn{1}{c|}{0.08} & 
\multicolumn{1}{c|}{0.02} & \multicolumn{1}{|c|}{0.02} & \multicolumn{1}{c|}{
0.03} & \multicolumn{1}{c|}{0.04} & \multicolumn{1}{c|}{0.03} & 
\multicolumn{1}{c|}{0.08} & \multicolumn{1}{c|}{0.14} & \multicolumn{1}{c|}{
0.07} & 0
\end{tabular}
$}$}\bigskip\ \bigskip\ \bigskip\ 

{\scriptsize $\fbox{$%
\begin{tabular}{lccccc|cccccccc}
\multicolumn{14}{c}{Table A.5} \\ 
\multicolumn{14}{c}{Import Input-Output Relations; U.S. 1987; 13 ISIC
industries; in percent} \\ 
\multicolumn{14}{c}{Shares of column industry's intermediate inputs from 12
partner industries; by row industry} \\ \hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{31} & \multicolumn{1}{c|}{32} & 
\multicolumn{1}{c|}{33} & \multicolumn{1}{c|}{34} & 351/2 & 
\multicolumn{1}{c|}{353/4} & \multicolumn{1}{c|}{355/6} & 
\multicolumn{1}{c|}{36} & \multicolumn{1}{c|}{37} & \multicolumn{1}{c|}{381}
& \multicolumn{1}{c|}{382/5} & \multicolumn{1}{c|}{383} & 384 \\ \hline
\multicolumn{1}{l|}{31} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.45}
& \multicolumn{1}{c|}{0.01} & \multicolumn{1}{c|}{0.24} & 0.60 & 
\multicolumn{1}{c|}{0.03} & \multicolumn{1}{c|}{0.03} & \multicolumn{1}{c|}{
0.08} & \multicolumn{1}{c|}{0.01} & \multicolumn{1}{c|}{0} & 
\multicolumn{1}{c|}{0.05} & \multicolumn{1}{c|}{0.01} & 0 \\ \hline
\multicolumn{1}{l|}{32} & \multicolumn{1}{c|}{0.06} & \multicolumn{1}{c|}{0}
& \multicolumn{1}{c|}{7.77} & \multicolumn{1}{c|}{1.04} & 0.14 & 
\multicolumn{1}{c|}{0.03} & \multicolumn{1}{c|}{2.97} & \multicolumn{1}{c|}{
0.97} & \multicolumn{1}{c|}{0.04} & \multicolumn{1}{c|}{0.03} & 
\multicolumn{1}{c|}{0.28} & \multicolumn{1}{c|}{0.11} & 1.88 \\ \hline
\multicolumn{1}{l|}{33} & \multicolumn{1}{c|}{0.06} & \multicolumn{1}{c|}{
0.30} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.46} & 0.03 & 
\multicolumn{1}{c|}{0.04} & \multicolumn{1}{c|}{0.57} & \multicolumn{1}{c|}{
2.27} & \multicolumn{1}{c|}{0.23} & \multicolumn{1}{c|}{0.53} & 
\multicolumn{1}{c|}{0.23} & \multicolumn{1}{c|}{0.31} & 3.22 \\ \hline
\multicolumn{1}{l|}{34} & \multicolumn{1}{c|}{1.34} & \multicolumn{1}{c|}{
0.39} & \multicolumn{1}{c|}{0.29} & \multicolumn{1}{c|}{0} & 2.49 & 
\multicolumn{1}{c|}{0.01} & \multicolumn{1}{c|}{3.02} & \multicolumn{1}{c|}{
3.45} & \multicolumn{1}{c|}{0.04} & \multicolumn{1}{c|}{0.23} & 
\multicolumn{1}{c|}{0.27} & \multicolumn{1}{c|}{0.49} & 0.04 \\ \hline
\multicolumn{1}{l|}{351/2} & \multicolumn{1}{c|}{3.45} & \multicolumn{1}{c|}{
11.70} & \multicolumn{1}{c|}{3.31} & \multicolumn{1}{c|}{6.98} & 0 & 
\multicolumn{1}{c|}{2.04} & \multicolumn{1}{c|}{41.74} & \multicolumn{1}{c|}{
17.69} & \multicolumn{1}{c|}{4.17} & \multicolumn{1}{c|}{1.46} & 
\multicolumn{1}{c|}{0.62} & \multicolumn{1}{c|}{2.60} & 0.32 \\ \hline
\multicolumn{1}{l|}{353/4} & \multicolumn{1}{c|}{0.43} & \multicolumn{1}{c|}{
0.46} & \multicolumn{1}{c|}{1.29} & \multicolumn{1}{c|}{1.84} & 1.35 & 
\multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.82} & \multicolumn{1}{c|}{1.70
} & \multicolumn{1}{c|}{0.89} & \multicolumn{1}{c|}{1.52} & 
\multicolumn{1}{c|}{0.21} & \multicolumn{1}{c|}{0.30} & 0.17 \\ \hline
\multicolumn{1}{l|}{355/6} & \multicolumn{1}{c|}{2.01} & \multicolumn{1}{c|}{
2.56} & \multicolumn{1}{c|}{7.16} & \multicolumn{1}{c|}{2.24} & 2.38 & 
\multicolumn{1}{c|}{0.20} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{0.41
} & \multicolumn{1}{c|}{0.54} & \multicolumn{1}{c|}{3.97} & 
\multicolumn{1}{c|}{2.84} & \multicolumn{1}{c|}{3.64} & 2.99 \\ \hline
\multicolumn{1}{l|}{36} & \multicolumn{1}{c|}{0.41} & \multicolumn{1}{c|}{
0.11} & \multicolumn{1}{c|}{1.84} & \multicolumn{1}{c|}{0.05} & 0.76 & 
\multicolumn{1}{|c|}{0.03} & \multicolumn{1}{c|}{1.15} & \multicolumn{1}{c|}{
0} & \multicolumn{1}{c|}{1.31} & \multicolumn{1}{c|}{1.36} & 
\multicolumn{1}{c|}{0.73} & \multicolumn{1}{c|}{1.72} & 0.91 \\ \hline
\multicolumn{1}{l|}{37} & \multicolumn{1}{c|}{0.01} & \multicolumn{1}{c|}{
0.01} & \multicolumn{1}{c|}{6.53} & \multicolumn{1}{c|}{0.26} & 0.25 & 
\multicolumn{1}{|c|}{0.01} & \multicolumn{1}{c|}{1.30} & \multicolumn{1}{c|}{
3.76} & \multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{67.32} & 
\multicolumn{1}{c|}{8.48} & \multicolumn{1}{c|}{14.32} & 3.82 \\ \hline
\multicolumn{1}{l|}{381} & \multicolumn{1}{c|}{1.34} & \multicolumn{1}{c|}{
0.09} & \multicolumn{1}{c|}{11.01} & \multicolumn{1}{c|}{1.03} & 2.54 & 
\multicolumn{1}{|c|}{0.34} & \multicolumn{1}{c|}{5.35} & \multicolumn{1}{c|}{
1.24} & \multicolumn{1}{c|}{1.50} & \multicolumn{1}{c|}{0} & 
\multicolumn{1}{c|}{4.71} & \multicolumn{1}{c|}{3.68} & 6.48 \\ \hline
\multicolumn{1}{l|}{382/5} & \multicolumn{1}{c|}{0.63} & \multicolumn{1}{c|}{
3.77} & \multicolumn{1}{c|}{2.50} & \multicolumn{1}{c|}{4.36} & 2.30 & 
\multicolumn{1}{|c|}{0.04} & \multicolumn{1}{c|}{3.44} & \multicolumn{1}{c|}{
1.74} & \multicolumn{1}{c|}{4.29} & \multicolumn{1}{c|}{4.67} & 
\multicolumn{1}{c|}{0} & \multicolumn{1}{c|}{6.37} & 7.27 \\ \hline
\multicolumn{1}{l|}{383} & \multicolumn{1}{c|}{0.02} & \multicolumn{1}{c|}{
0.05} & \multicolumn{1}{c|}{2.42} & \multicolumn{1}{c|}{0.16} & 0.09 & 
\multicolumn{1}{|c|}{0.01} & \multicolumn{1}{c|}{1.67} & \multicolumn{1}{c|}{
1.16} & \multicolumn{1}{c|}{1.83} & \multicolumn{1}{c|}{2.51} & 
\multicolumn{1}{c|}{23.79} & \multicolumn{1}{c|}{0} & 12.32 \\ \hline
\multicolumn{1}{l|}{384} & \multicolumn{1}{c|}{0.04} & \multicolumn{1}{c|}{
0.03} & \multicolumn{1}{c|}{1.14} & \multicolumn{1}{c|}{0.17} & 0.03 & 
\multicolumn{1}{|c|}{0.03} & \multicolumn{1}{c|}{0.08} & \multicolumn{1}{c|}{
0.14} & \multicolumn{1}{c|}{0.05} & \multicolumn{1}{c|}{0.17} & 
\multicolumn{1}{c|}{0.08} & \multicolumn{1}{c|}{0.04} & 0
\end{tabular}
$}$}\bigskip\ \bigskip\ \bigskip\ 

{\scriptsize \ $\fbox{$%
\begin{tabular}{lccccc|cccccccc}
\multicolumn{14}{c}{Table A.6} \\ 
\multicolumn{14}{c}{Technology Flow Matrix} \\ 
\multicolumn{14}{c}{Per cent of row industry going to column industry} \\ 
\hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{31} & \multicolumn{1}{c|}{32} & 
\multicolumn{1}{c|}{33} & \multicolumn{1}{c|}{34} & 351/2 & 
\multicolumn{1}{c|}{353/4} & \multicolumn{1}{c|}{355/6} & 
\multicolumn{1}{c|}{36} & \multicolumn{1}{c|}{37} & \multicolumn{1}{c|}{381}
& \multicolumn{1}{c|}{382/5} & \multicolumn{1}{c|}{383} & 384 \\ \hline
\multicolumn{1}{l|}{31} & \multicolumn{1}{c|}{97.6} & \multicolumn{1}{c|}{0.3
} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.3} & 1.2 & 
\multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.1} & \multicolumn{1}{c|}{0.0
} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.4} & 
\multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.0} & 0.0 \\ \hline
\multicolumn{1}{l|}{32} & \multicolumn{1}{c|}{0.7} & \multicolumn{1}{c|}{66.8
} & \multicolumn{1}{c|}{1.1} & \multicolumn{1}{c|}{2.5} & 0.9 & 
\multicolumn{1}{c|}{0.2} & \multicolumn{1}{c|}{1.6} & \multicolumn{1}{c|}{1.0
} & \multicolumn{1}{c|}{0.1} & \multicolumn{1}{c|}{0.5} & 
\multicolumn{1}{c|}{3.7} & \multicolumn{1}{c|}{17.7} & 3.1 \\ \hline
\multicolumn{1}{l|}{33} & \multicolumn{1}{c|}{1.7} & \multicolumn{1}{c|}{0.6}
& \multicolumn{1}{c|}{82.6} & \multicolumn{1}{c|}{0.6} & 0.6 & 
\multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.0
} & \multicolumn{1}{c|}{0.2} & \multicolumn{1}{c|}{0.4} & 
\multicolumn{1}{c|}{3.3} & \multicolumn{1}{c|}{2.3} & 7.7 \\ \hline
\multicolumn{1}{l|}{34} & \multicolumn{1}{c|}{26.0} & \multicolumn{1}{c|}{2.7
} & \multicolumn{1}{c|}{1.0} & \multicolumn{1}{c|}{57.7} & 2.6 & 
\multicolumn{1}{c|}{0.3} & \multicolumn{1}{c|}{1.4} & \multicolumn{1}{c|}{0.5
} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{1.3} & 
\multicolumn{1}{c|}{4.0} & \multicolumn{1}{c|}{2.1} & 0.2 \\ \hline
\multicolumn{1}{l|}{351/2} & \multicolumn{1}{c|}{1.6} & \multicolumn{1}{c|}{
2.8} & \multicolumn{1}{c|}{0.6} & \multicolumn{1}{c|}{2.6} & 70.3 & 
\multicolumn{1}{c|}{2.1} & \multicolumn{1}{c|}{9.0} & \multicolumn{1}{c|}{0.7
} & \multicolumn{1}{c|}{0.7} & \multicolumn{1}{c|}{3.3} & 
\multicolumn{1}{c|}{3.2} & \multicolumn{1}{c|}{2.2} & 0.9 \\ \hline
\multicolumn{1}{l|}{353/4} & \multicolumn{1}{c|}{0.3} & \multicolumn{1}{c|}{
2.0} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{1.0} & 4.3 & 
\multicolumn{1}{c|}{35.8} & \multicolumn{1}{c|}{0.8} & \multicolumn{1}{c|}{
0.8} & \multicolumn{1}{c|}{3.0} & \multicolumn{1}{c|}{5.5} & 
\multicolumn{1}{c|}{17.8} & \multicolumn{1}{c|}{2.8} & 26.1 \\ \hline
\multicolumn{1}{l|}{355/6} & \multicolumn{1}{c|}{7.8} & \multicolumn{1}{c|}{
2.2} & \multicolumn{1}{c|}{1.8} & \multicolumn{1}{c|}{1.9} & 7.3 & 
\multicolumn{1}{c|}{0.3} & \multicolumn{1}{c|}{49.1} & \multicolumn{1}{c|}{
1.0} & \multicolumn{1}{c|}{0.2} & \multicolumn{1}{c|}{3.1} & 
\multicolumn{1}{c|}{5.4} & \multicolumn{1}{c|}{3.7} & 16.1 \\ \hline
\multicolumn{1}{l|}{36} & \multicolumn{1}{c|}{0.9} & \multicolumn{1}{c|}{1.0}
& \multicolumn{1}{c|}{0.5} & \multicolumn{1}{c|}{0.2} & 2.1 & 
\multicolumn{1}{|c|}{0.6} & \multicolumn{1}{c|}{0.4} & \multicolumn{1}{c|}{
59.1} & \multicolumn{1}{c|}{7.7} & \multicolumn{1}{c|}{5.2} & 
\multicolumn{1}{c|}{7.3} & \multicolumn{1}{c|}{11.9} & 3.2 \\ \hline
\multicolumn{1}{l|}{37} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.2}
& \multicolumn{1}{c|}{0.3} & \multicolumn{1}{c|}{0.2} & 2.6 & 
\multicolumn{1}{|c|}{0.4} & \multicolumn{1}{c|}{0.4} & \multicolumn{1}{c|}{
0.2} & \multicolumn{1}{c|}{47.0} & \multicolumn{1}{c|}{19.9} & 
\multicolumn{1}{c|}{10.5} & \multicolumn{1}{c|}{10.5} & 7.7 \\ \hline
\multicolumn{1}{l|}{381} & \multicolumn{1}{c|}{2.2} & \multicolumn{1}{c|}{1.4
} & \multicolumn{1}{c|}{4.8} & \multicolumn{1}{c|}{0.6} & 1.5 & 
\multicolumn{1}{|c|}{0.7} & \multicolumn{1}{c|}{2.6} & \multicolumn{1}{c|}{
1.6} & \multicolumn{1}{c|}{3.3} & \multicolumn{1}{c|}{42.1} & 
\multicolumn{1}{c|}{29.4} & \multicolumn{1}{c|}{3.2} & 6.4 \\ \hline
\multicolumn{1}{l|}{382/5} & \multicolumn{1}{c|}{4.5} & \multicolumn{1}{c|}{
3.2} & \multicolumn{1}{c|}{1.4} & \multicolumn{1}{c|}{7.4} & 5.3 & 
\multicolumn{1}{|c|}{1.9} & \multicolumn{1}{c|}{4.9} & \multicolumn{1}{c|}{
2.6} & \multicolumn{1}{c|}{6.0} & \multicolumn{1}{c|}{5.5} & 
\multicolumn{1}{c|}{49.1} & \multicolumn{1}{c|}{4.1} & 4.0 \\ \hline
\multicolumn{1}{l|}{383} & \multicolumn{1}{c|}{0.1} & \multicolumn{1}{c|}{0.1
} & \multicolumn{1}{c|}{0.1} & \multicolumn{1}{c|}{0.2} & 0.5 & 
\multicolumn{1}{|c|}{0.0} & \multicolumn{1}{c|}{0.1} & \multicolumn{1}{c|}{
0.1} & \multicolumn{1}{c|}{0.6} & \multicolumn{1}{c|}{1.5} & 
\multicolumn{1}{c|}{13.1} & \multicolumn{1}{c|}{79.5} & 3.9 \\ \hline
\multicolumn{1}{l|}{384} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.0
} & \multicolumn{1}{c|}{0.1} & \multicolumn{1}{c|}{0.0} & 0.1 & 
\multicolumn{1}{|c|}{0.0} & \multicolumn{1}{c|}{0.1} & \multicolumn{1}{c|}{
0.0} & \multicolumn{1}{c|}{0.0} & \multicolumn{1}{c|}{0.0} & 
\multicolumn{1}{c|}{0.7} & \multicolumn{1}{c|}{0.0} & 98.9
\end{tabular}
$}$}\bigskip\ \bigskip\ \bigskip\ {\scriptsize {\normalsize \ \ }}%
{\footnotesize $\fbox{$%
\begin{tabular}{lcccc}
\multicolumn{5}{c}{Table A.7-1} \\ 
\multicolumn{5}{c}{Sensitivity Analysis: Timing and Knowledge Depreciation
Rate} \\ 
\multicolumn{5}{c}{All Industries; Dependent Variable: Log of TFP index} \\ 
\hline
\multicolumn{5}{c}{All regressors} \\ \hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{$
\begin{array}{c}
\text{Benchmark} \\ 
\text{as in text}
\end{array}
$} & \multicolumn{1}{c|}{Timing} & \multicolumn{2}{|c}{Knowledge
Depreciation Rate} \\ \hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{(T2.4)} & \multicolumn{1}{c|}{$
\begin{array}{c}
\text{Foreign R\&D (}b^f,b^{f,tm}\text{)} \\ 
\text{lagged by one period}
\end{array}
$} & $\delta =0$ & $\delta =0.2$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\beta _1\text{: Same Sector, } \\ 
\text{Domestic R\&D (}b\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.102^{**} \\ 
(0.043)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.097^{**} \\ 
(0.045)
\end{array}
$} & $
\begin{array}{c}
0.152^{**} \\ 
(0.069)
\end{array}
$ & $
\begin{array}{c}
0.109^{**} \\ 
(0.051)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\beta _2\text{: Other Sector,} \\ 
\text{Domestic R\&D (}b^{tm}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
1.235^{**} \\ 
(0.386)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.655^{**} \\ 
(0.354)
\end{array}
$} & $
\begin{array}{c}
1.024^{**} \\ 
(0.364)
\end{array}
$ & $
\begin{array}{c}
0.323 \\ 
(0.354)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\beta _3\text{: Same Sector,} \\ 
\text{Foreign R\&D (}b^f\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.848^{**} \\ 
(0.216)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.812^{**} \\ 
(0.218)
\end{array}
$} & $
\begin{array}{c}
1.034^{**} \\ 
(0.245)
\end{array}
$ & $
\begin{array}{c}
0.956^{**} \\ 
(0.312)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\beta _4\text{: Other Sector,} \\ 
\text{Foreign R\&D (}b^{f,tm}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
-1.808^{**} \\ 
(0.498)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
-1.294^{**} \\ 
(0.516)
\end{array}
$} & $
\begin{array}{c}
-1.733^{**} \\ 
(0.549)
\end{array}
$ & $
\begin{array}{c}
-0.902 \\ 
(0.598)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\text{F-statistic} \\ 
(\text{Degr. of freedom)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
24.61 \\ 
(4,2283)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
20.70 \\ 
(4,2179)
\end{array}
$} & $
\begin{array}{c}
23.25 \\ 
(4,2283)
\end{array}
$ & $
\begin{array}{c}
21.47 \\ 
(4,2283)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\text{Number of} \\ 
\text{Observations}
\end{array}
$} & \multicolumn{1}{c|}{2288} & \multicolumn{1}{c|}{2184} & 2288 & 2288
\end{tabular}
$}$\bigskip\ \bigskip\ \bigskip\  }\ {\footnotesize $\fbox{$%
\begin{tabular}{lcccc}
\multicolumn{5}{c}{Table A.7-2} \\ 
\multicolumn{5}{c}{Sensitivity Analysis: Timing and Knowledge Depreciation
Rate} \\ \hline
\multicolumn{5}{c}{All Industries; Dependent Variable: Log of TFP index;
variable $b^{f,tm}$ excluded} \\ \hline
\multicolumn{5}{c}{Variable $b^{f,tm}$ excluded} \\ \hline\hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{$
\begin{array}{c}
\text{Benchmark} \\ 
\text{as in text}
\end{array}
$} & \multicolumn{1}{c|}{Timing} & \multicolumn{2}{|c}{$
\begin{array}{c}
\text{Knowledge} \\ 
\text{Depreciation Rate}
\end{array}
$} \\ \hline
\multicolumn{1}{l|}{} & \multicolumn{1}{c|}{(T2.3)} & \multicolumn{1}{c|}{$
\begin{array}{c}
\text{Foreign R\&D (}b^f,b^{f,tm}\text{)} \\ 
\text{lagged by one period}
\end{array}
$} & $\delta =0$ & $\delta =0.2$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\beta _1\text{: Same Sector, } \\ 
\text{Domestic R\&D (}b\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.171^{**} \\ 
(0.033)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.150^{**} \\ 
(0.035)
\end{array}
$} & $
\begin{array}{c}
0.233^{**} \\ 
(0.049)
\end{array}
$ & $
\begin{array}{c}
0.137^{**} \\ 
(0.034)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\beta _2\text{: Other Sector,} \\ 
\text{Domestic R\&D (}b^{tm}\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.551 \\ 
(0.417)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.235 \\ 
(0.428)
\end{array}
$} & $
\begin{array}{c}
0.691^{*} \\ 
(0.419)
\end{array}
$ & $
\begin{array}{c}
-0.7\times 10^{-3} \\ 
(0.415)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\beta _3\text{: Same Sector,} \\ 
\text{Foreign R\&D (}b^f\text{)} \\ 
\text{(s.e.)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.952^{**} \\ 
(0.259)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
0.959^{**} \\ 
(0.277)
\end{array}
$} & $
\begin{array}{c}
0.699^{**} \\ 
(0.260)
\end{array}
$ & $
\begin{array}{c}
0.947^{**} \\ 
(0.263)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\text{F-statistic} \\ 
(\text{Degr. of freedom)}
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
24.33 \\ 
(3,2284)
\end{array}
$} & \multicolumn{1}{c|}{$
\begin{array}{c}
23.00 \\ 
(3,2180)
\end{array}
$} & $
\begin{array}{c}
20.51 \\ 
(3,2284)
\end{array}
$ & $
\begin{array}{c}
25.02 \\ 
(3,2284)
\end{array}
$ \\ \hline
\multicolumn{1}{l|}{$
\begin{array}{c}
\text{Number of} \\ 
\text{Observations}
\end{array}
$} & \multicolumn{1}{c|}{2288} & \multicolumn{1}{c|}{2184} & 2288 & 2288
\end{tabular}
$}$ }

\end{document}