0% found this document useful (0 votes)
18 views28 pages

Rdad 005

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
18 views28 pages

Rdad 005

Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 28

Review of Economic Studies (2023) 90, 3315–3342 https://doi.org/10.

1093/restud/rdad005
© The Author(s) 2023. Published by Oxford University Press on behalf of The Review of Economic Studies Limited.
Advance access publication 25 January 2023

Estimating Production Functions


of Multiproduct Firms

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


NELLI VALMARI
ETLA Economic Research

First version received May 2016; Editorial decision November 2022; Accepted January 2023 (Eds.)

Multiproduct firms constitute a considerable share of firms and account for an even greater share of
production. Nevertheless, the vast majority of production function estimates are based on the assumption
that firms are single-product manufacturers. This assumption is due to a lack of data on how firms allocate
their inputs across their various product lines. I provide a strategy for estimating product-specific input
allocations and production functions of multiproduct firms involved in monopolistic competition. The
strategy is based on using firms’ product prices and output demand in solving for product-level input
allocations.
Key words: Multiproduct firm, Production function, Productivity

JEL codes: D24, L11, L25

1. INTRODUCTION
Firms’ production functions are estimated in order to identify important values like marginal
costs, returns to scale, and productivity differences between firms. These attributes are essential
in evaluating how changes in, for example, competition and regulation affect market outcomes.
Production function estimates are typically based on the often implicit assumption that the firm
produces all of its output with a firm-level production technology that is independent of the
firm’s product set and that the firm is equally productive in manufacturing all of its products.
If a firm produces a range of products, however, it is possible that the firm uses multiple pro-
duction technologies, in addition to facing distinct demand for these products. Recognising this
is important because a substantial share of firms are multiproduct firms, and an even greater
share of products are provided by these firms. For example, in the US manufacturing sector
from 1987 to 1997, 39% of firms manufactured more than one product, and these multiprod-
uct firms accounted for 87% of the sector’s output (Bernard et al., 2010). In international trade,
multiproduct firms are even more prominent: they accounted for more than 99% of US exports
in 2000 (Bernard et al., 2007). Moreover, empirical findings in the international trade literature
suggest that multiproduct firms have productivity differences across product lines.1 The reason
for assuming a firm-level production technology that is independent of the product set is simply

1. Reductions in barriers to international trade are shown to raise firms’ productivity and reduce their product
scopes, i.e. the number of products produced, presumably because firms drop the least productive products from their
selections of exported products (e.g. Baldwin et al., 2005; Bernard et al., 2011; Mayer et al., 2014).

The editor in charge of this paper was Aureo de Paula.

3315
3316 REVIEW OF ECONOMIC STUDIES

pragmatic: in a typical data set of a cross-section of firms, the input allocation across the firm’s
product lines is unobservable.
In this paper, I present a strategy for estimating product-level production functions of firms
that are multiproduct producers and face monopolistic competition in the output market. Pro-
duction functions may vary across products, and there may be productivity differences across
products within firms. As usual, inputs are observed only at the firm or establishment level. The
challenge is to solve for the product-level inputs, which are functions of, among other factors, the
unobservable productivity levels. The key to the estimation strategy is that for firms that are not

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


price-takers in the output markets, the products’ output demand may be utilised to estimate the
optimal product-level input allocations. The estimation strategy requires data on firms’ product
prices, which fortunately are available in several data sets used in the literature.2 My strategy for
estimating product-specific production functions3 builds on the structural production function
literature that includes Olley and Pakes (1996), Levinsohn and Petrin (2003), Doraszelski and
Jaumandreu (2013), Ackerberg et al. (2015), and Gandhi et al. (2020). I show that, under the two
model assumptions presented in this section, product-level inputs can be solved parametrically
from the firm’s static profit maximisation problem as functions of the observable data and the
parameters to be estimated. While within-firm differences in output demand4 and demand shocks
provide a source for identifying variation, supply- and demand-side instruments can be used to
identify the production function parameters. The only demand function parameters that need to
be estimated are the own-price elasticities of demand. I demonstrate the estimation strategy by
estimating production functions for products in the wood industry and find that the production
technologies are statistically different across products.
In general, there are several reasons why an applied economist may want to allow for product-
specific production technologies instead of a firm-level technology that is independent of the
product set the firm produces. First, products may differ in how intensive their production pro-
cesses are in terms of inputs such as labour. Suppose a firm-level Cobb–Douglas production
function is estimated, whereas the production technologies are in fact product-specific. In that
case, the estimated output elasticity of, say, labour is a linear combination of all the product-
specific output elasticities of labour and of all the product-specific output elasticities of other
inputs such as capital and materials (Mundlak, 1963). Second, returns to scale may take place at
the product level instead of the firm level and may vary across products. Third, there may be pro-
ductivity differences across products but within firms. Fourth, there may be economies of scope
due to producing multiple products. In other words, when a firm-level production technology is
assumed, even if some firms produce a variety of products, the estimated production function is
likely to be misspecified. Consequently, the production function parameter estimates may not be
consistent. Furthermore, the inferences on returns to scale, productivity, and marginal costs may
be false. An additional advantage of physical product-specific production function estimates is
that they enable computing products’ price-cost markups.

2. Other data sets reporting outputs and revenues at the product-firm or product-plant levels are available from,
for example, Belgium (Statistics on the Production of Manufactured Goods by Statistics Belgium), Colombia (Annual
Manufacturers Survey by the National Administrative Department of Statistics of Colombia), India (Prowess Data by
the Centre for Monitoring the Indian Economy), and the United States (Longitudinal Research Database by the U.S.
Census Bureau).
3. In this paper, I estimate Cobb–Douglas production functions but, as shown in Supplementary Material,
Appendix A, the identification strategy is also compatible with the translog production function.
4. The demand function estimated in this paper is isoelastic, but also other functional forms can be used, as
shown in Supplementary Material, Appendix A.
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3317

The first assumption I make in order to solve for the product-level input allocation concerns
economies of scope and flexible manufacturing. Economies of scope are obtained when simulta-
neous manufacturing of multiple products results in lower fixed or variable costs of production.
They are a potential reason for firms to produce multiple products (e.g. Panzar, 1989). An alter-
native supply-side explanation for the existence of multiproduct firms is that firms can add new
products to their product assortments without making large investments in production technol-
ogy, even if the product-specific variable costs of the new products are higher. This is referred
to as flexible manufacturing, and it is assumed in several theoretical studies concerning multi-

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


product firms (e.g. Eckel and Neary, 2010; Mayer et al., 2014). Flexible manufacturing does not
entail economies of scope in the form of variable costs, but some theoretical models allow for
economies of scope in the form of lower fixed costs.5 Instead, an essential feature of flexible
manufacturing is the capability of a multiproduct firm to produce one or a few of its products
more efficiently than the rest of its products; these are referred to as its core competency6 (e.g.
Eckel and Neary, 2010).7 The estimation strategy of this paper is based on the assumption that
there are no economies of scope, but firms may have core competencies in producing some of
their products, as postulated in the theory of flexible manufacturing.8 Therefore, the estimation
strategy is not suitable for industries where products’ production technologies are joint because
product lines have joint inputs or by-products of a product line are used as inputs in another
product line.9
The second assumption I make in order to solve for product-level inputs is that capital and
labour have already been chosen in the previous period and can be costlessly reallocated across
product lines in response to demand and productivity shocks. This assumption is most likely to
hold for data on plants producing products in a single industry, as opposed to conglomerates or
multiplant firms observed only at the firm level. The reason for assuming labour to be predeter-
mined is the institutional features of the Finnish labour market. However, the estimation strategy
can, under certain conditions, also be applied when the labour input is flexible.
The production functions are estimated as physical production functions; i.e. the output is
measured in physical quantities instead of value added or gross output value. By doing so, the
omitted price bias discussed by Klette and Griliches (1996) can be avoided. The omitted price
bias arises when output is measured in value added or gross output value, while output prices and
input choices are correlated due to firms’ pricing power. If this correlation and demand shocks
are not taken into account, estimates of the production function parameters and returns to scale
are likely to be biased (De Loecker, 2011).10 Another advantage of estimating physical pro-
duction functions is that physical productivity changes can be disentangled from price changes,
which is not possible when estimating revenue production functions with conventional methods
(Garcia-Marin and Voigtlander, 2019). In order to estimate physical production functions, the

5. For example, Eckel and Neary (2010) and Mayer et al. (2014) assume economies of scope in the form of lower
fixed costs, whereas Qiu and Zhou (2013) and Flach and Irlacher (2018) do not.
6. “Core competency” has an alternative definition in the multiproduct firm literature: it is the most profitable
activity of a firm (Bernard et al., 2011).
7. There are empirical findings consistent with firms having core competencies in producing some of their
products (Arkolakis et al., 2021; Dhyne et al., 2021).
8. However, the estimation strategy may be adjusted to allow for the number of products to have an effect on the
unobservable product-specific productivity level, as in De Loecker et al. (2016).
9. In other words, the firm’s production correspondence is additively separable in the production functions for
each product.
10. An additional problem arises when firm-level revenue production functions of multiproduct firms are esti-
mated. As the composite output (i.e. the product-level outputs aggregated to the firm level) depends on the products’
relative prices, so do the elasticity estimates of the firm-level function, even when controlled for the endogeneity between
product prices and input choices (Mundlak, 1963).
3318 REVIEW OF ECONOMIC STUDIES

output quantities have to be observable and, for any given product, comparable across manufac-
turers. This data requirement is a matter of how heterogeneous the products in the industry are
and, eventually, how detailed the product classification is.
Another data requirement relates to dealing with product selection for which the production
function estimation strategy does not account. It is well known in the literature that firm produc-
tivity and capital are correlated among the observations some distance above the market entry
threshold, which may give rise to selection bias (e.g. Wedervang, 1965). This kind of selection
may take place also at the product level. To avoid selection bias, one needs a data set with a low

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


probability of firm entry and exit and a low probability of change in the firm’s product assort-
ment. To this end, data on large, established firms with little variation in product assortment are
desirable.
The present contributes to the literature in which multiproduct firms’ product-dependent pro-
duction technologies are estimated. Early research considers how product-specific production
functions can be aggregated to the firm level (e.g. Mundlak, 1963, 1964), and how marginal
costs can be obtained by estimating joint cost functions (e.g. Brown et al., 1979; Caves et al.,
1980). More recently, product-level production functions have been estimated using the assump-
tion that firm-level inputs are allocated across products in proportion to products’ revenue shares
(e.g. Foster et al., 2008; Collard-Wexler and Loecker, 2015) or in proportion to the number of
products (De Loecker, 2011).
Some of the most recent methods for estimating multiproduct production technologies nei-
ther make any assumptions on input allocation nor estimate the allocation. The technological
relationship between individual outputs along the firm’s production possibilities frontier has
been estimated by Malikov and Lien (2021). A transformation function where the output quan-
tity of a given product is related to the firm-level inputs and the output quantities of the other
products the firm produces has been estimated by Dhyne et al. (2020, 2021). As the output
quantities of the other products are endogenous to the productivity shock of a given product,
Dhyne et al. (2020, 2021) instrument them with lagged values of the output quantities or inputs
lagged even further back. Their estimation methodology is suitable for estimating production
technologies for the joint production of products.
Product-specific production functions have been estimated by De Loecker et al. (2016). The
purpose of the estimation is to evaluate how marginal costs and price cost markups respond
to trade liberalisation. De Loecker et al. use data on single-product firms and the estimation
strategy of Ackerberg et al. (2015) to estimate product-specific production function parameters
which are assumed to be the same for single- and multiproduct firms. This enables De Loecker
et al. to estimate the parameters without simultaneously solving for the unobservable input allo-
cations. They assume that output quantities are set in the period before production and that
flexible inputs are set to minimise costs, conditional on the chosen output quantities and firm-
level predetermined inputs. All the predetermined and flexible inputs are allocated to product
lines such that the input expenditure shares are constant across the inputs for any given product.
Economies of scope may exist, but productivity differences between product lines are assumed
away. De Loecker et al. show that both cost efficiency and profitability vary across the different
products a firm produces.
Another solution for estimating product-level production functions is provided by Orr (2022).
In contrast to De Loecker et al. (2016), who make no assumptions concerning output demand or
market structure, and more similar to the present study, Orr uses information from output demand
in solving for the unobservable product-level inputs. His methodology focuses on uncovering
within-firm productivity dispersion while assuming that production function parameters do not
vary across industries. Orr first estimates multiproduct firms’ input allocations using estimates of
the shape of the firm’s demand function. He then estimates product-level production functions
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3319

using the estimated input allocations. Orr finds that plants in machinery manufacturing have
sizeable within-plant variation in efficiency.
The estimation strategy of the present study differs from the strategies discussed above
because it enables estimation of non-joint and product-specific production functions with pro-
ductivity differences across products, while the unobservable product-level inputs are solved
accordingly. This requires, in contrast to the methodologies discussed above, simultaneous
estimation of product-level input allocations and production functions.
The model and the estimation strategy are presented in Sections 2 and 3. In Section 4, I intro-

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


duce the data set and provide further details of the estimation procedure. Empirical results are
presented in Section 5. In Section 6, I discuss the key identifying assumptions and data require-
ments of the estimation strategy, how they relate to the current production function literature,
and the kinds of industries for which the estimation strategy is suitable. Section 7 concludes.

2. THE MODEL
The model consists of product-specific production and demand functions and of assumptions
about the timing of production decisions. Firm-level production functions are typically estimated
without considering demand for the products. Estimation of product-specific production func-
tions, however, requires solving for the unobservable product-level input allocations, which in
turn requires the econometrician to specify the output demand functions.11 In demonstrating the
estimation strategy, I assume Cobb–Douglas production and isoelastic demand functions but,
as shown in Supplementary Material, Appendix A, the estimation strategy is compatible also
with other functional forms, including demand functions that allow for substitution and com-
plementarity patterns between other products and output of other producers. The identifying
assumptions are discussed in more detail in Section 6.1.

2.1. Production
In an industry, a group of firms produces N similar or related products i, i ∈ {1, . . . , N }. Firm
j produces at least one of these products, Oi jt = 1 denotes that firm j produces title i at time
t and Oi jt = 0 otherwise, and firm j does not make products in other industries. The physical
output of product i that firm j produces at time t is denoted by Q i jt . The firm uses three inputs
in making product i: materials Mi jt , labour L i jt , and capital K i jt .12 Labour and capital are sub-
stitutable across the product lines of the firm.13 All the factors of production are continuously
divisible and exclusive across product lines. This means that they can be flexibly allocated across
the different lines and that any given share of a firm-level input is used in only one product line.
Outputs or by-products of other product lines are not used as factors of production. All the fac-
tors other than Mi jt , L i jt , and K i jt that affect the firm’s production volume of product i at time
t are comprised in productivity, ωi jt . Productivity may represent (product line-specific) factors

11. De Loecker (2011) estimates a demand system together with production functions to control for unobserved
product prices and demand shocks. In a single-product setting, Doraszelski and Jaumandreu (2013) allow for imperfect
competition in the output market, but they need to specify only the own-price elasticity of demand.
12. Materials Mi jt may be an aggregate of a variety of materials. Quantities of different types of materials add up
to Mi jt with weights. The econometrician does not need to observe these weights provided that the model assumptions
hold and that the quantities of and expenditures on the different material types are observable, as discussed in Section 4.1
13. As discussed in Section 2.3, firm-level L jt and K jt are chosen at t − 1 and they are predetermined at the time
of production. However, if the econometrician rather assumes L jt to be flexible, i.e. to be chosen at t, and the require-
ments for making this assumption are fulfilled as discussed in Section 6.1, the assumption of labour substitutability
across product lines can be relaxed.
3320 REVIEW OF ECONOMIC STUDIES

such as management and organisation of production, downtime due to, for example, maintenance
work, and defect rates in the manufacturing process (Ackerberg et al., 2015). Allowing for pro-
ductivity differences across product lines within firms is in line with the concept of flexible
manufacturing, as discussed in the Introduction. The production technology is product-specific
and independent of production of other products; that is, the firm’s production correspondence
is additively separable in the production functions for each product. This implies that there are
no economies of scope in the form of lower variable costs.14 The production function takes the
form of Cobb–Douglas, which is a special case of the translog production function considered

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


in Supplementary Material, Appendix A. In addition, the output elasticities of materials β Mi ,
labour β Li , and capital β K i are product-specific. The product-specific constant of the production
function is denoted by β0i . Hence, the production function is written as
β β β  
Q i jt = exp (β0i ) Mi jtMi L i jtLi K i jtK i exp ωi jt . (1)

Productivity ωi jt follows an exogenous first-order Markov process. It can be divided into a con-
ditional expectation of productivity, which is a function of the productivity attained at t − 1, and
a deviation from the conditional expectation of productivity; that is, a productivity shock ξi jt :15
 
ωi jt = E ωi jt | ωi jt−1 + ξi jt . (2)

The productivity shock ξi jt is mean zero and independent across products, firms, and time.

2.2. Demand
Competition in the output market is monopolistic, that is, price changes of other products or by
other producers have only a negligible effect on the demand for product i of firm j. Firm j faces
a downward-sloping and isoelastic demand curve for product i at time t. The demand for and
price of product i of firm j at time t are denoted by Q i jt and Pi jt , respectively. The demand
function is written as
  η
Q i jt = exp αi j Pi jti exp(εi jt ). (3)

The demand for product i of firm j depends on factors unobservable to the econometrician and
denoted by αi j . These factors vary across products and firms but are constant over time. Any
shocks to the product- and firm-specific demand levels are captured by εi jt . These shocks can be
caused by changes in buyers’ preferences or income or by the number of buyers in the market,
for example. The demand shock εi jt is mean zero and independent of any information at time
t − 1. The price elasticity of demand ηi is product-specific. As firm j faces its own demand
curve for product i, it operates like a monopolist in the elastic part of the demand function.
Hence, to ensure positive input allocations to product line i, ηi is assumed to be lower than −1.
As competition is monopolistic, Pi jt is determined as a function of Q i jt , ηi , αi jt , and εi jt , as
discussed in Section 2.3.

2.3. Timing of production decisions


The firm’s decision-maker makes the product selection at t − 1, before the production period
begins. The choice of whether to make product i depends on the profit the firm’s decision-maker

14. However, it is possible to allow for economies of scope in the form of lower variable costs by modelling
productivity ωi jt as a function of the producer’s product scope, as De Loecker et al. (2016) do.
15. Defining the structure of the productivity process is essential for structural methods of production function
estimation, such as Olley and Pakes (1996).
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3321

expects from manufacturing any product i ∈ {1, . . . , N } at time t. Product choices are not cor-
related with the productivity shocks ξi jt or the demand shocks εi jt that take place only at time t.
The three types of inputs Mi jt , L i jt , and K i jt differ in how they are determined. The firm-level
labour16 L jt and capital stock K jt are predetermined at the time of production. L jt is chosen
in the previous period, at t − 1, and changing labour from L jt−1 to L jt may involve adjustment
costs. In addition, K jt is chosen at t − 1. It is determined in a dynamic process as a function
of the previous period’s capital stock K jt−1 and investment I jt−1 , as K jt = f (K jt−1 , I jt−1 ).
Both L jt and K jt depend on the (expected) prices of labour, capital, and materials. L jt and

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


K jt are also subject to independent variation. The product-level inputs L i jt and K i jt are allo-
cated
 across product lines in the  period of production, subject to the firm-level constraints
i=1,...,N ;Oi jt =1 L i jt ≤ L jt and i=1,...,N ;Oi jt =1 K i jt ≤ K jt .
The product-level materials Mi jt and the firm-level materials M jt , on the other hand, are
chosen at the time of production. Setting Mi jt and M jt does not involve adjustment costs; that is,
the choices of Mi jt and M jt do not have dynamic implications. Firm j does not have monopsony
power in the materials market, and thus the price level of materials, henceforth called simply
the price of materials, P jtM , is determined exogenously to M jt . This implies that there are no
cost economies of scope or scale in the form of lower input prices. Instead, the price level of
materials varies across firms due to, for example, differences in transport costs (Grieco et al.,
2016). Even if the materials input consists of different types of materials, P jtM does not depend
on the composition of M jt . In addition, P jtM is not fully determined by P jt−1 M
.
The timing regarding the firm’s static profit maximisation problem is as follows. At time
t, the firm has active production lines for titles i, i ∈ {1, . . . , N }, with Oi jt = 1 denoting an
active product line and Oi jt = 0 otherwise. The firm employs a predetermined capital stock
K jt and labour L jt . The productivity shocks ξi jt and the demand shocks εi jt are realised and
become observable to the firm. The firm also observes the price of materials P jtM , and chooses
the quantities of product-level materials Mi jt . The firm also decides how to allocate its labour
L jt and capital stock K jt among the different product lines in which the firm is active; that is, it
sets L i jt and K i jt . Most of the timing assumptions of this model are similar to the assumptions
previously made in the production function literature. The assumptions are compared to those in
the previous literature in Section 6.1.

2.4. The firm’s optimisation problem


Solving the product-level inputs and identifying the production function parameters requires
solving the firm’s static profit maximisation problem, which consists of setting Mi jt , L i jt , and
K i jt for the product lines in which the firm is active in order to maximise the sum of product-level
profits.17 Denoting the firm’s static profit by  jt , the static profit maximisation problem is


N
 
max  jt = Pi jt Q i jt − P jtM Mi jt
Mi jt ,L i jt ,K i jt
i=1|Oi jt =1


N 
N
s.t. L i jt ≤ L jt and K i jt ≤ K jt . (4)
i=1|Oi jt =1 i=1|Oi jt =1

16. L jt is often assumed to be a flexible input. I assume L jt to be predetermined because it is more realistic
for the Finnish labour market, as discussed in Section 6.1. However, the empirical model can be adjusted to allow for
flexible labour input if the wage (level) is observable and firms are price-takers in the labour market.
17. Firm-level labour and capital are predetermined so their costs are not included in the static profit function.
3322 REVIEW OF ECONOMIC STUDIES

Substituting in the inverse demand, Pi jt = (Q i jt (exp(αi jt + εi jt ))−1 )1/ηi jt , and the production
function, the static profit maximisation problem becomes


N  −(1/ηi )
max  jt = exp(αi j + εi jt )
Mi jt ,L i jt ,K i jt
i=1|Oi jt =1
   1/ηi +1
β β β
× exp (β0i ) Mi jtMi L i jtLi K i jtK i exp ωi jt − P jtM Mi jt

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024



N 
N
s.t. L i jt ≤ L jt and K i jt ≤ K jt . (5)
i=1|Oi jt =1 i=1|Oi jt =1

The optimisation problem yields a Lagrangian equation with two constraints. The constraints
account for not exceeding L jt and K jt in setting L i jt and K i jt for each active product line.
 maximises profit, L jt and K
More precisely, given that the firm jt are always fully utilised, and
the constraints are binding as i=1,...,N ;Oi jt =1 L i jt = L jt and i=1,...,N ;Oi jt =1 K i jt = K jt . The
Lagrangian is


N
 
Lagr = exp(αi j + εi jt )−(1/ηi )
i=1|Oi jt =1
   1/ηi +1
β β β
× exp (β0i ) Mi jtMi L i jtLi K i jtK i exp ωi jt − P jtM Mi jt
⎛ ⎞ ⎛ ⎞
N  N
+ λ L jt ⎝ L jt − L i jt ⎠ + λ K jt ⎝ K jt − K i jt ⎠ . (6)
i=1|Oi jt =1 i=1|Oi jt =1

The first-order conditions of static profit maximisation for firm j, ∀ j ∈ {1, . . . , J } at time t,
∀t ∈ {1, . . . , T } are

∂ Lagr 1  −(1/ηi )
= + 1 exp(αi j + εi jt )
∂ Mi jt ηi
   1/ηi +1 β Mi
β β β
× exp (β0i ) Mi jtMi L i jtLi K i jtK i exp ωi jt − P jtM
Mi jt
= 0 ∀i ∈ {1, . . . , N } | Oi jt = 1, (7)

∂ Lagr 1  −(1/ηi )
= + 1 exp(αi j + εi jt )
∂ L i jt ηi
   1/ηi +1 β Li
β β β
× exp (β0i ) Mi jtMi L i jtLi K i jtK i exp ωi jt − λ L jt
L i jt
= 0 ∀i ∈ {1, . . . , N } | Oi jt = 1, (8)

∂ Lagr 1  −(1/ηi )
= + 1 exp(αi j + εi jt )
∂ K i jt ηi
   1/ηi +1 βK i
β β β
× exp (β0i ) Mi jtMi L i jtLi K i jtK i exp ωi jt − λ K jt
K i jt
=0 ∀i ∈ {1, . . . , N } | Oi jt = 1, (9)
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3323

∂ Lagr N
= L jt − L i jt = 0, (10)
∂λ L jt i=1|O =1 i jt

∂ Lagr N
= K jt − K i jt = 0. (11)
∂λ K jt i=1|O =1 i jt

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


3. IDENTIFICATION AND ESTIMATION STRATEGY
Firm-level Cobb–Douglas production functions have been estimated in numerous studies. Esti-
mation of the product-specific production functions in the present study differs from the
estimation of firm-level functions in one important aspect: the product-level inputs are unob-
servable to the econometrician. This implies that all the elements in the production function are
unobservable: the input quantities, the output elasticities of the inputs, and productivity. In other
words, not only are the inputs endogenous to the unobservable productivity, which is a standard
problem in production function estimation, but they are also unobservable. Clearly, these two
problems are closely related. The observables to the econometrician are the product-level out-
put quantities and prices Q i jt and Pi jt , the firm-level predetermined inputs L jt and K jt , and the
firm-specific price of the materials input P jtM . All these observables are measured without error.
My identification strategy builds on the structural production function literature and espe-
cially two remarks. The first remark is that the product-level inputs can be solved parametrically
from the firm’s static profit maximisation problem as functions of the observable data and the
parameters to be estimated, as in equations (12)–(14). This is possible because the unobserv-
able productivity terms can be substituted away by using the parametric form of the production
function and the observable output quantities. The second remark is that when competition
in the output market is imperfect, firms make their production decisions not only as a func-
tion of supply-side factors but also as a function of the demand for the products. Intuitively,
the higher the demand for a given product, the more inputs the firm is willing to allocate to
the product line. Product–firm-specific differences in demand generate within-firm variation in
the profitability of manufacturing various products, which combine with supply- and demand-
side instruments to allow the identification of product-level input allocations and the production
function parameters.
The estimation strategy does not consider product or firm selection. It is therefore suitable
for a setting where the likelihood of firm entry and exit is low, and changes in firms’ product
assortments are infrequent. Firm and product selection are discussed in Section 6.1.
The production and demand functions are estimated by the generalised method of moments
(GMM). In Section 3.1, I first show how the product-level input allocations and the productivity
shock can be written in terms of the observable data and parameters to be estimated.18 I then dis-
cuss how the production functions can be identified using instrumental variables familiar from
estimation of (firm-level) production functions. Alternatively, the production functions can be
identified using optimal instruments, as shown in Supplementary Material, Appendix A, which
is my choice in the empirical application of this paper. In Section 3.2, I present an overidenti-
fying moment condition for the output elasticities of flexible inputs such as materials,19 which

18. I provide solutions for the product-level input allocations for alternative functional forms of production and
demand in Supplementary Material, Appendix A.
19. The identification strategy does not require firm-level flexible inputs to be observable. But as data sets like
the one used in this paper often include firm-level expenditures on flexible inputs, they can, together with the firm-level
price of materials, be used in overidentifying moment conditions such as this one.
3324 REVIEW OF ECONOMIC STUDIES

is based on the firm’s input demand function, as opposed to the production function. Finally, in
Section 3.3, I discuss how the own-price elasticities of demand are identified.20 They are the only
parameters of the demand function that need to be estimated. The model can be estimated in one
step, as in the empirical application of this paper. Alternatively, the own-price elasticities may be
estimated first and then be used in estimating the product-level inputs and production functions.

3.1. Identification of the production function

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


The production functions are identified by, first, using the assumption of static profit maximi-
sation and the parametric form of the model to write for the unobservable input allocations.
Second, the productivity shocks are written using the Markovian structure of the productiv-
ity process. Third, the identifying moments are written using instruments familiar from the
literature.
I show first how to solve for the product-level inputs as a function of observable data and
the parameters to be identified. The optimal input choices are determined by the first-order con-
ditions of the firm’s static profit maximisation problem21 in equations (7)–(11). As equations
(7)–(9) show, the input choices depend on the productivity terms ωi jt , which are unobservable
to the econometrician. However, the unobservable ωi jt can be substituted away from equations
(7)–(9) with the production function in equation (1) inverted for ωi jt ; that is, the right-hand side
β β β
of ωi jt = log(Q i jt (exp(β0i )Mi jtMi L i jtLi K i jtK i )−1 ). In addition, the right-hand side of the inverse
demand function Pi jt = exp(αi j + εi jt )−(1/ηi ) Q 1/ηi that appears in equations (7)–(9) can be sub-
stituted away with the left-hand side of the inverse demand function; that is, the observable
Pi jt . Consequently, the only demand function elements that remain in equations (7)–(9) are the
own-price elasticity ηi and the observable Pi jt . The Lagrangian multiplier λ L jt in equation (8) is
eliminated by substitution among equations (8) and (10). Similarly, λ K jt in equation (9) is elim-
inated by substitution among equations (9) and (11). After these substitutions, the first-order
conditions in equations (7)–(11) can be rewritten as follows:

1 β Mi
Mi jt = + 1 Pi jt Q i jt M where i ∈ {1, . . . , N } | Oi jt = 1, (12)
ηi P jt

1
ηi
+ 1 Pi jt Q i jt β Li L jt
L i jt =   where i ∈ {1, . . . , N } | Oi jt = 1, (13)
i=1,...,N ;Oi jt =1 ηi + 1 Pi jt Q i jt β Li
1


1
ηi
+ 1 Pi jt Q i jt β K i K jt
K i jt =   where i ∈ {1, . . . , N } | Oi jt = 1. (14)
i=1,...,N ;Oi jt =1 ηi + 1 Pi jt Q i jt β K i
1

These solutions22 for product-level inputs are expressed only in terms of the observable data
(Pi jt , Q i jt , P jtM , L jt , K jt ) and parameters to be estimated (ηi , β Mi , β Li , β K i ).

20. I discuss identification of the own-price elasticity of demand of alternative demand functions in Supplemen-
tary Material, Appendix A.
21. Doraszelski and Jaumandreu (2013) also define the first-order conditions for static profit maximisation. They
solve for the productivity term by parametrically inverting the intermediate input demand function thus obtained.
22. The demand function, as specified in equation (3), does not allow for substitution and complementar-
ity between different products and products of different producers. However, as shown in Supplementary Material,
Appendix A, equations (12)–(14) are also the solutions for the product-level inputs when the demand function is modified
to allow for substitution and complementarity.
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3325

After solving for the product-level inputs, the productivity shock ξi jt can be written, given
the productivity process as defined in equation (2), as
 
ξi jt = ωi jt − E ωi jt |ωi jt−1 , (15)

where ωi jt is inverted from the production function in equation (1); that is, ωi jt =
β β β
log(Q i jt /exp(β0i )Mi jtMi L i jtLi K i jtK i ), Mi jt , L i jt , and K i jt are as defined in equations (12)–(14), and
the lagged ωi jt−1 , Mi jt−1 , L i jt−1 , and K i jt−1 are defined accordingly. In this way, ξi jt is written

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


as a function of observables (Pi jt , Q i jt , P jtM , L jt , K jt , Pi jt−1 , Q i jt−1 , P jt−1 M
, L jt−1 , K jt−1 ) and
the parameters (ηi , β Mi , β Li , β K i , β0 , and the parameters in E[ωi jt | ωi jt−1 ]) to be estimated.
The output elasticities of inputs β Mi , β Li , and β K i can be identified by using instruments
familiar from the structural production function literature.23,24 The exogenous P jtM , the prede-
termined L jt and K jt , and the lagged Pi jt−1 , P jt−1 M
, L jt−1 , and K jt−1 are all orthogonal to ξi jt .
Product-level materials Mi jt are chosen as a function of P jtM . The higher P jtM is, the lower is
Mi jt ; that is, the two variables are correlated.  The product-level labour L i jt is chosen in static
profit maximisation such that L jt − i=1,...,N ;Oi jt =1 L i jt = 0. Hence, the higher the firm-level
L jt , the higher the product-level L i jt , as equation (13) shows. Similarly, the firm-level K jt and
the product-level K i jt are correlated because  the product-level K i jt in equation (14) is set in static
profit maximisation such that K jt − i=1,...,N ;Oi jt =1 K i jt = 0. In the same way, the lagged P jt−1 M
,
25
L jt−1 , and K jt−1 determine the lagged Mi jt−1 , L i jt−1 , and K i jt−1 , respectively. In addition to
the supply-side factors, the product-level inputs are chosen as functions of the product-specific
demand. Furthermore, the demand side is a source of product-specific, within-firm identifying
variation. The lagged output price Pi jt−1 increases in the product-firm-specific demand level αi j
and is thus correlated with the profit maximising Mi jt , L i jt , and K i jt .26
It is important to note that in the solution for E[ωi jt | ωi jt−1 ], β Mi , β Li , and β K i interact with
the parameters governing E[ωi jt | ωi jt−1 ]. As a consequence, the production function parameters
can be separately identified (except for β0 ) only if the current supply-side variables—P jtM , L jt ,
M
and K jt —vary independently from the lagged supply-side variables (P jt−1 , L jt−1 , and K jt−1 ).27
M M M
In other words, P jt has to vary over time such that P jt is not functionally dependent on P jt−1 .
In addition, L jt and K jt have to vary over time such that the lagged supply-side variables,
M
including P jt−1 , do not fully determine L jt and K jt . These assumptions are stated in Section
2.3. In addition, the firm chooses L jt and K jt based on the same information set, but L jt and
K jt are subject to sufficient independent variation.28
To summarise, β Mi , β Li , and β K i can be identified with instruments that are familiar from the
estimation of firm-level production functions. The difference in the moment conditions is that

23. For a discussion on identification of gross output production functions, and especially on avoiding collinearity
problems, see Gandhi et al. (2020).
24. The discussion on identification of β Li and β K i is based on the assumption that the probability of change in
the firm’s product assortment is low and therefore product selection does not require particular attention. The potential
firm or product selection bias is discussed in Section 6.1.
25. In addition, Mi jt−1 , L i jt−1 , and K i jt−1 are solved for because they are needed for solving ωi jt−1 , which
is the explanatory variable in E[ωi jt | ωi jt−1 ].
26. Lagged output prices have previously been used as instruments by Doraszelski and Jaumandreu (2013).
27. Potential functional dependence problems in production function estimation are discussed in Doraszelski and
Jaumandreu (2013), Ackerberg et al. (2015), and Gandhi et al. (2020). Gandhi et al. (2020) use a support condition on
the regressors adapted from Newey et al. (1999) to avoid the functional dependence problem.
28. As P jtM is observed after K jt and L jt have been set, it is a source of variation for Mi jt that does not determine
K jt and L jt .
3326 REVIEW OF ECONOMIC STUDIES

the productivity shocks are defined at the product-firm level. The moment conditions29 are
       
E ξi jt | P jtM = E ξi jt | L jt = E ξi jt | K jt = E ξi jt | P jt−1
M
     
= E ξi jt | L jt−1 = E ξi jt | K jt−1 = E ξi jt | Pi jt−1 = 0
where i ∈ {1, . . . , N }. (16)

As a consequence of the mean independence conditions in equation (16), β Mi , β Li , and β K i


can be identified using optimal instruments. They are discussed and derived in Supplementary

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


Material, Appendix B. The GMM objective function formed using the moments in equation
(16) is not convex because β Mi , β Li , and β K i enter the moments non-linearly. The optimisation
algorithm is discussed in Section 4.5.
In addition to the output elasticities β Mi , β Li , and β K i , the production function has parameters
that govern the first-order Markov process of productivity, E[ωi jt | ωi jt−1 ], as well as a constant,
β0i . As discussed in Section 4.3, E[ωi jt | ωi jt−1 ] is approximated by a polynomial. Because the
variables in the polynomial are exogenous, they are used as instruments for themselves.

3.2. Overidentifying restriction from firm-level demand for the flexible input
In addition to estimating the production functions, I estimate the firm-level demand function
for the flexible input M jt . The firm-level M jt comprises the unobservable product-level materi-
als choices Mi jt in equation (12). Observing the firm-level M jt is not necessary for identifying
the production functions, but if M jt is observable,30 it can be used to write an overidentifying
restriction for the output elasticity of the flexible input, β Mi . In practice, firm-level expenditures
on materials are often observable in financial statement data. Given the firm-level price of mate-
rials, P jtM , a firm-level measure of materials input quantity M jt can be obtained, as discussed in
Section 4.1.
If the econometrician  is confident that M jt is measured with zero measurement error, the
restriction of M jt = i=1,...,N ;Oi jt =1 Mi jt can be imposed. I require the firm-level M jt to be
measured with a multiplicative mean zero measurement error M jt . As discussed in Section 4.1,
the source of measurement error in M jt is not P jtM but the variable of firm-level expenditures on
materials in the financial statement data. The firm-level demand for the flexible input is written
as
  
M jt = Mi jt 1 + M jt , (17)
i=1,...,N ;Oi jt =1

where Mi jt is defined as in equation (12). The firm-level input demand function can be iden-
tified using the product-level sales Pi jt Q i jt as the instruments. Pi jt Q i jt is correlated with the
underlying input choice Mi jt for two reasons. First, the higher the product-firm-specific demand
level αi j + εi jt and hence Pi jt , the higher the profit maximising Mi jt . Second, the higher the
productivity and the resultant output Q i jt , the higher the input choice Mi jt . Since Pi jt Q i jt is
uncorrelated with the measurement error M jt , Pi jt Q i jt is a valid instrument for identifying β Mi .
One more moment condition for identifying β Mi is therefore
 
E M jt | Pi jt Q i jt = 0 where i ∈ {1, . . . , N }. (18)

29. These moment conditions do not identify the price elasticity of demand, ηi . The moment conditions that
identify ηi are discussed and presented in Section 3.3.
30. Observing the firm-level materials input M jt does not eliminate the need for the observable materials
price P jtM .
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3327

As the above moment condition is an overidentifying one, it can also be left out of estima-
tion. But if M jt is observable, allowing for non-zero measurement error and including the
overidentifying assumption in estimation uses the available information to its fullest extent.

3.3. Identification of the price elasticity of demand


There are two parameters in the demand function: αi j and ηi . The product-firm-specific αi j is
a constant. Identification of ηi requires an instrument because the price is endogenous.31 The

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


materials price P jtM , the firm-level labour L jt , and the firm-level capital stock K jt correlate with
the product price, while they are uncorrelated with the demand shock εi jt . Hence, ηi is identified
with the following moment conditions:
     
E εi jt | P jtM = E εi jt | L jt = E εi jt | K jt = 0 where i ∈ {1, . . . , N }. (19)

If Ti j is small, however, αi j is an incidental parameter. As discussed in Section 3.1, identifica-


tion of the production function does not require αi j to be identified, but the potential incidental
parameter problem may lead to an inconsistent estimator of ηi . To avoid this, identification of ηi
requires an instrument that is uncorrelated with αi j . As P jtM is determined exogenously, it is not
correlated with αi j and is therefore also a valid instrument when αi j is an incidental parameter.32

4. DATA AND EMPIRICAL IMPLEMENTATION


4.1. Data
I use the Longitudinal Database on Plants in Finnish Manufacturing (LDPM) and the Industrial
Output data from Statistics Finland for the years 2004–2011 (updated in 2012). The two data sets
include plants that belong to manufacturing firms with at least twenty employees and a subset
of plants of firms with less than twenty employees. The reporting units are mainly plants. The
only exceptions are in the Industrial Output data, where a few plants that belong to the same firm
report jointly. For these reporting units, I aggregate the observations in the LDPM accordingly.
I estimate product-specific production functions of plants declaring “Manufacture of wood
and of products of wood and cork, except furniture; manufacture of articles of straw and plaiting
materials”, as Division 16 of Eurostat’s PRODCOM (Production communautaire) classifica-
tion is described. This industry is a good example of an industry in which firms manufacture
several different products and presumably employ multiple technologies. The products are clas-
sified according to eight-digit PRODCOM codes that are supplemented by national ten-digit
subclasses. The titles are provided in Table 1. From the production perspective, products within
the fairly narrowly defined titles appear to be comparable in physical quantities. At the same
time, product differentiation is likely for several products, and while there are a large number
of producers in the industry, competition in the output market is likely to be monopolistic. The
output is sold to several other industries, including construction, trade, manufacture of paper and
paper products, and generation of (bio)energy.33 For each product a plant produces in a given

31. For a discussion of the instruments used in demand estimation, see, for example, Ackerberg et al. (2007).
32. The discussion of the identification of ηi is based on the assumption that the probability of change in the
firm’s product assortment is low, and therefore product selection does not require particular attention. Firm and product
selection are discussed in Section 6.1.
33. The output sold to the paper and paper products industry and the energy sector are by-products of wood
manufacturing. The data used in this paper have seven products that can be considered by-products: 16.10.23.03 Conif-
erous wood in chips or particles; 16.10.23.05 Non-coniferous wood in chips or particles; 16.10.41.00.10 Sawdust;
16.10.41.00.20 Woodchips; 16.10.41.00.40 Lathes, borders, etc.; 16.10.41.00.60 Bark; 16.10.41.00.80 Other wood waste
3328 REVIEW OF ECONOMIC STUDIES

year, I observe the output measured in a physical unit, Q i jt , as well as the sales revenue. These
two factors yield the average price of the product in a given year, Pi jt .
Similarly for the intermediate products and materials, I observe physical quantities and
expenditures by the PRODCOM titles. These data are used to obtain the “price” of materials,
P jtM , computed as the Elteto–Koves–Szulc (EKS) multilateral price index (e.g. Hill, 2004; Neary,
2004). For firm j, it can be expressed as follows (suppressing subscripts t for time):
    1/J

J 
PF q j , q j , p j , p j

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


P jM =   , (20)
j  =1
PF q j  , qb , p j  , pb

   
where q j and p j are the quantity and price vectors of firm j  , and PF (q j , q j , p j , p j ) is the
bilateral Fisher price index between firm j and firm j  , j  = 1, . . . , J (J is the number of firms),
which is given by
 1/2
  

qj ∗ pj qj ∗ pj
PF q j , q j , p j , p j =   × ,
q ∗p
j j q j ∗ p j

   j j
where q j ∗ p j = n=1,...,N qi pi (N is the number of intermediate product and material titles);
 
similarly for PF (q j , qb , p j , pb ), where b stands for the base firm chosen. The EKS multilateral
index satisfies the circularity (transitivity) requirement, which implies that the same index is
obtained irrespective of whether firms are compared with each other directly or through their
relationships with other firms (Hill, 2004; Neary, 2004). The EKS multilateral index is thus well
suited for my purpose of comparing firms when no representative firm exists and bundles of
products differ between firms.
In addition to the physical quantities of and expenditures on the intermediate products and
materials by PRODCOM title, I observe plant-level intermediate product and material purchase
values in the LDPM, which I denote by M jtL D P M . To obtain a plant-level measure of intermediate
product and material use that does not comprise input price differences across plants, M jt , I
deflate M jtL D P M by the price index for intermediate products and materials, P jtM . While P jtM is
assumed to be measured without error, M jtL D P M and hence M jt are allowed to be measured with
error, as discussed in Section 3.2.
The labour input is measured in labour costs, which comprise salary and social payments.
The capital stock is estimated using the perpetual inventory method, K jt = (1 − δ)K jt−1 +
I jt−1 , where δ = 0.1 and I jt is investment.34
The estimation strategy poses certain requirements on the estimation sample. First, obser-
vations with missing variables cannot be used in estimation. Second, each product has to be
observed in at least four pairs of observations, with each observation pair being from two
consecutive years in a given plant.35 The four observation pairs may be from, for example, five

(excluding sawdust, woodchips, bark, lathes, borders, pellets, briquettes, etc.). The data show that about two thirds of the
wood manufacturing plants have either 0% or 100% sales in these by-products, which suggests that shared inputs in the
production of “main products” and by-products, which are assumed away in the estimation strategy, are not important.
34. I have set δ = 0.1 following the examples of, among others, Levinsohn and Petrin (2003), who estimate
production functions of Chilean manufacturing firms, and Doraszelski and Jaumandreu (2013), who estimate production
functions of Spanish manufacturing firms.
35. The requirement of at least four observation pairs can cut down the data set considerably if there are many
products that only very few plants or firms produce. For most industries, however, the aforementioned requirement is
not likely to affect the sample considerably.
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3329
TABLE 1
Product titles
PRODCOM Title
16.10.10.33 Coniferous wood; sawn or chipped lengthwise, sliced or peeled, of a thickness >6 mm,
end-jointed, sanded or planed
16.10.10.33.10 Spruce wood (Picea abies Karst.), sanded or planed, end-jointed, sawn or chipped
lengthwise, sliced or peeled, of a thickness >6 mm
16.10.10.33.20 Pine wood (Pinus sylvestris L.), sanded or planed, end-jointed, sawn or chipped lengthwise,
sliced or peeled, of a thickness >6 mm

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


16.10.10.35 Spruce wood (Picea abies Karst.), fir wood (Abies alba Mill.)
16.10.10.37 Pine wood (Pinus sylvestris L.)
16.10.10.50 Wood, sawn or chipped lengthwise, sliced or peeled, of a thickness >6 mm
(excluding coniferous and tropical woods and oak blocks, strips and friezes)
16.10.21.10 Coniferous wood continuously shaped (including strips and friezes for parquet flooring,
not assembled)
16.10.23.03 Coniferous wood in chips or particles
16.10.23.05 Non-coniferous wood in chips or particles
16.10.41.00.10 Sawdust
16.10.41.00.20 Woodchips
16.10.41.00.40 Lathes, borders, etc.
16.10.41.00.60 Bark
16.10.41.00.80 Other wood waste (excluding sawdust, woodchips, bark, lathes, borders, pellets, briquettes,
etc.)
16.21.11.00 Plywood, veneered panels and similar laminated wood, of bamboo
16.21.12.14 Plywood consisting solely of sheets of wood (excluding of bamboo), each ply not
exceeding 6 mm thickness,
with at least one outer ply of non-coniferous wood (excluding tropical wood)
16.21.12.17 Plywood consisting solely of sheets of wood (excluding of bamboo), each ply not
exceeding 6 mm thickness
(excluding products with at least one outer ply of tropical wood or non-coniferous wood)
16.21.13.13 Particle board, of wood
16.21.21.18.30 Veneer for plywood, cross-banded plywood and other wood, of coniferous wood, sawn
lengthwise, sliced or peeled, of a thickness <= 6 mm
(excluding end-jointed, planed, sanded and board for manufacturing pencils)
16.21.21.18.80 Veneer for plywood, cross-banded plywood and other wood, of hardwood, sawn length-
wise, sliced or peeled, of a thickness <= 6 mm
(excluding end-jointed, planed, sanded and board for manufacturing pencils)
16.21.22.00 Densified wood, in blocks, plates, strips or profile shapes
16.22.10.60 Parquet panels of wood (excluding those for mosaic floors)
16.23.11.10 Windows, French-windows and their frames, of wood
16.23.11.50 Doors and their frames and thresholds, of wood
16.23.19.00.12 Carpenter’s produce for walls, of wood
16.23.19.00.16 Carpenter’s produce for stairs, of wood
16.23.19.00.26 Components for sauna, of wood
16.23.19.00.32 Panel elements (also glulam and cellular panels), of wood
16.23.19.00.36 Ceiling elements, of wood
16.23.19.00.42 Glulam beams and columns
16.23.19.00.46 Vertical and horizontal beams (excluding glulam beams and columns)
16.23.19.00.52 Log frames for buildings of wood
16.23.19.00.90 Other carpenter’s produce, of wood
(excluding doors, windows, produce for floors, walls, stairs and sauna, panel and ceiling
elements, beams, columns and log frames)
16.23.20.00.20 Residential buildings of wood, for permanent habitation
16.23.20.00.40 Residential buildings of wood, for recreational use
(continued)
3330 REVIEW OF ECONOMIC STUDIES
TABLE 1
Continued
PRODCOM Title
16.23.20.00.60 Saunas of wood (outdoor saunas, assembled or prefabricated)
16.23.20.00.90 Buildings of wood (assembled or prefabricated) (excluding residential buildings and
saunas)
16.24.11.35 Box pallets and load boards of wood (excluding flat pallets)
16.24.13.20 Cases, boxes, crates, drums and similar packings of wood (excluding cable drums)
16.24.13.50 Cable-drums of wood

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


16.29.14.90 Other articles of wood (excluding pallet collars)

consecutive years in a single plant, or two consecutive years in four plants. This is because,
for each product, there are four parameters to be estimated by solving the non-linear optimisa-
tion problem,36 and because estimating the first-order Markov process of productivity evolution
requires sequences of at least two observations. To be included in the estimation sample, a plant-
level observation has to contain at least one product that the plant also produced in the previous
period or in the following period. Observations that do not fulfil the aforementioned criteria are
dropped from the sample.
Recall that measurement error in output is assumed to be zero. Unfortunately, there is no
other output variable that could be used to verify the accuracy of the product-level sales rev-
enue variables. The only other output variable available is the plant-level gross output value
reported in the LDPM. Gross output value is defined as the sum of sales revenue, deliveries
to other plants of the firm, changes in inventories, production for own use, and other business
revenue, deducting capital gains and acquisition of merchandise. Not surprisingly, gross out-
put value is not equal to the sum of product-level sales revenues from production in all of the
plants. There are several possible explanations for this. Plants may produce output that is not
included in the sales revenue from production (due deliveries to other plants of the firm, positive
changes in inventories, production for own use), or the sales revenue data may include output
produced in some previous year (due to negative changes in inventories). Moreover, because
capital gains and acquisition of merchandise are deducted from gross output value, it is not pos-
sible to make strong inferences about potential measurement error in output. Unfortunately, the
various components of gross output value are not reported in the LDPM, and hence I cannot
identify why gross output value may differ from sales revenue. However, to reduce the likelihood
of using observations with major measurement error in output, I use only those observations for
which the ratio of the sum of sales revenue to gross output value is at least 0.6 but not more
than 1.4.
In practice, the aforementioned requirements on data restrict the estimation sample as fol-
lows. I start with 3660 product-plant-year observations (exactly 1200 plant-year observations)
and sixty-six products once the observations with any missing variables have been excluded. I
first drop products with fewer than four observations pairs (step 1). I then drop (1) those plant-
year observations for which at least one product has been dropped in step 1, as well as (2)
plant-year observations for which the ratio of the sum of sales revenue (reported in the Indus-
trial Output data) to gross output value (reported in the LDPM) is less than 0.6 or more than 1.4
(step 2). I repeat these two steps sequentially until no observations drop out. Repeating steps 1

36. The rest of the parameters are concentrated out of the non-linear optimisation problem, as discussed in
Section 4.3.
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3331
TABLE 2
Descriptive statistics
Percentiles

5th 25th 50th 75th 95th Mean Std. dev.


Salesa 2.42 6.87 14.48 33.05 80.02 27.12 46.21
Capitala 0.30 1.25 3.71 11.55 27.30 8.83 16.04
Laboura 0.65 1.22 2.38 4.03 10.04 4.18 9.16
Materialsa 1.74 5.09 11.56 31.57 70.40 23.88 38.69

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


Price of materialsb 0.42 0.64 0.86 1.22 2.38 1.04 0.63
Number of products 1 1 3 4 7 3.25 2.35
Observations 904
Relative pricec 0.21 0.83 0.95 1.05 1.73 1 0.97
Observations 2936
a Measured in 100,000 euros.
b Measured by the EKS multilateral price index.
c Relative to the product-year-specific mean.

and 2 is required because executing step 2 may result in some of the remaining products having
less than four observation pairs. Similarly, dropping products brings about plant-year observa-
tions for which step 2 has to be carried out again. After repeating these two steps, there are 2936
product-plant-year observations (904 plant-year observations) and forty-two products left. This
estimation sample comes from 190 plants during an eight-year time period.
In this estimation sample, 673 out of 904 plant-year observations have a preceding observa-
tion. These 673 plant-year observations have altogether 2280 product-plant-year observations,
of which 2053 are of products that the plant also produced in the previous year. This means that
the product entry rate, excluding plants entering the data, is 10%. The product exit rate, exclud-
ing plants exiting the data, is 7%. Completely changing the product selection is very uncommon:
there is only one plant-year observation with a totally different product selection in year t than
in t − 1, with only one product produced in each year. Plants’ product assortments range from
one to 17 products, with 659 out of 904 plant-year observations containing at least two product-
level observations. Product assortments vary across plants such that there are no typical product
combinations. The correlation between making two products is low for most product pairs: the
absolute value of the correlation coefficient is lower than 0.05 (0.1) [0.2] for 63% (81%) [91%]
of all product pairs.
The estimation sample is summarised in Table 2. The descriptive statistics show that the
plants vary in size, as measured by their sales and by their input use (capital, labour, and materi-
als, reported in hundreds of thousands of euros in Table 2). For example, the fifth percentile plant
has sales of 240,000 euros in a year, while the 95th percentile plant has sales of 8 million euros.
The price of materials, which is measured by the EKS multilateral price index, also varies across
plants. For the fifth percentile plant, the materials price index is 0.42, while for the 95th per-
centile plant the index is 2.38. A median plant produces three products, and the 95th percentile
plant produces seven products. In Table 2, the product-specific output prices are characterised as
relative prices, where the relative price is the ratio of the price to the product-year-specific mean.
The relative output prices also have a rather wide distribution, with the fifth percentile relative
price being 0.21 and the 95th percentile being 1.73.
Output prices vary not only between plants but also to a large extent within plants. Table 3
summarises the mean differences between product prices at the plant level. In a two-product
plant, the mean difference between the relative prices is 0.69. In a three-product plant, the mean
difference is 0.35 between the highest and the second highest and 0.27 between the second
3332 REVIEW OF ECONOMIC STUDIES
TABLE 3
Within-plant price differences
Mean differences between relative prices, ordered from highest to lowest
(Standard deviation in brackets)

Number of products 1st and 2nd 2nd and 3rd 3rd and 4th 4th and 5th Other
2 0.69
(1.07)
3 0.35 0.25

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


(0.78) (0.27)
4 0.45 0.15 0.27
(2.78) (0.25) (0.29)
5 0.74 0.15 0.12 0.34
(3.21) (0.28) (0.17) (0.31)
≥6 0.23 0.10 0.07 0.07 0.14
(0.26) (0.13) (0.09) (0.12) (0.20)

highest and the lowest price. Plants with larger scopes also have within-plant price differences,
as shown in Table 3. This suggests that demand varies across the different products that a plant
produces.

4.2. Choice of instruments


Instead of using the moment conditions with traditional instruments in equation (16), I use the
moment conditions with optimal instruments37 as defined in Supplementary Material, Appendix
B (equations (A.19)–(A.21)), as well as the overidentifying moment condition in equation (18).
The demand equations are identified using the moment conditions in equation (19). As four
moment conditions are sufficient for exact identification of the model, there are three overi-
dentifying restrictions in this set of moments. Some of the eight- or 10-digit products have
at least four but fewer than seven observation pairs; with these cases, I cannot use all seven
moment conditions. Instead of dropping observations of the product entirely, I drop some of the
overidentifying moments for these products. For product i with only four observations pairs, I
adopt moments E[ξi jt | z Mi jt ] = 0, E[ξi jt | z Li jt ] = 0, E[ξi jt | z K i jt ] = 0, and E[εi jt | P jtM ] = 0.
Moment E[ M jt | Pi jt Q i jt ] = 0 (E[εi jt | L jt ] = 0) [E[εi jt | K jt ] = 0] is used when there are at
least five (six) [seven] observation pairs. This gives a total of 272 moments.

4.3. Estimation of the constants and the productivity process


The production and demand functions involve parameters that can be concentrated out of the
non-linear optimisation problem that is solved for estimating β Mi , β Li , β K i , and ηi . In the
production function, these parameters are the product-specific constant β0i and the parame-
ters governing the productivity process E[ωi jt | ωi jt−1 ], for fixed β Mi , β Li , β K i , and ηi . The
productivity process is estimated nonparametrically by approximating it with a second-order
polynomial of the lagged productivity term ωi jt−1 (β Mi , β Li , β K i , ηi ), denoted by g(ωi jt−1 ). Esti-
mates of the parameters in g(ωi jt−1 ), denoted by γi , and β0i , are obtained by regressing the

37. The objective function appears smoother and the estimates are less responsive to the starting values when the
optimal instruments are used instead of the traditional instruments. This is because the functional forms are exploited to
a fuller extent.
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3333

productivity level implied by a given set of parameter values, ωi jt (β Mi , β Li , β K i , ηi ), on the


second-order polynomial terms of the implied lagged productivity, ωi jt−1 (β Mi , β Li , β K i , ηi ).
The demand function has one parameter, the product-firm-specific constant αi j , that can be
concentrated out of the non-linear optimisation problem. For a given ηi , the estimation equation
is qi jt − ηi pi jt = αi j + εi jt , which can be estimated by ordinary least squares (OLS).

4.4. Parameter restrictions

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


Every product i is related to four parameters that are estimated by solving the non-linear optimi-
sation problem: the price elasticity ηi and the output elasticities β Mi , β Li , and β K i . If I defined
the parameters at the eight- or 10-digit level, I would need to estimate 42 × 4 = 168 param-
eters by solving the non-linear optimisation problem. At least in my setting, this is too many.
Instead, I define these parameters at the three-digit level,38 which yields two product categories:
“Sawmilling and planing of wood” (PRODCOM code 161) and “Products of wood, cork, straw
and plaiting materials” (PRODCOM code 162). This specification implies estimating 2 × 4 = 8
parameters by solving the non-linear optimisation problem. The parameters governing the pro-
ductivity process g(ωi jt−1 ) are also specified at the three-digit level. The production function
constant β0i is specific to the product as defined at the eight- or 10-digit level.39 The demand
constant αi j is specific to the eight- or 10-digit product and the plant.
There are fifteen products in PRODCOM category 161 and 27 titles in PRODCOM category
162. A plant produces on average 2.17 titles in category 161 and 1.08 titles in category 162.
Of the plants in the sample, 56% produce at least one product in category 161, and 61% of the
plants produce at least one product in category 162. Of the plants that produce any product in
category 161, 86% produce at least two titles in that category. Similarly, 43% of the plants that
produce any product in category 162 produce more than one title in that category.

4.5. Estimation algorithm


The parameters are estimated by iterated GMM. The estimation algorithm for computing the
moments is as follows. At the beginning of each outer GMM iteration, I compute the optimal
instruments given the starting values for ηi , β Mi , β Li , and β K i . First, I compute the productivity
level implied by the starting values for ηi , β Mi , β Li , and β K i , which I call h i jt instead of ωi jt
(step 1). Second, I estimate the parameters of the implied productivity process g(h i jt−1 ) by
OLS, where h i jt is the dependent variable, and the explanatory variables are polynomial terms
of h i jt−1 , as in the Markov process for productivity (step 2). Finally, I compute the optimal
instruments as functions of the data, the starting values for ηi , β Mi , β Li , and β K i , and the implied
estimates of the parameters in g(h i jt−1 ). On the second and any subsequent outer GMM iteration,
the starting values for ηi , β Mi , β Li , and β K i are the parameter estimates obtained in the previous
outer GMM iteration.
For each inner GMM iteration, I compute the moments, given some values for ηi , β Mi ,
β Li , and β K i , and the optimal instruments computed at the beginning of the outer GMM iter-
ation. First, I compute h i jt as in step 1 of the outer GMM iteration. Second, I estimate the
parameters of g(h i jt−1 ) as in step 2 of the outer GMM iteration. Third, I compute the produc-
tivity shock ξi jt and the measurement error in the firm-level flexible input M jt , as described in
equations (15) and (17), respectively. The demand shock εi jt is obtained from equation (3), as

38. The implications of this restriction are discussed in Section 6.1.


39. Different products’ physical output productivities cannot be meaningfully compared unless the different
products’ productivity distributions are normalised in some way.
3334 REVIEW OF ECONOMIC STUDIES

the product-firm-specific constant αi j is estimated by OLS given the output and price data and
some value for ηi . Finally, I compute the moments.
As multiple parameters are estimated by solving the non-linear optimisation problem, I
need to make sure that the estimation routine reaches the global minimum among the local
minima of the GMM objective function. I experiment with various minimisation algorithms;
the Gauss–Newton algorithm turns out to be the least sensitive to starting values. I also run the
estimation routine with a large set of alternative starting values.40 The GMM estimator uses the
efficient weighting matrix. Since the parameter estimates are obtained in one step, the asymptotic

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


standard errors are computed using standard GMM formulas and numerical derivatives.

5. RESULTS
The estimation results are presented in Table 4. The production and demand functions are esti-
mated for the products in “Sawmilling and planing of wood” (PRODCOM 161) and “Products of
wood, cork, straw, and plaiting materials” (PRODCOM 162). All the output elasticities of inputs
and price elasticities of demand are statistically significant.41 In addition, the estimates of the
output elasticities are statistically different for the technologies in the two product groups. The
output elasticity of materials β Mi is considerably higher for PRODCOM 162 than for PROD-
COM 161 (0.73 and 0.37, respectively). The output elasticity of labour β Li is again considerably
lower for PRODCOM 162 than for PRODCOM 161 (0.13 and 0.36, respectively). The order of
these estimates is not what we may expect as the output titles in PRODCOM 162 are more
processed than the output titles in PRODCOM 161. However, a plausible explanation for these
estimates relates to the composition of the material inputs used for PRODCOM 161 and PROD-
COM 162. The materials used in manufacturing PRODOM 161 are raw in the sense that they
have barely been processed before being transported to the plant. At least some of the materi-
als and intermediate products used in manufacturing PRODCOM 162, on the other hand, are
products of other industries. In other words, some labour and capital inputs have already been
used in manufacturing the material inputs for PRODCOM 162, raising its value. As a con-
sequence, β Mi for PRODCOM 162 is also rather high. The output elasticity of capital is of
the same magnitude for PRODCOM 161 and PRODCOM 162 (β K i is 0.20 and 0.18, respec-
tively). Returns to scale are different across the products in PRODCOM 161 and PRODCOM
162: the technology for the products in PRODCOM 161 is subject to decreasing returns to
scale (β Mi + β Li + β K i = 0.93 < 1), while the technology for the titles in PRODCOM 162 has
increasing returns to scale (β Mi + β Li + β K i = 1.04 > 1). In short, the various products in the
product groups “Sawmilling and planing of wood” and “Products of wood, cork, straw, and plait-
ing materials”, which many multiproduct firms produce simultaneously, are not manufactured
with a single production technology.
The demand for the titles in PRODCOM 161 is more price elastic than the demand for the
titles in PRODCOM 162, as ηi for the titles within PRODCOM 161 is −1.29, while ηi for the
titles within PRODCOM 162 is −1.13. This is intuitive because products of wood, cork, straw,
and plaiting materials are likely to be more differentiated than the output of sawmilling and
planing of wood. Hansen’s J-test does not reject the null hypothesis of valid overidentification
restrictions (Prob[Chi-sq.(26442 )>J] is 0.4632).

40. The starting values for β Mi , β Li , and β K i range between 0.15 and 0.5, and the starting values for ηi between
−8 and −1.5.
41. The forty-two constants β0i and the parameters governing the productivity process g(ωi jt−1 ) are not
reported.
42. The number of moment conditions is 272.
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3335
TABLE 4
Parameter estimates
Parameter estimate (standard error)

PRODCOM 161 PRODCOM 162


Materials 0.37 0.73
(0.008) (0.002)
Labour 0.36 0.13
(0.011) (0.003)

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


Capital 0.20 0.18
(0.008) (0.003)
Price elasticity of demand −1.29 −1.13
(0.020) (0.004)
Prob[Chi-sq.(264) > J] 0.4632
Number of obs. 2053
PRODCOM 161 Sawmilling and planing of wood.
PRODCOM 162 Products of wood, cork, straw, and plaiting materials.

6. DISCUSSION
In Section 6.1, I discuss the assumptions and data requirements underlying the identification
and compare them to the assumptions and requirements of other structural production func-
tion papers. In Section 6.2, I discuss the extent to which the empirical strategy accommodates
economies of scope and differentiated products.

6.1. Identification
Most of the assumptions underlying the identification strategy are familiar from the struc-
tural production function literature (e.g. Olley and Pakes, 1996; Levinsohn and Petrin, 2003;
Doraszelski and Jaumandreu, 2013; Ackerberg et al., 2015; Gandhi et al., 2020). I also make
some novel assumptions. One of the key assumptions in all the estimation strategies concerns
the timing of the input choices with respect to the evolution of the productivity process. As in
other structural production function models, at least one flexible (product-level) input is required
to solve the firm’s static profit maximisation problem. In this paper, firm-level labour is consid-
ered to be a predetermined input. The reason for this is the environment in which the data have
been generated: employment protection legislation plays a significant role in Finland.43 How-
ever, the empirical model may be adjusted to allow for flexible labour input if the wage (level)
is observable and the assumptions concerning materials input and the exogeneous price of mate-
rials are also plausible for labour and wage. In that case, instead of L jt and L jt−1 , wage and
lagged wage are valid instruments for identifying β Li .
I further specify that, while the firm-level L jt and K jt are predetermined, the product-level
allocations of labour and capital, L i jt and K i jt , are flexible. This assumption not only facilitates
solving for the product-level L i jt and K i jt but also allows firms to reallocate labour and capital in
response to demand and productivity shocks. Whether the assumption of zero adjustment costs

43. The OECD indicators of employment protection (OECD, 2013) measure the strictness of legislation on indi-
vidual and collective dismissals and the strictness of hiring employees on temporary contracts. The measures are based
on information about statutory and case law, collective bargaining agreements, and advice from officials from OECD
member countries and country experts. According to these indicators, the Finnish labour market was at the OECD
average in terms of the strictness of employment protection during the 2004–2011 period. Based on this measure,
predetermined labour input is a realistic assumption.
3336 REVIEW OF ECONOMIC STUDIES

holds depends on the set of technologies the firm uses and whether the firm is a single-plant firm
or observed at the plant level. While most multiproduct firms manufacture products in a single
industry, a considerable share of capital and labour is likely to be applicable to several product
lines, even if predetermined at the plant level. On the other hand, the adjustments producers
are willing to make in response to demand and productivity shocks are likely to be moderate in
magnitude. Hence, the assumption of costless reallocation is likely to be realistic to the extent
that single-plant firms, or multiplant firms observed at the plant-level, active in a single industry
are willing to make reallocations due to demand and productivity shocks.

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


One more difference in the assumptions on timing and residuals between this and other struc-
tural estimation strategies is that I assume away productivity shocks once the product-level inputs
have been set. I also assume away measurement error in output Q i jt . I make these assumptions
in order to solve for the unobservable input allocations while controlling for the unobservable
productivity ωi jt .
In addition to the timing assumptions, the proxy methods are based on the assumption that
productivity ωi jt is the only scalar unobservable that affects the proxy variable. Unobservable
inter-firm variation in, say, the input price is assumed away. I also need to make the scalar
unobservable assumption. But in contrast to the other estimation strategies, the materials price44
P jtM is required for identifying β Mi . In this model, the scalar unobservability assumption implies
that P jtM does not depend on the quantity of materials, Mi jt . This assumption is plausible for
industries in which firms do not have monopsony power in the raw materials markets; that is, for
firms that are not large relative to the materials markets.45
In contrast to most of the other structural methods, this paper requires the econometrician
to observe the output prices and specify the output demand function. Moreover, the structural
method is applicable only for firms that are not price-takers in the output market. Because dif-
ferences in the product-firm-specific demand level, αi j , and product-firm-time-specific demand
shocks, εi jt , generate within-firm variation in the profitability of manufacturing various prod-
ucts, the product-level input allocations and production function parameters can be identified
with valid instruments. However, the econometrician has to identify only the own-price elas-
ticities of the products that the firm produces, even when substitution and complementarity
between different products and products of different producers exist, as shown in Supplementary
Material, Appendix A. The demand function estimated in the empirical application is isoelas-
tic, but the product-level inputs are solved also for linear and exponential demand functions in
Supplementary Material, Appendix A.
I make the assumption of monopolistic competition; that is, the assumption that price changes
in substitutes and complementary products have only a negligible effect on the demand for a
given product of a specific producer. This assumption is likely to be most appropriate in an
industry where a large number of firms produces differentiated products. Assuming isoelastic
demand, as in this paper, implies that for a given product, price-cost markups are constant across
producers. Alternative functional forms of demand, such as the exponential and linear demand
considered in Supplementary Material, Appendix A, break the assumption of constant price-cost

44. In the wood products industry considered in the present study, input prices vary across plants because of
transportation costs that constitute a considerable share of the cost of wood purchases.
45. This assumption is plausible, at least, in the wood products industry. First, the most important material for
wood product plants or firms is raw wood. The majority of raw wood is purchased by the paper and paper manufacturing
industry. Thus, demand shocks for wood products are not likely to have an important effect on wood prices. Second, the
Natural Resources Institute Finland (Luke) publishes monthly statistics on wood prices. These statistics are based on
more than 90% of the volume purchased by the forest industry. This suggests that buyers of raw wood are well informed
about the current price levels of various raw wood types, leaving little room for negotiation.
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3337

markups for a given product. Identification of the isoelastic demand function of the empirical
application is based on two assumptions. First, the level of demand for a given product of a
given firm is constant over time, except for mean independent demand shocks. If appropriate,
potential industry-level changes in demand can be accounted for by estimating product-time-
specific fixed effects, as shown in Supplementary Material, Appendix A. Second, changes in
input prices and predetermined input stocks shift the supply curve, while the demand curve,
including the demand shock εi jt , is not affected. Using material prices and predetermined input
stocks as instruments is a standard practice.

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


The estimation strategy takes into account neither firm nor product selection that takes place
due to the expected profitability of manufacturing any particular product in the industry. The
demand-side factor that affects the profitability of manufacturing product i is the product-firm
specific level of demand αi j . The supply-side factors are, for example, the expected product-
specific productivity level the firm would have for the product, as well as the stock of capital
and labour the firm employs at t − 1. As is well attested in the literature, firm selection leads to
firm productivity and capital being negatively correlated among the observations that are some
distance above the threshold of market entry and exit (e.g. Wedervang, 1965). If this is not taken
into account, the estimated output elasticity of capital is biased downwards. In general, large,
established firms have a low likelihood of exit, so selection is expected to be less of a problem
when the data are on large firms.46
Product selection regarding supply-side factors may be a concern for the following reason.
A large stock of firm-level capital or labour may enable the firm to profitably open, or not close,
a product line even if its productivity is low. As equations (13) and (14) show, the product-level
K i jt and L i jt increase in the firm-level K jt and L jt , respectively. Hence, it is possible that the
product-level K i jt or L i jt is negatively correlated with the product-specific ωi jt . However, firms
are likely to add (drop) products to (from) their product selection in descending (ascending) order
of ωi jt . This implies that the more multiproduct firms there are, or the more products firms have
in their assortments,47 the lower the correlation between ωi jt and K i jt or L i jt is likely to be.48
Infrequent changes in firms’ product assortments is another indication of the low probability of
product-level entry and exit in the data.49
Product selection also requires consideration of output demand. Recall from Section 3.3 that
αi j is an incidental parameter, and therefore consistent estimation of ηi requires an instrument
that is uncorrelated with αi j . Materials price P jtM is assumed to be exogenous and hence uncor-
related with αi j . However, if P jtM is correlated over time, it is possible that P jtM is positively
correlated with αi j among the observations close to the product-level entry–exit threshold. In
addition, for this reason the estimation strategy is suitable for a setting where the probability

46. According to the two data sets I use, the annual exit rate of plants in the wood products industry is 5.6%.
Unfortunately, the data set is not informative about whether the exits from the data are real plant closures or whether the
plants have not replied to the survey. Most of the plants are large, though, which suggests that the probability of plant
exit is low.
47. I use data on 904 plant-year observations in which 659 plants produce at least two products; the average
scope of all plants is 3.25 products.
48. De Loecker et al. (2016) estimate product-specific production functions using a panel of single-product firms,
some of which may be multiproduct firms in other periods. They estimate the production functions correcting for the
selection to become a multiproduct firm using the insights of Olley and Pakes (1996) and compare the results to those
estimated without correcting for such selection. They conclude that “the use of the unbalanced panel of single-product
firms (which includes firms that are always single-product and firms that ultimately transition to a multi-product status)
likely alleviates most of the concerns about the selection bias” (De Loecker et al., 2016, p. 483).
49. I use data where product entry rate, excluding plants entering the data, is 10%, and product exit rate, excluding
plants exiting the data, is 7%.
3338 REVIEW OF ECONOMIC STUDIES

of change in the firm’s product assortment is low. Alternatively, the model may be extended to
control for, first, firm selection and, second, selection into various product lines by computing
propensity scores for entry, as in Olley and Pakes (1996).
The production function parameters are estimated by solving a non-linear optimisation prob-
lem. This implies that the parameters may not be estimated as specific to the most detailed
product classification available, as discussed in Section 4.4. In the empirical example in the
present study, the production functions are estimated at the eight- or 10-digit level of the PROD-
COM classification, but the output elasticities of inputs are estimated as specific to the three-digit

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


classification. The implication of this restriction is that the elasticity estimates are weighted aver-
ages of the true product-specific elasticities within the three-digit product class. Consequently,
the estimated product-level inputs and the implied productivity levels are also determined by,
among other factors, weighted averages of the true product-specific elasticities within the three-
digit product class. The extent to which the restriction of constant elasticities within a product
class may bias the production function estimates depends on the true product-specific elasticities.
Presumably, the more detailed the product class definitions, the more similar the true elasticities
within a given class are likely to be and the smaller any potential estimation biases. It is impor-
tant to note, however, that the elasticity estimates obtained under these parameter restrictions are
likely to be different than the estimates that would be obtained by estimating production func-
tions for outputs and inputs aggregated to the three-digit level. An essential difference between
these two cases is that when estimating production functions for aggregated outputs and inputs,
returns to scale are specified to take place at the aggregate level, which may be a misspecifica-
tion and one more source of estimation bias. For example, in the case of aggregated output and
inputs, elasticity estimates are biased even if the true production functions are identical across
product lines but do not have constant returns to scale.
The estimation strategy is parametric, which requires the econometrician to carefully con-
sider the choices of functional form. As shown in Supplementary Material, Appendix A, the
identification strategy is not restricted to the Cobb–Douglas production function and the isoelas-
tic demand function estimated in the empirical application. The solutions for product-level input
allocations when production technology takes the form of a translog, another common func-
tional form in production function estimation, and demand is linear, log-linear, or exponential,
are provided in Supplementary Material, Appendix A. The requirement of the functional form is
that it has to allow for closed-form solutions for product-level inputs. An example of a produc-
tion technology that does not allow for this is the constant elasticity of substitution production
function.

6.2. Industry characteristics


The model assumptions and data requirements underlying the estimation strategy are more suit-
able for some industries than others. Two model assumptions are about economies of scope and
product characteristics.

6.2.1. Economies of scope and flexible manufacturing. Economies of scope may arise
due to lower average fixed or average variable costs of production. Average fixed costs decrease
when the producer’s fixed costs are spread over a greater number of products. For example, open-
ing a new product line does not necessarily require larger premises and thus higher rental costs.
The estimation strategy of the present study assumes away fixed costs, as usual in production
function estimation.
Economies of scope due to lower variable costs arise if, for example, the production process
of a product yields by-products that can be used as inputs for other products. Alternatively, there
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3339

can be some units of inputs that are joint between product lines, leading to a decrease in average
variable costs. These kinds of inputs cannot be solved for when only firm-level inputs are observ-
able. The estimation strategy therefore assumes away the aforementioned types of economies of
scope and is not suitable for industries where such economies of scope are plausible. However,
not allowing for economies of scope in the form of lower variable costs is in line with the theory
of flexible manufacturing (e.g. Eckel and Neary, 2010). Instead, the estimation strategy allows
firms to have productivity differences between product lines and hence core competencies in
making some products, which is an essential feature of flexible manufacturing, as discussed in

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


the Introduction. Whether the assumption of flexible manufacturing holds depends on the par-
ticular industry. For example, joint inputs in petroleum refining lead to lower variable costs and
economies of scope, so the estimation strategy is not appropriate. The estimation strategy is,
however, compatible with modelling productivity as a function of the producer’s product scope
and thus allowing for (dis)economies of scope in the form of higher (lower) product-specific total
factor productivity and hence (dis)economies of scope due to variable costs, as in De Loecker
et al. (2016).

6.2.2. Product heterogeneity. The estimation strategy is suitable for industries where com-
petition is imperfect, where product differentiation is likely. Hence, the product classification
has to be detailed enough for the physical output quantities of a given product to be compara-
ble across manufacturers.50 At the same time, the data have to include a sufficient number of
observations of the product for identification to be possible. If the physical quantities of a prod-
uct cannot be compared due to unobservable quality variation or different units of output, the
estimation strategy is not suitable for that industry.
A large share of manufacturing industries make products that are comparable in physical
units even across manufacturers. Examples include food products, textiles, paper and paper
products, printing and reproduction of recorded media, chemicals and chemical products, basic
pharmaceutical products and pharmaceutical preparations, rubber and plastic products, other
non-metallic mineral products, basic metals, and fabricated metal products. Industries that do
not make products comparable in physical units are likely to manufacture products that involve
advanced technology. Certain transport equipment such as ships and boats, or air and spacecraft,
are good examples. If an econometrician wished to estimate production functions for cars, for
example, (s)he should consider whether subcompact cars such as the Citroë n C1 and Renault
Twingo, or entry-level luxury cars such as the Audi A4 and Mercedes-Benz C-Class, are com-
parable in physical units, and if not, whether the product classes can be defined more narrowly
without making the number of observations in product classes too small.
In short, product differentiation is not a challenge to identification as long as quality differ-
ences are observable, or if differentiation is horizontal and does not require different production
technologies. Ultimately, this requirement is a matter of whether the output data are sufficient
regarding the output characteristics. Many national statistics authorities provide data sets that are
likely to be sufficiently rich for several manufacturing industries. There are also several manu-
facturing industries, such as automakers, whose output is not reported in sufficient detail in these
data sets and would have to be complemented with additional output data such as car type and
characteristics.

50. In the empirical example used in the present study, unobservable quality differences or different units of
output are not admissible for a product defined at the eight- or 10-digit level of the PRODCOM classification.
3340 REVIEW OF ECONOMIC STUDIES

7. CONCLUSION
This paper contributes to the substantial empirical literature on production function estimation,
which underlies an even larger body of applied economic research. The standard assumption
made in production function estimation is that firms produce all of their output with a firm-level
production technology that is independent of the firm’s product set and that the firm is equally
productive in manufacturing all of its products. However, an empirical fact is that a remarkable
share of firms are multiproduct firms, and it is possible that they use several, product-specific

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


production technologies. The empirical literature, with a few exceptions, disregards this fact in
production function estimation because data sets do not report how firms allocate their inputs to
the various product lines. Building on previous structural estimation strategies, I provide a strat-
egy for estimating product-specific production functions of firms that are not price-takers in the
output market. The empirical model does not require data on input allocations to various prod-
uct lines but does require data on firms’ output prices. The optimal input allocations are solved
parametrically by, first, specifying the output demand and the firm’s static profit maximisation
problem. The unobservable productivity terms are then substituted for by using the definition
of productivity and the observable output quantities. While differences in output demand and
product-firm-time-specific demand shocks generate within-firm variation in the profitability of
manufacturing various products, supply- and demand-side instruments are used to identify the
product-level input allocations and the production function parameters. The empirical model
also accommodates productivity differences between a firm’s product lines, which the empiri-
cal literature suggests is important in determining firms’ production and product market entry
decisions and hence market outcomes.
I demonstrate the estimation strategy by estimating product-specific production functions
for products in the industry “Manufacture of wood and of products of wood and cork, except
furniture; manufacture of articles of straw and plaiting materials”. I find that the technologies
used in “Sawmilling and planing of wood” (PRODCOM 161) and “Products of wood, cork,
straw and plaiting materials” (PRODCOM 162) are statistically different from each other. These
findings show that, at least in this industry, the single-product technology assumption should not
be imposed on multiproduct firms.
Several questions remain for future research. First, there are industries where production
technologies are joint because product lines have joint inputs or because by-products of a
product line are used as inputs in another product line. Applied econometricians would appre-
ciate identification strategies that accommodate these kinds of economies of scope. Second,
while it is important to consider endogenous product selection when estimating product-specific
production functions, those estimates, likewise, are useful for studying how firms select their
products.
Acknowledgments. I thank Aureo de Paula for carefully reading the manuscript and providing helpful comments.
I am grateful to my PhD thesis supervisor Otto Toivanen for his constructive comments and support. I also thank
Dan Ackerberg, Liran Einav, Jeremy Fox, Pinelopi Goldberg, Pekka Ilmakunnas, Peter Nyberg, David Rivers, Matti
Sarvimäki, Paul Scott, Valerie Smeets, Janne Tukiainen, Fr édèric Warzynski, and three anonymous referees for valuable
comments and discussions, along with seminar participants at the Helsinki Center of Economic Research, the European
Economic Association Annual Congress 2013, and the 11th CEPR School on Applied Industrial Organization.

Supplementary Data
Supplementary data are available at Review of Economic Studies online.

Data Availability Statement


The data are confidential micro data from Statistics Finland. The data are compiled of the following two data sets:
(1) industrial output for the years 2004–2011 (Statistics Finland, 2012a) and (2) LDPM for the years 2004–2011
VALMARI ESTIMATING PRODUCTION FUNCTIONS OF MULTIPRODUCT FIRMS 3341

(Statistics Finland, 2012b). These data sets are provided by the Research Services department of Statistics Finland.
Statistical legislation and data protection and confidentiality practices specified in legislation are applied in releasing
the data. Release of data is subject to a user licence. Instructions for applying for a user licence are provided here:
https://www.tilastokeskus.fi/tup/mikroaineistot/hakumenettely en.html. The replication package can be accessed here:
http://doi.org/10.5281/zenodo.7474464.

REFERENCES
ACKERBERG, D., BENKARD, C. L., BERRY, S. and PAKES, A. (2007), “Chapter 63 Econometric Tools for Analyzing

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


Market Outcomes,” in Heckman, J. L. and Leamer, E. E. (eds) Handbook of Econometrics (Vol. 6) (Amsterdam:
Elsevier) 4171–4276.
ACKERBERG, D., CAVES, K. and FRAZER, G. (2015), “Identification Properties of Recent Production Function
Estimators”, Econometrica, 83, 2411–2451.
ARKOLAKIS, C., GANAPATI, S. and MUENDLER, M.-A. (2021), “The Extensive Margin of Exporting Products: A
Firm-Level Analysis”, American Economic Journal: Macroeconomics, 13, 182–245.
BALDWIN, J. R., CAVES, R. and GU, W. (2005), “Responses to Trade Liberalization: Changes in Product Diversifi-
cation in Foreign- and Domestic-controlled Plants,” Economic Analysis (EA) Research Paper Series 31, Statistics
Canada, Analytical Studies Branch.
BERNARD, A. B., JENSEN, J. B., REDDING, S. J. and SCHOTT, P. K. (2007), “Firms in International Trade”, Journal
of Economic Perspectives, 21, 105–130.
BERNARD, A. B., REDDING, S. J. and SCHOTT, P. K. (2010), “Multiple-Product Firms and Product Switching”,
American Economic Review, 100, 70–97.
BERNARD, A. B., REDDING, S. J. and SCHOTT, P. K. (2011), “Multiproduct Firms and Trade Liberalization”,
Quarterly Journal of Economics, 126, 1271–1318.
BROWN, R. S., CAVES, D. W. and CHRISTENSEN, L. R. (1979), “Modelling the Structure of Cost and Production for
Multiproduct Firms”, Southern Economic Journal, 46, 256–273.
CAVES, D. W., CHRISTENSEN, L. R. and TRETHEWAY, M. W. (1980), “Flexible Cost Functions for Multiproduct
Firms”, Review of Economics and Statistics, 62, 477–481.
COLLARD-WEXLER, A. and LOECKER, J. D. (2015), “Reallocation and Technology: Evidence from the US Steel
Industry”, American Economic Review, 105, 131–171.
DE LOECKER, J. (2011), “Product Differentiation, Multiproduct Firms, and Estimating the Impact of Trade Liberal-
ization on Productivity”, Econometrica, 79, 1407–1451.
DE LOECKER, J., GOLDBERG, P. K., KHANDELWAL, A. K. and PAVCNIK, N. (2016), “Prices, Markups, and Trade
Reform”, Econometrica, 84, 445–510.
DHYNE, E., PETRIN, A., SMEETS, V. and WARZYNSKI, F. (2020), “Theory for Extending Single-product Production
Function Estimation to Multi-product Settings” (Working Paper).
DHYNE, E., PETRIN, A., SMEETS, V. and WARZYNSKI, F. (2021), “Multi-product Firms, Import Competition, and
the Evolution of Firm-product Technical Efficiencies” (Working Paper).
DORASZELSKI, U. and JAUMANDREU, J. (2013), “R&D and Productivity: Estimating Endogenous Productivity”,
Review of Economic Studies, 80, 1338–1383.
ECKEL, C. and NEARY, J. P. (2010), “Multi-product Firms and Flexible Manufacturing in the Global Economy”,
Review of Economic Studies, 77, 188–217.
FLACH, L. and IRLACHER, M. (2018), “Product Versus Process: Innovation Strategies of Multiproduct Firms”,
American Economic Journal: Microeconomics, 10, 236–277.
FOSTER, L., HALTIWANGER, J. and SYVERSON, C. (2008), “Reallocation, Firm Turnover, and Efficiency: Selection
on Productivity or Profitability?”, American Economic Review, 98, 394–425.
GANDHI, A., NAVARRO, S. and RIVERS, D. (2020), “On the Identification of Gross Output Production Functions”,
Journal of Political Economy, 128, 2973–3016.
GARCIA-MARIN, A. and VOIGTLANDER, N. (2019), “Exporting and Plant-Level Efficiency Gains: It’s in the
Measure”, Journal of Political Economy, 127, 1777–1825.
GRIECO, P. L. E., LI, S. and ZHANG, H. (2016), “Production Function Estimation with Unobserved Input Price
Dispersion”, International Economic Review, 57, 665–690.
HILL, R. J. (2004), “Constructing Price Indexes Across Space and Time: The Case of the European Union”, American
Economic Review, 94, 1379–1410.
KLETTE, T. J. and GRILICHES, Z. (1996), “The Inconsistency of Common Scale Estimators When Output Prices Are
Unobserved and Endogenous”, Journal of Applied Econometrics, 11, 343–361.
LEVINSOHN, J. and PETRIN, A. (2003), “Estimating Production Functions Using Inputs to Control for Unobserv-
ables”, Review of Economic Studies, 70, 317–341.
MALIKOV, E. and LIEN, G. (2021), “Proxy Variable Estimation of Multiproduct Production Functions”, American
Journal of Agricultural Economics, 103, 1878–1902.
MAYER, T., MELITZ, M. and OTTAVIANO, G. I. P. (2014), “Market Size, Competition, and the Product Mix of
Exporters”, American Economic Review, 104, 495–536.
MUNDLAK, Y. (1963), “Specification and Estimation of Multiproduct Production Functions”, Journal of Farm
Economics, 45, 433–443.
3342 REVIEW OF ECONOMIC STUDIES

MUNDLAK, Y. (1964), “Transcendental Multiproduct Production Functions”, International Economic Review, 5,


273–284.
NEARY, P. J. (2004), “Rationalizing the Penn World Table: True Multilateral Indices for International Comparisons of
Real Income”, American Economic Review, 94, 1411–1428.
NEWEY, W. K., POWELL, J. L. and VELLA, F. (1999), “Nonparametric Estimation of Triangular Simultaneous
Equations Models”, Econometrica, 67, 565–603.
OECD (2013), “OECD Indicators of Employment Protection” https://www.oecd.org/els/emp/oecdindicatorsofemploy
mentprotection.htm.
OLLEY, G. S. and PAKES, A. (1996), “The Dynamics of Productivity in the Telecommunications Equipment Industry”,
Econometrica, 64, 1263–1297.

Downloaded from https://academic.oup.com/restud/article/90/6/3315/7005276 by guest on 03 November 2024


ORR, S. (2022), “Within-Firm Productivity Dispersion: Estimates and Implications”, Journal of Political Economy, 130,
2771–2828. https://www.journals.uchicago.edu/doi/abs/10.1086/720465?journalCode=jpe.
PANZAR, J. C. (1989), “Chapter 1 Technological Determinants of Firm and Industry Structure”, in Schmalensee, R.
and Willig, R. (eds) Handbook of Industrial Organization (Vol. 1, 1st edn) (Amsterdam: Elsevier) 3–59.
QIU, L. D. and ZHOU, W. (2013), “Multiproduct Firms and Scope Adjustment in Globalization”, Journal of
International Economics, 91, 142–153.
Statistics Finland (2012a), “Industrial Output” https://stat.fi/en/statistics/documentation/tti.
Statistics Finland (2012b), “Longitudinal Database on Plants in Finnish Manufacturing (LDPM)”.
WEDERVANG, F. (1965), Development of a Population of Industrial Firms: The Structure of Manufacturing Industries
in Norway, 1930–1948 (Oslo: Scandinavian University Books (Universitetsforlaget)).

You might also like