Missing Growth From Creative Destruction: American Economic Review 2019, 109 (8) : 2795-2822

American Economic Review 2019, 109(8): 2795–2822
https://doi.org/10.1257/aer.20171745
Missing Growth from Creative Destruction†
By Philippe Aghion, Antonin Bergeaud, Timo Boppart,

Peter J. Klenow, and Huiyu Li*
For exiting products, statistical agencies often impute inflation from

surviving products. This understates growth if creatively-destroyed
products improve more than surviving ones. If so, then the market
share of surviving products should systematically shrink. Using
entering and exiting establishments to proxy for creative destruc-
tion, we estimate missing growth in US Census data on non-farm
businesses from 1983 to 2013. We find missing growth (i) equaled
about one-half a percentage point per year; (ii) arose mostly from
hotels and restaurants rather than manufacturing; and (iii) did not
accelerate much after 2005, and therefore does not explain the sharp
slowdown in growth since then. (JEL E23, E31, L14, L15, O30, O41)
Whereas it is straightforward to compute inflation for an unchanging set of goods

and services, it is much harder to separate inflation from quality improvements and
variety expansion amidst a changing set of items. In the US Consumer Price Index
(CPI), over 3 percent of items exit the market each month (Bils and Klenow 2004).
In the Producer Price Index (PPI) the figure is over 2 percent per month (Nakamura
and Steinsson 2008).
The Boskin Commission (Boskin et al. 1996) highlighted the challenges of mea-
suring quality improvements when incumbents upgrade their products. It also main-
tained that the CPI does not fully capture the benefits of brand new varieties. We
argue that there exists a subtler, overlooked bias in the case of creative destruction.
When the producer of the outgoing item does not produce the incoming item, the
standard procedure at statistical offices is to resort to some form of imputation.
Imputation inserts the average price growth among a set of surviving products that
* Aghion: Collège de France, 3 Rue d’Ulm, 75005 Paris, France, and London School of Economics (email:
[email protected]); Bergeaud: Banque de France, 31 Rue Croix des Petits Champs, 75001 Paris, France (email:
[email protected]); Boppart: Institute for International Economic Studies, Stockholm University,
SE-106 91 Stockholm, Sweden (email: [email protected]); Klenow: Department of Economics, Stanford
University, Stanford, CA 94305 (email: [email protected]); Li: Federal Reserve Bank of San Francisco,
Economics Research, 101 Market Street, San Francisco, CA 94105 (email: [email protected]). Gita Gopinath
was the coeditor for this article. We thank Raouf Boucekkine, Pablo Fajgelbaum, Kevin Fox, Bart Hobijn, Colin
Hottman, and Stephen Redding for excellent discussions and Victoria De Quadros for superb research assistance.
Ufuk Akcigit, Robert Feenstra, Xavier Jaravel, Chad Jones, Per Krusell, Torsten Persson, Ben Pugsley, John Van
Reenen, four referees, and numerous seminar participants provided helpful comments. Any opinions and conclu-
sions expressed herein are those of the authors and do not necessarily represent the views of the Federal Reserve
System, the US Census Bureau, the Bank of France, or the Eurosystem. All results have been reviewed to ensure
that no confidential information is disclosed. The authors have no relevant or material financial interests that relate
to the research in this paper.
†
Go to https://doi.org/10.1257/aer.20171745 to visit the article page for additional materials and author
disclosure statements.
2795
2796 THE AMERICAN ECONOMIC REVIEW AUGUST 2019
were not creatively destroyed.1 We think this misses some growth because inflation
is likely to be below-average for items subject to creative destruction.2
Creative destruction is believed to be a key source of economic growth. See
Aghion and Howitt (1992); Akcigit and Kerr (2018); and Aghion, Akcigit,
and Howitt (2014). We therefore attempt to quantify the extent of “missing growth,”
the difference between actual and measured productivity growth, due to the use of
imputation in cases of creative destruction. Our estimates are for the US nonfarm
business sector over the past three decades.
In the first part of the paper we develop a growth model with (exogenous) inno-
vation to provide explicit expressions for missing growth. In this model, innova-
tion may either create new varieties or replace existing varieties with products of
higher quality. The quality improvements can be performed by incumbents on their
own products, or by competing incumbents and entrants (creative destruction). The
model predicts missing growth due to creative destruction if the statistical office
resorts to imputation.
In the second part of the paper we estimate the magnitude of missing growth
based on our model. We use microdata from the US Census on employment at all
private nonfarm businesses to estimate missing growth from 1983 to 2013. We look
at employment shares of continuing (incumbent), entering, and exiting plants. If
new plants produce new varieties and carry out creative destruction, then the inroads
they make in incumbents’ market share should signal their contribution to growth.
Our findings can be summarized as follows. First, missing growth from impu-
tation is substantial: roughly one-half a percentage point per year, or around one-
third of measured productivity growth. Second, missing growth is concentrated in
hotel, restaurants, and retail trade rather than manufacturing. Third, missing growth
accelerated only modestly after 2005, an order of magnitude smaller than needed to
explain the slowdown in measured growth.
Example: The following numerical example illustrates how imputation can miss
growth. Suppose that (i) 80 percent of products in the economy experience no inno-
vation in a given period and are subject to a 4 percent inflation rate; (ii) 10 percent
of products experience quality improvement without creative destruction, with their
quality-adjusted prices falling 6 percent (i.e., an inflation rate of −6 percent); and
(iii) 10 percent of products experience quality improvement due to creative destruc-
tion, with their quality-adjusted prices also falling by 6 percent. The true inflation
rate in this economy is then 2 percent. Suppose further that nominal output grows
at 4 percent, so that true productivity growth is 2 percent after subtracting the 2 per-
cent true inflation rate. What happens if the statistical office resorts to imputation
in cases of creative destruction? Then it will not correctly decompose growth in
nominal output into its inflation and real growth components. Imputation means
1
US Government Accountability Office (1999) details CPI procedures for dealing with product exit. For the
PPI, “If no price from a participating company has been received in a particular month, the change in the price
of the associated item will, in general, be estimated by averaging the price changes for the other items within the
same cell for which price reports have been received” (US Bureau of Labor Statistics 2015, p.10). The BLS makes
explicit quality adjustments, such as using hedonics, predominantly for goods that undergo periodic model changes
by incumbent producers (Groshen et al. 2017).
2
A similar bias due to creative destruction could arise at times of regular rotation of items in the CPI and PPI
samples.
VOL. 109 NO. 8 AGHION ET AL.: MISSING GROWTH FROM CREATIVE DESTRUCTION 2797
that the statistical office will ignore the goods subject to creative destruction when
computing the inflation rate for the whole economy, and only consider the products
that were not subject to innovation plus the products for which innovation did not
involve creative destruction. Thus, the statistical office will take the average inflation
rate for the whole economy to be equal to
 _
1   ⋅ (− 6%) = 2.9%.
8   ⋅  4% +  _
9 9
Presuming it correctly evaluates the growth in nominal GDP to be 4 percent, the

statistical office will (incorrectly) infer that the growth rate of real output is
4% − 2.9% = 1.1%.
This in turn implies “missing growth” in productivity amounting to
2% − 1.1% = 0.9%.
This ends our example, which hopefully clarifies the main mechanism by which
imputation can miss growth from creative destruction.3
We want to clarify two things in the context of this example. First, items can be
goods or services. The BLS prices not only Universal Product Codes (UPC) at grocery
and drug stores, but also rooms at hotels, dishes on restaurant menus, services pro-
vided by auto repair shops, and treatments at dental and medical facilities. An individ-
ual item is priced over time at a given outlet (location). Second, when an entire outlet
in the BLS sample exits the market, the BLS imputes inflation for all of the items in
the exiting establishment. This is true in both the CPI and PPI. Our empirical approach
hopes to capture creative destruction which results in establishment exit altogether and
which might be missed due to BLS imputation at these exit points.
Our estimates are therefore a form of outlet bias. Boskin et al. (1996), Hausman
and Leibtag (2009), and Moulton (2018) describe outlet bias as when a new outlet
sells an identical item to that in an existing store, only at a lower price. The BLS
does not compare prices across stores, so it misses such price reductions associated
with new outlets. Our notion of outlet bias is much broader, as we do not require the
new outlets to sell identical items to existing stores. The new outlets may sell better
quality versions, or a wider variety of items. And outlet bias can occur in all sectors,
not just grocery stores and mass merchandisers. This distinction can help explain why
our estimate of outlet bias (54 basis points per year) is over five times larger than the
estimates of Boskin et al. (1996) and Moulton (2018) (10 and 8 basis points per year,
respectively).
Our paper touches recent literatures on secular stagnation and growth mea-
surement. Gordon (2012) argues that innovation has run into diminishing returns,
inexorably slowing total factor productivity (TFP) growth.4 Syverson (2017) and
3
This example is stylized. In practice, imputation by the BLS is carried out within the items category or
category-region. See the US Government Accountability Office (1999).
4
Related studies include Jones (1995), Kortum (1997), and Bloom et al. (2018).
Byrne, Fernald, and Reinsdorf (2016) conclude that understatement of growth in

the information and communications technology (ICT) sector cannot account for
the productivity slowdown since 2005. In contrast to these latter studies, we look at
missing growth for the whole economy.
More closely related to our analysis are Feenstra (1994); Bils and Klenow (2001);
Bils (2009); Broda and Weinstein (2010); Erickson and Pakes (2011); Byrne, Oliner,
and Sichel (2015); and Redding and Weinstein (2018).5 We make two contributions
relative to these important papers. First, we compute missing growth for the entire
private nonfarm sector from 1983 to 2013.6 Second, we focus on a neglected source
of missing growth, namely imputation in the event of creative destruction. The miss-
ing growth we identify is likely to be exacerbated when there is error in measuring
quality improvements by incumbents on their own products.7
In Section I we lay out an environment relating missing growth to creative
destruction. In Section II we estimate missing growth using US Census data on plant
market shares. Section III compares our estimates to those in three existing studies.
Section IV concludes.
I. A Model of Missing Growth
In this section we relate our measure of missing growth to Feenstra (1994). We

modify his approach to focus on bias from new producers of a given product line, as
opposed to all new products. This allows us to highlight the role of creative destruc-
tion in missing growth, as opposed to quality bias from incumbent improvement of
their own products.
A. Relating Missing Growth to Market Share Dynamics
Time is discrete and in each period consumption has a constant elasticity of sub-
stitution (CES) structure
  σ    
_
Ct = (∫0   [qt( j) ct( j)]  dj) 
σ−1
Nt   σ−1
_     

(1) σ
,
where c t( j)denotes quantity and qt( j)the quality of variety j. The variable Ntis the
number of varieties available, which can change over time. Here σ > 1denotes the
constant elasticity of substitution between varieties.
Feenstra (1994) corrects for biases in the US import price indices of six manufacturing goods, in particular
5
due to an expanding set of available product varieties. Bils and Klenow (2001) use the US Consumer Expenditure
surveys to estimate “quality Engel curves” and assess the unmeasured quality growth of 66 durable goods which
account for 12 percent of consumer spending. Bils (2009) uses CPI microdata to decompose the observed price
increases of durable goods into quality changes and true inflation. Broda and Weinstein (2010) look at missing
growth from entry and exit of products in the nondurable retail sector, using an AC Nielsen database. Byrne, Oliner,
and Sichel (2015) look at missing growth in the semiconductor sector.
Broda and Weinstein (2010) used AC Nielsen data from 1994 and 1999–2003. This database is heavily
6
weighted toward nondurables, particularly food. Bils and Klenow (2004) report a product exit rate of about 2.4 per-
cent per month for nondurables (1.2 percent a month for food) versus about 6.2 percent per month for durable
goods. Hence, it is important to analyze missing growth across many sectors of the economy, including durables.
Unlike Broda and Weinstein (2010), we do not assume that the BLS makes no effort to quantify such quality
7
improvements. Bils (2009) estimates that the BLS subtracted 0.7 percentage points per year from inflation for dura-
bles over 1988–2006 due to quality improvements. For the whole CPI, Moulton and Moses (1997) calculate that the
BLS subtracted 1.8 percentage points in 1995.
Let πtdenote true inflation, the log first difference between t − 1and tin the min-
imum cost of acquiring one unit of composite C . Following Feenstra (1994) we can
decompose true inflation as
SI,t−1
σ − 1 ( SIt,t )
(2) ˆ t  −  _
πt = π  log _
1        t     ,
where πˆ  tis inflation for a subset of products denoted I t, and S It,τis the share in nom-
inal expenditure spent on this set of products at time τ. Here πˆ  tcan be constructed
from the nominal expenditure shares s It,τ( j) ≡ Sj,τ /SIt,τand quality-adjusted prices
of products pτ( j) / qτ( j)within the subset It as
pt( j)/qt( j)
( pt−1( j)/qt−1( j))

(3) πˆ  t ≡ ∫j∈ It ( { It,t−1( ) It,t( )}i∈It)
      f j, s i , s i log ___________
   
    dj,
where f ( j, {sIt,t−1(i), sIt,t(i)}i∈It)is the Sato-Vartia weight.8

Feenstra (1994) chose Itto be the set of US import category-country pairs that
overlapped between t − 1and t. If the US import price index is captured by π  ˆ t, then the
second term in (2) reflects inflation bias due to the changing set of category-country
pairs. Intuitively, if the market share of the subset of products used for measurement
changes, then this subset is not representative. In particular, if the market share of
the subset Itis shrinking, then the relative price of the subset must be rising. For a
given change in market shares, the bias is smaller when the σ elasticity is higher,
because a smaller relative price change is needed to explain the observe change in
market shares.
Broda and Weinstein (2010) chose Itto be the set of UPC codes in AC Nielsen
scanner data with positive sales in adjacent periods. They argue that quality is fixed
over time for a product with a given UPC code. They therefore infer bias in measured
inflation à la Feenstra from the change in market share of overlapping UPC codes.
This assumes that all quality growth is missed in inflation measurement, including
improvements by incumbents on their own products. In practice, statistical offices
do attempt to measure quality changes and net them out of inflation, in particular for
incumbent product upgrades.9
Our focus is on whether imputation leads to missing growth; in particular, in
cases of creative destruction. We therefore define the subset of products Itas those
produced by the same producer in overlapping periods. This is broader than the
set used by Broda-Weinstein since it includes successive generations of products
in a given line produced by the same firm. The omitted set of goods is exiting and
entering producer-product pairs. As we argue in online Appendix A, the BLS infers
8
Sato-Vartia weights are
s (j) − s (j)
   
   
It,t It,t−1
____________
log sI,t(j) − log sI,t−1(j)
f (j, {sI,t−1(i), sI,t(i)}i∈I)
= _____________
    
    .
∫

t t
sI,t(i) − sI,t−1(i)
i∈  I    
    di
t t t
____________
log sIt,t(i) − log sIt,t−1(i)
t t
See Table 1 in Groshen et al. (2017) and Table B1 in Moulton (2018).

9
inflation for these entering and exit producer-product pairs using the inflation rate
for continuing pairs. If the market share of overlapping pairs is shrinking over time,
however, then the inflation rate of continuing pairs overstates overall inflation and
understates real growth. Thus, when the market share of continuing pairs shrinks,
we infer missing growth from creative destruction and/or brand new varieties.
We make these links explicit in the next subsection by adding a supply side to the
model with growth coming from quality improvements by incumbent producers on
their own products, creative destruction, and new variety creation. In the next sec-
tion we describe how we implement this modified Feenstra approach on US Census
establishment data.
B. Framework of Creative Destruction and Missing Growth
A representative household inelastically supplies L units of labor each period and

maximizes (1) subject to a budget constraint Πt  + Wt L = ∫0 N  t pt( j) ct( j) dj, where Πt
and Wt Ldenotes profits and labor earnings.
On the supply side, each variety c( j)is produced one-for-one with labor
ct( j) = lt( j),

by a monopolistically competitive producer, where lt( j)is the amount of labor used

to produce good jin period t. Given the isoelastic demand that results from the
household problem, it is optimal for each producer to set the price
(4) pt( j) =  _ σ     

W,
σ − 1 t
where W tis the nominal wage that equalizes across firms in the competitive labor
market. We assume the nominal wage follows an exogenously given path over time
which will imply that the other nominal variables like prices, p ( j), profits, Π, and
nominal output will grow at the same rate over time.
We model technical change as product innovation.10 At each point in time, and
for each variety jthere is a common exogenous probability of creative destruction
λd ∈ [0, 1). That is, with probability λdthe incumbent firm of input jis replaced
by a new producer. We assume that the new producer (who may be an entrant or an
incumbent firm) improves upon the previous producer’s quality by a factor γd  >  1.
The previous producer cannot profitably produce due to limit pricing by the new
producer.11 If jis an existing variety where quality is improved upon by a new pro-
ducer, we have
qt( j) = γd qt−1
( j).
10
This modeling choice matters if process innovation is more easily captured by the statistical office. Yet,
across firms and plants with price information in the Census of Manufacturing, we find that firm/plant revenues
increase without a decline in unit prices. This suggests that innovations are rather of the product than of the process
type. Hottman, Redding, and Weinstein (2016) provide similar evidence for retail prices of consumer nondurable
manufacturers.
We assume γd > σ/(σ − 1)and Bertrand competition within each market, which allows the new producer,
11
who produces a better product at the same cost as the current producer, to drive the current producer out of the
market.
We refer to this innovation process as creative destruction.

In addition, for products jwhere the incumbent producer is not eclipsed by cre-
ative destruction, there is each period an exogenous arrival rate λi ∈ [0, 1)of an
innovation that improves their quality by factor γi > 1. Hence, if jis a variety
where quality is improved upon by the incumbent producer, we have
qt( j) = γi qt−1
( j).
We call this incumbent own innovation. The producer of jchanges with creative
destruction, whereas it stays the same with incumbent own innovation. The arrival
rates and step sizes of creative destruction and incumbent own innovation are con-
stant over time and across varieties.
Finally, each period t, a flow of λn Nt−1new product varieties ι ∈
(Nt−1, Nt] are
created and available to final goods producers from tonward. Consequently, the law
of motion for the number of varieties is
Nt = (1 + λn) Nt−1.

We allow the (relative) quality of new product varieties to differ from the “average”
quality of preexisting varieties by a factor γn .12
To summarize, there are three sources of growth in this framework. First, the
quality of some products increases due to creative destruction. Second, for some
other products quality increases as a result of incumbent own innovation. Third, new
product varieties are invented which affects aggregate output, because the utility
function (1) features love-for-variety.
True, Measured, and Missing Growth.—There are no investment goods so that

aggregate nominal output equals aggregate nominal consumption expenditure.13
Nominal expenditure can be expressed as Pt Ct where
1/(1−σ)
Pt ≡ (∫
0   (pt( j) / qt( j))  dj) 
Nt

1−σ

is the quality-adjusted ideal price index. In this economy the real (gross) output
growth gt = Ct / Ct−1
is given by nominal output growth divided by the inflation
rate Pt / Pt−1
.
12
More formally, we assume that a firm that introduces in period ta new variety ι starts with a quality that
equals γ [0, Nt−1]in period t − 1, that is
n  >  0times the “average” quality of preexisting varieties j ∈
 _     

(
qt(ι) = γn  _ )
∫  t−1 qt−1 ( j) σ−1 dj   , ∀
N
1
N       ι ∈ (Nt−1, Nt].

1 σ−1
t−1 0
Average quality here is the geometric average, which depends on the elasticity of substitution. We do not put further
restrictions on the value of γn so that new products may enter the market with above-average (γn   >  1), average
(γn = 1), or below-average quality (γn < 1). As the relative quality of new varieties is a free parameter, express-
ing it this way is without loss of generality.
13
In online Appendix C, we allow for capital and show how this affects the interpretation of our missing growth
estimates relative to measured growth.
As we show in online Appendix B, the true growth rate in the economy is

_
  1    

(5) [ (
g = 1 + λd γ  σ−1
d    −  1) ( ) (
 +  1 − λd λi γ  σ−1
i    −  1) ]
 + λn γ  n    .
σ−1 σ−1
This equation shows how the arrival rates and step sizes affect the growth rate. The
n γ  σ−1
term λ n  captures the effect of variety expansion on growth, and the growth rate
is increasing in λnand γn . The term (1 − λd) λi ( γ  σ−1
i    −  1)summarizes the effect of
incumbent own innovation on growth. The term λ d ( γ  σ−1
d    −  1)captures the effect
from creative destruction on the growth rate.
Next we turn to measured growth. We posit that the statistical office resorts to
imputation in response to producer-product exit (we argue this is realistic in online
Appendix A). In so doing they presume the set of continuing producer-product pairs
is representative of the economy-wide inflation rate.14 For products that are not sub-
ject to creative destruction, we assume that statistical agency observes unit prices
correctly. Their quality adjustments allow them to retrieve the true frequency and
step size of quality improvements by incumbents on their own products. As we show
in online Appendix B, under these assumptions the measured real growth rate is
_
  1    

(6) gˆ t  = [1 + λi(γ  iσ−1
    −  1)]  .
σ−1
We then define the log difference between true growth and measured growth as
missing growth (MG). Combining (5) and (6) allows us to approximate missing
growth as15
λd( γ  σ−1
d    −  1) + λn γ  n   − λd λi(
σ−1
γ  σ−1
i    −  1)
(7) MGt ≈  _________________________
           .
σ − 1
The first two terms are growth from creative destruction and new varieties, respec-
tively. They are not missed entirely because growth is imputed based on incum-
bent own innovations in the event of creative destruction, which accounts for the
last term. Growth is missed when the rate of innovation from creative destruction
and new varieties exceeds that imputed from continuing producer-product pairs.
Equation (7) allows us to theoretically decompose missing growth into its sources:
creative destruction and variety expansion.
Under the appropriate definition of the subset of products I t, missing growth can
again be expressed in terms of formula (2) as
SI,t−1
σ − 1 ( SIt,t )
(8) ˆ t  − πt =  _
MGt = π  log _
1        t    
 ,
where measured inflation π 

ˆ tis as defined in equation (3). Importantly, as dis-
cussed, we define the set of overlapping products It in (3) as the set of continuing
14
BLS imputation is actually carried out within categories or category-regions. See the US Government
Accountability Office (1999).
15
The exact growth rate is
σ−1 σ−1 σ−1
(
log[1 + λd(γ d   − 1) + (1 − λd) λi(γ i   − 1) + λn γ n  ] − log 1 + λi(γ i   − 1)
σ−1
)
MGt = ___________________________________________________
              .
σ − 1
p roducer-product pairs. Growth is missed when the market share of these continu-
ing pairs shrinks over time.
In online Appendix D, we derive missing growth when the quality improvement
of incumbents is not perfectly measured. In particular, we show that missing growth
due to creative destruction would be larger if quality improvements by incumbents
are understated, precisely due to imputation.
II. Estimates of Missing Growth
Here we present estimates of missing growth using data on the market share of
entering establishments (plants), surviving plants, and exiting plants. This approach
does not allow us to differentiate between the different sources of missing growth
(creative destruction versus variety expansion), but it provides a simple and intuitive
quantification which avoids having to estimate the size and frequency of the various
types of innovations.
A. Measuring the Market Share of Continuers
Our goal is to quantify missing growth in the aggregate economy over a time
horizon of several decades. We therefore base our estimates of missing growth on
the Longitudinal Business Database (LBD), which covers all nonfarm business sec-
tor plants with at least one employee. We use the employment information in this
dataset to infer the market share of continuing plants, SIt,t.
Ideally we would have data at the product level for each firm. Unfortunately,
such data do not exist for the aggregate US economy outside of the Census of
Manufacturing, or consumer nondurables in the AC Nielsen scanner data. In lieu
of such ideal data, we suppose that firms must add plants in order to produce new
products. Such new plants could be at entering firms or at existing firms. And the
products produced by new plants could be brand new varieties or the result of cre-
ative destruction. Moreover, we assume that all incumbent own innovation occurs at
existing plants. Under these assumptions, we can use continuing plants as a proxy
for continuing incumbent products.
These assumptions are admittedly strong. They require that firms do not add
products through existing plants. Bernard, Redding, and Schott (2011) find that US
manufacturing plants do start up production in new industries. Our assumption may
be a better approximation outside manufacturing, such as in retail where location is
a key form of product differentiation. Related, if creative destruction occurs through
process innovations embodied in new establishments (e.g., new Walmart outlets),
then the market share of new plants should in principle capture them.
If existing plants do introduce new varieties or carry out creative destruction, then
our approach is likely to understate missing growth. As we explain below, our baseline
specification will assess market shares of new plants after a five-year lag. Hence, the
critical assumption is that plants do not add new products after the age of five years.
We can offer two facts that provide some reassurance here. First, employment
growth is much lower after age five than for younger plants (Haltiwanger, Jarmin,
and Miranda 2013). Second, plant exit rates do fall with age, but not very sharply
after age five (Garcia-Macia, Hsieh, and Klenow 2018). If plants add varieties after
age five, then one would expect exit rates to fall rapidly beyond age five.
Our baseline estimates use employment data to measure market shares, rather
than revenue or even payroll data. In our model these variables are all proportional
to each other across products. But in practice they differ. Annual revenue data are
only available at the firm level. Plant-level revenue data are available to us only
every five years in the manufacturing sector. As a robustness check we will report
results with revenue for manufacturing. And we will show robustness to using pay-
roll rather than employment for all sectors.
To be more exact, we calculate market shares using plant data as follows. Let B
denote the first year of operation and Ddenote the last year of operation of a plant.
Then the continuing plants Itare those plants who operated in both t − 1and t, that
is, all plants with B ≤ t − 1and D ≥ t. Define Etas the group of plants that first
operated in period t (B = t, D ≥ t) and X tas the group of plants that last operated
in period t (B ≤ t, D = t). Let L( t, )denote the total employment in period t of
plants belonging to group  . We then measure the ratio S It,t−1/SIt,ton the right-hand
side of (2) as
L(t − 1, It)
_______________
S It,t−1   
     
_ L(t − 1, It)  +  L(t − 1, Xt−1)
_______________
(9)        =          .
SIt,t _________ L(t, It)
       
L(t, It)  +  L(t, Et)
According to (9), missing growth is positive whenever the employment share of

continuing plants shrinks between t − 1and t.
We define a period tas a calendar year, and map Dto the last year a plant is in
LBD. We set Bequal to k ≥ 0years after the plant first appears in the LBD. If τ is
= τ + k. We use k = 5 in
the first year the plant appears in the database, we set B
our baseline specification.
We stress that, under our assumptions, equation (9) should capture missing
growth from creative destruction and brand new varieties. The missing growth from
creative destruction could be missed by the BLS at times of forced item substitu-
tions or regular item rotations.
B. Elasticities of Substitution
To quantify missing growth, we need the elasticity of substitution across plants

in all nonfarm business sectors. Hall (2018) estimates markups for 18 two-digit sec-
tors using KLEMS data for the period 1988–2015.16 He regresses industry output
growth on share-weighted input growth instrumented by military purchases and oil
prices, in the hopes that these are orthogonal to residual productivity growth.
Hall’s methodology is designed to estimate average markups within two-digit
industries, not elasticities of substitution across two-digit industries. Thus, there is
no discrepancy between the average product-level markups we seek and Hall’s level
16
Hall estimates that markups are modestly (but not precisely) trending up over time. His sample period is close
to our 1983–2013 LBD sample time frame.
Table 1—Average Missing Growth, 1983–2013 by Two-Digit Sectors
Contribution
Missing
NAICS Sector name Hall σ growth (ppt) (ppt) (%)
72 Hotels and restaurants 2.82 2.70 0.18 33.9
44–45 Retail trade 3.22 1.23 0.15 28.6
54 Professional services 4.23 1.07 0.05 8.6
52 Finance and insurance 3.17 0.60 0.03 5.4
81 Other services 4.03 0.54 0.02 4.4
48–49 Transportation and warehousing 4.23 0.58 0.02 3.7
71 Arts, entertainment 2.92 1.51 0.02 3.6
51 Information 3.56 0.88 0.02 3.4
42 Wholesale trade 4.70 0.29 0.01 2.5
31–33 Manufacturing 3.44 0.05 0.01 2.2
56 Administrative and support services 26.0 0.21 0.01 1.7
55 Management of companies 6.26 0.13 0.01 0.9
53 Real estate 12.1 0.24 0.00 0.6
22 Utilities 9.33 0.16 0.00 0.3
21 Mining 4.85 0.37 0.00 0.3
62 Health care ∞ 0 0 0
23 Construction ∞ 0 0 0
61 Education ∞ 0 0 0
Total 0.54
Notes: Entries are the elasticity of substitution and missing growth for different two-digit sectors. Sectors are
ordered by their contribution to missing growth over the period 1983–2013.
of aggregation. Missing growth does add a residual to Hall’s regression. Because

missing growth is not affected by Hall’s instruments in our model, however, his
identifying assumption remains valid in our setting.17
We convert Hall’s (2018) estimated markups to elasticities assuming the markups
equal σ/(σ − 1). We display the implied elasticities for each two-digit sector in the
first few columns of Table 1. In a few of the industries, Hall (2018) reports markups
that are less than 1. In these industries (health care, construction, and education), we
assume σ → ∞. As shown in the table, Hall’s (2018) implied σranges from below
3 to over 26 in the industries with markups in excess of 1.
C. Results
Our baseline results calculate missing growth from 1983–2013 using LBD data.
The LBD contains data on employment and payroll going back to 1976, but the
payroll data feature a number of implausible outliers before 1989. We therefore use
employment data for our baseline estimates. Using the employment data, we iden-
tify entrants beginning in 1977; 1983 is the earliest year we can calculate missing
growth (market share growth of survivors between 1982 and 1983) because we use
plants that have been in the data for at least five years (1977 to 1982 at the begin-
ning). We examine missing growth starting in 1983.18
17
If, contrary to our model, creative destruction results in different markups for the incoming item than the
markups that prevailed previously on outgoing items, then missing growth could create an omitted variable bias for
Hall’s regression.
18
More specifically, we calculate missing growth in each year tusing data from years t − 6, t − 5, t − 1,
and t. We use data in year t − 6to identify plants that have been in the data for at least five years in t − 1 and
In the fourth column of Table 1, we report our missing growth estimates for
each two-digit sector using Hall’s (2018) implied σ . We order the sectors by their
contribution to missing growth over the period 1983–2013, which totals 54 basis
points per year.19 The biggest contributors, by far, are NAICS 72 (hotels and
restaurants) and 44–45 (retail trade); they contribute 33 of the 54 basis points, or
over 62 percent of the total. Their contribution may reflect the geographic spread
of outlets by big national chains. No other sector contributes more than 5 basis
points.20
The dominant role played by hotels, restaurants, and retail trade hearkens back
to studies finding that large entering establishments carried productivity growth
in these sectors. See Foster, Haltiwanger, and Krizan (2006) and Jarmin, Klimek,
and Miranda (2009) for evidence on retail trade, and Hanner et al. (2011) for doc-
umentation of the big-box revolution among grocery stores, mass merchandisers,
and restaurants. We stress that the growth generated by this revolution may not have
been fully captured in official statistics.
Perhaps the most surprising result in Table 1 is the small contribution of manu-
facturing. Within this sector we estimate missing growth of about 5 basis points per
year, so that it contributes less than 1 basis point per year to overall missing growth.
By focusing on domestic production, we are overlooking the gains in import variety
emphasized in the trade literature, such as Feenstra (1994) and Broda and Weinstein
(2006).
How much bigger might missing growth be in manufacturing if we take into
account import variety? Subtracting exports and adding imports, missing growth
would be
_ domestic continuer domestic sales + import continuer sales

_________________________________________
  1     
Δlog               .
1 − σ domestic sales
Domestic sales refer to sales in the domestic market by domestic producers plus
imports. This expression assumes the BLS correctly measures quality growth for
continuing importers, just as for continuing domestic producers.
c alculate their total employment in year t − 1. Then we calculate the share of this total employment belonging to
plants that survive to year t. Similarly, we use year t − 5data to identify plants that are in the data for at least five
years in tand calculate the tperiod employment share of plants surviving from period t − 1.
19
We aggregate using Törnqvist averages of the current and previous year employment shares of the sector; this
is a second-order approximation to any smooth utility aggregator of two-digit sectors. We calculate the contribution
of a sector in a year by multiplying the missing growth in that sector with the Törnqvist employment share. Then we
take averages over 1983 to 2013 to arrive at the average contribution reported in the table.
20
The variation in missing growth across sectors in Table 1 suggests that official relative price trends across
sectors may be biased considerably. Taken at face value, our estimates alter the pattern of Baumol’s cost disease
across sectors.
Let continuer domestic sales denote the sum of domestic continuer domestic
sales and import continuer sales and let dom be a shorthand for domestic. We can
multiply and divide to decompose the key log first difference into three terms:
(10) Δlog  ______________
  
continuer
      
dom  sales = Δlog   __________________
   
dom
     
continuer 
dom sales 
dom sales 
dom producer dom sales
MG formula
dom producer dom sales

+  Δ
log  _________________
               
 
 dom sales
ACR correction
+ Δlog  
    __________________
   
   
     .
continuer dom sales
 dom continuer dom sales
Relative price correction
The first term on the right-hand side is our missing growth formula, only restricted
to the domestic sales of domestic producers. The second term is an ACR correction,
following Arkolakis, Costinot, and Rodríguez-Clare (2012), for welfare gains from
a rising import share. The third term should be decreasing in the relative (official)
price of imports; if positive (because of falling import prices) this will explain some
of the rising import share without resorting to missing growth.
To calculate the first term, we use the Census of Manufacturing from 1987 to
2012, extrapolating for 2013. 1987 is the first year when establishment-level export
data are available. We calculate domestic sales of a domestic establishment by sub-
tracting its exports from its sales. Then we use the methods described for Table 7 to
calculate the first term, replacing establishment level total sales with domestic sales.
To calculate the ACR term, we aggregate HS-level US import and export data from
Schott (2008) and merge with public tabulations of the value of shipments from the
Census of Manufacturing and the Annual Survey of Manufacturers.21 The trade data
begin in 1989, so the earliest year we can calculate the ACR term is 1990.
We do not have data to identify import continuers in the third term in equation
(10). Hence, we use an approximation and an assumption:
(     )
import continuer sales
Δlog  __________________
  continuer dom sales   ≡ Δlog 1 +  __________________
      
dom continuer dom sales dom continuer dom sales
import continuer sales

__________________
    
≈ Δ    
dom continuer dom sales
importer salesinitial
?
  (σ − 1)(ΔToT)  ×  ____________________
=     
     .
dom producer dom salesinitial
The terms of trade (ToT) appear in the last equality because under our CES frame-
work, σ
 − 1times the change in (official) terms of trade equals the growth in import
21
Import and export data are available in our dataset on the AER article page: https://doi.org/10.1257/
aer.20171745.
Table 2—Missing Growth with Trade in Manufacturing
1990–2013
Domestic missing growth 0.06
ACR adjustment 0.33
Relative price adjustment −0.17
Total in manufacturing 0.22
Contribution to aggregate missing growth 0.05
Notes: Entries are percentage points per year. Each entry is an average
over yearly values from 1990 to 2013. Rows 1 to 4 are calculated using
equation (10) with σ
= 3.44from Hall (2018). Row 5 multiplies missing
growth by manufacturing employment share.
continuer sales relative to domestic continuer sales. To convert this growth rate to
change in levels, we need to multiply by import continuer sales relative to domestic
continuer domestic sales in the initial period. We assume that the relative sales of
continuers is representative of all sales by domestic producers versus imports (hence
the question mark). We calculate the terms of trade as the export price index divided
by import price index. We use data from Atkeson and Burstein (2008) for 1990 to
2006 and data from the BLS for 2007 onward.22 We use data from Schott (2008) to
calculate imports relative to domestic producers’ domestic sales.
We present the results in Table 2. As mentioned above, the sample starts in 1990
because this is the earliest year for which we have data to calculate growth in imports
and domestic sales by domestic producers for manufactured goods.23 If we look only
at domestic sales of domestic producers, missing growth is 6 basis points per year
in manufacturing (row 1 in Table 2). But imports rose relative to the domestic sales
of domestic producers, opening the door to more missing growth through (unmea-
sured) rising import variety and import quality. This ACR adjustment is substantial
at 33 basis points per year (row 2). Roughly half of this, or 17 basis points per year,
is accounted for by the falling (official) relative price of imports, as shown in row 3.
This portion should be captured by official price indices. When we add up these
components, we arrive at 22 basis points of missing growth per year within manu-
facturing from 1990–2013. This compares to our estimate of only 5 basis points per
year when we ignored imports and exports altogether.
The final row in Table 2 shows that manufacturing contributes only 5 basis points
a year to aggregate missing growth. This is 4 basis points more than our baseline
estimate. Still, the bulk of our missing growth comes from outside manufacturing.
Perhaps manufacturing contributes less than expected in part because rising import
penetration has led to a loss of domestic variety, as emphasized for Canada in the
empirical study by Hsieh et al. (2016).
22
We resort to using two separate datasets because the BLS only provides export and import price indices for
the manufacturing sector from 2005.
23
This is the primary reason we do not insert this trade correction for manufacturing into our baseline estimates.
We also hesitate because we do not adjust for imports outside non-manufacturing, and because we need to assume
that continuers are representative of all sellers in terms of the ratio of imports to domestic sales. And, while adjust
for import variety is appropriate for estimating real domestic consumption growth, it may not be appropriate for
domestic GDP growth.
Table 3—Measured versus True Growth with Hall’s σ
Missing Measured True % of growth

growth growth growth missed
1983–2013 0.54 1.87 2.41 22.4

1983–1995 0.52 1.80 2.32 22.4
1996–2005 0.48 2.68 3.16 15.2
2006–2013 0.65 0.98 1.63 39.9
Notes: Entries are percentage points per year. Missing growth is calculated
using equation (2). The market share is measured as the employment share
of plants in the Census Longitudinal Business Database (LBD) as in (9).
These baseline results assume a lag k = 5and use the two-digit elasticities
of substitution in Hall (2018). Measured growth is calculated as the BLS
MFP series + R&D contribution expressed in labor-augmenting terms.
True growth is the sum of measured growth and missing growth.
Returning to our baseline estimates, Table 3 compares our missing growth to offi-
cial TFP growth. The entries are annual percentage points. As mentioned, we find
0.54 percentage points of missing growth per year from 1983–2013. BLS measured
TFP growth over the same interval was 1.87 percentage points per year.24 If we add
our missing growth to the BLS TFP series we arrive at “true” growth of 2.41 percent
per year. Thus, our baseline estimate is that over one-fifth of true growth is missed.
Table 3 also breaks the 30-year sample into three subperiods: 1983–1995 (an ini-
tial period of average official growth), 1996–2005 (a middle period of rapid official
growth), and 2006–2013 (a final period of low official TFP growth). Did missing
growth contribute to the speed-up or slow-down? Our estimates say yes, but mod-
estly at most. Missing growth slowed down by 4 basis points when official growth
accelerated by 88 basis points in the middle period. And missing growth sped up by
17 basis points when official growth dropped 170 basis points in the final period.
While missing growth did not accelerate significantly in the aggregate, it did so
in the Information sector. In this sector, missing growth rose from 0.22 percent per
year over 1983–2013 to 0.90 over 1996–2005 and to 1.90 over 2006–2013. Due to
its small employment share, however, the Information sector contributed a small
amount to overall missing growth: 0.00 over 1983–2013, 0.02 over 1996–2005, and
0.04 over 2006–2013.
D. Robustness and Discussion
In this section we discuss how our estimates of missing growth are affected by
the elasticity of substitution, the lag used to defined new plants, and the data used to
measure market shares (e.g., using payroll instead of employment data).
Elasticities of Substitution.—As a robustness check, we use the Visa data to esti-

mate σfor the two sectors responsible for 57 percent of the total missing growth in
24
We put BLS TFP growth in labor-augmenting form, and include the BLS estimates of the contribution of
R&D and intellectual property to TFP growth. The BLS multifactor productivity series uses real output growth from
the Bureau of Economic Analysis (BEA). The vast majority of the price indices that go into constructing BEA real
output growth come from the BLS (see US Bureau of Economic Analysis 2014), even if the BEA weights sectors
differently than in the aggregate CPI or PPI.
Table 1: retail trade and restaurants.25 Following the methodology of Dolfen et al.
(2018), we estimate σby using the location of Visa cardholders versus physical
stores and converting distance into effective price variation, based on the opportu-
nity cost of time and the direct costs of travel. We add the differential travel costs to
the average spending per visit at the store. Using all cards, we regress relative visits
on relative prices across stores to estimate the elasticity of substitution in a given
sector:
cij qj pjk
  + τij
( pjk  + τik )
(11)   c     ) = log(_
log(_   q     ) − (σ − 1)log _
     
.
ik k
Here idenotes a cardholder, j and k are competing merchants, crefers to the number
of visits, q denotes residual quality (assumed to be orthogonal to customer distance
to merchant jversus k ) , pjkis the average card spending per visit at merchants j and
k, and the τ s are estimated costs of travel.26
The key identifying assumption here is that people do not locate closer to com-
peting establishments for which they have an idiosyncratic preference. Notice this
would be hard to do across all NAICS at once. If they do, however, this would over-
state the elasticity of substitution and understate missing growth.
We only compare stores of competing chains within three-digit NAICS (e.g.,
general merchandisers Walmart versus Target, or grocery stores Trader Joe’s ver-
sus Whole Foods). Across restaurants (NAICS 722) we estimate an elasticity of
substitution of 2.92, not far from Hall’s (2018) estimate of 2.82. Across retail estab-
lishments (NAICS 44–45), we estimate an elasticity of 5.01, higher than Hall’s esti-
mate of 3.22.
In Table 4 we report missing growth for retail trade and restaurants combined,
where we aggregate these two industries using Törnqvist employment shares. Due to
the higher σ estimate for retail, the Visa data imply lower missing growth (1.26 per-
centage points per year) than when we use Hall (2018) σ values (1.68), and a corre-
spondingly smaller contribution to aggregate missing growth of 23 versus 30 basis
points per year.
These Visa-based estimates take advantage of differences in the time to shop at
one outlet versus another. We converted distance into effective price variation to
estimate the elasticity of demand across competing outlets. But we do not view the
market share gains of entering outlets as necessarily coming from locating closer
to customers or allowing them to shop and check out faster, time savings which are
usually outside the measurement of market GDP. Entering outlets, instead, can gain
market share because of a combination of selling higher quality products, selling a
wider variety of products, and providing higher quality service. When this happens,
we argue it results in missing growth in market GDP.
25
In Table 1, Retail Trade contributed 28.6 percent of the total missing growth whereas hotels and
Restaurants contributed 33.9 percent. We calculate the contribution of retail trade plus restaurants by
28.6% + 33.9% × 83.8% = 57.0%,where 83.8percent is the share of restaurant employment (NAICS 722) in
hotels and restaurants employment (NAICS 72) in the 2012 Census of Accommodation and Food Services.
26
Dolfen et al. (2018) estimate $ 0.79in direct costs (fuel, depreciation) and $ 0.80in indirect costs (opportunity
cost of time based on after-tax wages), for a total of $3.18 per round-trip mile. We follow them in using cardholder
and store locations to calculate driving distance to stores, and multiplying the distance by this cost per mile. In using
average spending per visit across the two merchants for pjk , we are assuming that the price and bundle of items
bought is the same at competing merchants, other than quality differences.
Table 4—Average Missing Growth for Retail Trade

and Restaurants
Using Using
Hall σs Visa σ
s
Missing growth in the sector 1.68 1.26
Contribution to overall missing growth 0.30 0.23
Notes: Hall σs are taken from Hall (2018) whereas Visa σ
s have been com-
puted from Visa data following Dolfen et al. (2018). Entries are in percent-
age point per year, on average between 1983 and 2013.
Table 5—Missing Growth with Different Elasticities σ
Missing growth
Higher elasticities Benchmark Lower elasticities
1983–2013 0.43 0.54 0.72
1983–1995 0.42 0.52 0.69
1996–2005 0.38 0.48 0.64
2006–2013 0.52 0.65 0.87
Notes: The Benchmark column uses σvalues from Hall (2018). The Higher elasticities are
25 percent higher than the σ − 1values estimated by Hall, and the Lower elasticities are
25 percent lower than Hall’s. Entries are in percentage point per year.
We next show the broader sensitivity of missing growth to the degree of substi-
tution across plants. As shown in (8), missing growth is proportional to 1 / (σ − 1).
Thus, missing growth declines monotonically as we raise σin Table 5. The higher is
σ, the smaller the quality and variety improvements by new plants to needed explain
the observed decline in the observed market share of continuing plants.
In our model, a given product’s share of revenue, the wage bill, employment, and
profits are all the same. This is due to the technology, preferences, and market struc-
ture we have assumed (physical output proportional to employment for all products,
a common wage for all workers in the economy, and a common markup across all
products). In the data, however, these series are not identical so it is useful to gauge
robustness to looking at the wage bill and revenue instead of employment.
Using Payroll to Measure Market Shares.—Table 6 compares missing growth

based on employment versus payroll. The payroll data allow us to do this only from
1989 onward. The estimates are quite similar overall, but exhibit a bigger dip and
bounce back across the three subperiods.
Using Revenue to Measure Market Shares.—For manufacturing only, we have

access to revenue data through the Census of Manufacturing (CMF). The CMF is
available every 5 years. To calculate missing growth with lag k = 5in a census
year t, we define incumbents as plants that are in the census in year tand t − 10
and calculate the market share growth of incumbents between year t − 5and t. We
convert this growth over 5 years to annual growth by taking logs and dividing by
5. We then use σ = 3.44for manufacturing from Hall (2018) to convert the result-
ing annualized growth into missing growth. Table 7 provides our missing growth
Table 6—Missing Growth with Employment versus Payroll

(PPT per Year)
Employment Payroll
1989–2013 0.59 0.56
1989–1995 0.66 0.72
1996–2005 0.48 0.31
2006–2013 0.65 0.73
Table 7—Missing Growth with Revenue versus Employment
Revenue Employment Employment

(CMF) (CMF) (LBD)
1982–2012 0.04 −0.05 −0.06
1982–1997 0.09 0.05 0.14
1997–2007 0.06 −0.25 −0.37
2007–2012 0.48 0.33 0.06
Notes: All entries are for manufacturing only and are given in percentage point per year. For
CMF, entries are averages over quinquennial data. For LBD, entries are averages of annual
data.
e stimates based on revenue versus employment for manufacturing. The time periods
are altered slightly to coincide with Census years.
The market share of continuing manufacturing plants shrinks more in terms
of revenue than employment, so that missing growth based on equation (8) is
higher when market shares are measured in terms of revenue. As shown, there is
little missing growth in manufacturing except for the end period, so on average
this increase is relative to a small contribution. The last column of the table dis-
plays missing growth from LBD manufacturing plants for the same periods. The
CMF and LBD yield similar missing growth for manufacturing, except for over
the period 2007–2012.
Different Lag k .—Our baseline results in Table 3 evaluate the market share of
entrants after a lag of k = 5years. Table 8 indicates how the results change if we
instead look at market shares with no lag after entry. With k = 0missing growth is
much smaller, averaging 23 basis points per year rather than 54 basis points. Though
not reported in Table 8, for k = 3we obtain estimates of missing growth closer to
k = 5. Increasing the lag beyond the baseline to k = 7years increases missing
growth only slightly compared to k = 5.
Missing Growth with More Sectors.—Our baseline estimate of missing growth

assumes a given elasticity of substitution across all plants within a two-digit sector.
We also calculated missing growth at a more detailed level and then aggregated the
missing growth rates using Törnqvist employment shares within each subsector as
weights (again, a second-order approximation to any smooth utility aggregator).
Within subsectors we set the elasticity of substitution equal to the Hall (2018) esti-
mate for the two-digit sector containing the subsector.
Table 8—Missing Growth with Different Lags k

(PPT per Year)
k = 5 k = 0
1983–2013 0.54 0.23
1983–1995 0.52 0.24
1996–2005 0.48 0.27
2006–2013 0.65 0.14
Table 9—Missing Growth with Disaggregated Sectors
2-digit 3-digit 4-digit 5-digit

1983–2013 0.54 0.56 0.57 0.65
1983–1995 0.52 0.53 0.46 0.58
1996–2005 0.48 0.54 0.63 0.68
2006–2013 0.65 0.64 0.70 0.72
Notes: Entries are percentage points per year. The first column repeats our baseline results.
Other columns are Tornqvist employment-weighted averages of missing growth within differ-
ent NAICS (2002) n-digit levels.
Table 9 compares our benchmark missing growth estimates to those obtained by

aggregating up missing growth in this way up from the three-, four-, or five-digit
sectoral level. We present our baseline two-digit estimates as well, for comparison.
The estimates are gently increasing in the level of disaggregation (from 54 basis
points to 65), and are broadly similar across subperiods.27
Declining Dynamism and Missing Growth.—One may wonder why our miss-
ing growth estimates do not trend downward along with rates of entry, exit, and
job reallocation across firms, the “declining dynamism” documented by Decker
et al. (2014). The answer is three-fold. First, we look at plants (establishments), not
firms. Second, our market share equation for missing growth is tied to the net entry
rate (weighted by employment), not the gross job creation rate due to entrants. Put
differently, the growth of survivors’ market share is influenced by the difference
between the job creation rate due to new plants and the job destruction rate due to
exiting plants. Unlike gross flows at the firm level, we see no trend in the net job
creation rate of plants over 1983–2013 in the LBD. Finally, we look at market shares
five years after the plant appears in the LBD, rather than immediately upon entry.
Table 10 illustrates these points of distinction between our market share approach
and declining dynamism. Across the first two columns, missing growth drops dra-
matically when moving from plants to firms. Many new plants are at existing firms,
and they are bigger on average than new plants in new firms.28
27
We also aggregated sectoral missing growth rates using average employment shares over the entire
1983–2013 period to fully eliminate any trends due to changes in sectoral composition. The results were very
similar to those in Table 9.
28
We identify firms using the LBD firm ID, which can change with mergers and acquisition events. We also
calculated firm-level missing growth using an alternative firm ID for each plant, where we set the firm ID of a plant
to its initial firm ID throughout the lifetime of the plant. We find higher firm-level missing growth (0.31 percent per
year versus 0.15 over 1983–2013) after netting out M&A activity in this way.
Table 10—Missing Growth versus Declining Dynamism (PPT per Year)
Plant level Firm level Net entry Gross entry

1983–2013 0.54 0.15 0.29 0.29
1983–1995 0.52 0.17 0.27 0.41
1996–2005 0.48 0.16 0.32 0.29
2006–2013 0.65 0.10 0.27 0.09
The third column of Table 10, labeled Net entry, shows how big missing growth
would be using firm-level data and assuming all firms were of the same size (i.e.,
had the same level of employment). In this case, missing growth is larger than the
firm-level estimate because entering firms tend to be smaller than the average firm.
Finally, the gross entry rate has declined more than the net entry rate, and our
missing growth estimates focus on the net entry rate. To illustrate this point, the
last column of Table 10 shows missing growth if all firms were equal-sized and the
exit rate was fixed to the 1983–2013 average. Missing growth declines even more
precipitously when measured in this counterfactual way. In the data, the exit rate fell
along with the entry rate, dampening the decline in missing growth.
To recap, declining dynamism is seen most strikingly in the gross entry rate,
which is several steps removed from our missing growth calculation. We base our
missing growth estimates on plant dynamics, rather than firm dynamics, because we
think it is much defensible to assume plants do not add new products than to assume
that firms do not do so.
III. External Validation
Here we compare our sectoral missing growth estimates with those in two prom-
inent papers that used more detailed data on prices and quantities, and to those
obtained using a separate, indirect inference approach.
A. Bils (2009) and Broda and Weinstein (2010)
Bils (2009) uses scanner data to estimate quality bias in the CPI for consumer
durables. He estimates a bias of 1.8 percent per year from 1988 to 2006. To obtain
a comparable estimate, we first restrict our attention to Census retail NAICS for
durable consumer goods, which include 441 (motor vehicle and parts dealers),
442 (furniture and home furnishings stores), and 443 (electronics and appliance
stores). Comparing stores of competing chains in these categories, we estimate
σ = 7 .9using Visa data. With this Visa-based σ
and the market share of continuers
in the LBD in these industries, we arrive at missing growth of 0.36 percentage points
per year from 1988 to 2006. This compares to 1.8 percentage points per year in Bils
(2009). Bils’s number is understandably higher because he includes understated
improvements of incumbent products, whereas our focus is solely on new outlets.
Similarly, we can restrict attention to grocery and drug store retailers to facil-
itate comparison to Broda and Weinstein (2010). We map this to retailers selling
nondurables other than gasoline: categories 445 (food and beverage stores) and
446 (health and personal care stores). Comparing Visa spending at stores within
these industries, we estimate σ
= 6.0. We then combine this σ
with data on con-
tinuer shares in the LBD for these industries over the period 1994–2003. We find
annual missing growth of 0.43 percentage points per year, compared to the Broda
and Weinstein (2010) estimate of 0.8 percent per year. Again, our estimate should
be smaller in that we focus on new outlets, whereas their estimate would capture
improvements in products at continuing retailers.
B. Indirect Inference Approach
Our market share approach is simple and intuitive. It does, however, require
that a plant not add new product lines; all creative destruction must occur through
new plants. Furthermore, it cannot separate out expanding variety from creative
destruction. We therefore entertained an alternative “indirect inference” method-
ology which does not rely on such assumptions. This methodology starts from the
decomposition of missing growth into its creative destruction and variety expansion
components. We use the exact growth rate, but for intuition the approximate growth
decomposition is useful:
λd( γ  σ−1

d    −  1) + λn γ  n   − λd λi(
σ−1
γ  σ−1
i    −  1)
MGt ≈  _________________________
           ,
σ − 1
where λd (λn) is the sum of the arrival rate of creative destruction (new variety)
innovation by incumbents and entrants.
We estimate missing growth and its decomposition by first estimating the fre-
quency and size of the various types of innovation, i.e., by estimating parameters
(λi, γi, λd, γd, λn, γn) . This in turn requires more data moments than when using
the market share approach. This method is built on Garcia-Macia, Hsieh,
and Klenow (2018)—henceforth, GHK—who back out arrival rates and step sizes
(λi, γi, λd, γd, λn, γn) by inferring parameter values to mimic moments on firm dynam-
ics in the LBD. In the Appendix we describe the GHK algorithm in some detail. We
modify it to incorporate how measured growth can differ from true growth; GHK
assumed that growth was measured perfectly.
Table 11 reports our indirect inference estimates, both parameter values and the
missing growth they imply. We find more missing growth under this indirect infer-
ence approach (99 basis points per year on average) than under the market share
approach (54 basis points per year). But, like the market share approach, the indirect
inference approach yields no big acceleration in missing growth to account for the
sharp slowdown in measured growth over the period 2003–2013. Finally, the indi-
rect inference approach implies that the vast majority (over 80 percent) of missing
growth is due to creative destruction. Expanding varieties play a limited role.
IV. Conclusion
In this paper we lay out a model with incumbent and entrant innovation to assess
the unmeasured TFP growth resulting from creative destruction. Crucial to this
missing growth is the use of imputation by statistical agencies when producers no
longer sell a product line. Our model generates an explicit expression for missing
TFP growth as a function of the frequency and size of creative destruction ver-
sus other types of innovation.
Based on the model and US Census data for all nonfarm businesses, we esti-
mated the magnitude of missing growth from creative destruction over the period
1983–2013. Our approach uses the market share of surviving, entering, and exit-
ing plants. We found (i) missing growth from imputation was substantial at around
one-half of a percentage point per year, or over one-fifth of measured productivity
growth; (ii) missing growth was concentrated in hotels, restaurants, and retail trade
rather than manufacturing; and (iii) it has accelerated modestly since 2005, so that it
accounts for only about one-tenth of the sharp growth slow-down since then.
We may be understating missing growth because we assumed there were no
errors in measuring quality improvements by incumbents on their own products.
We think missing growth from imputation is over and above (and amplified by) the
quality bias emphasized by the Boskin Commission.29
Our analysis could be extended in several interesting directions. One would be to
look at missing growth in countries other than the United States. A second extension
would be to revisit optimal innovation policy. Based on Atkeson and Burstein (forth-
coming), the optimal subsidy to R&D may be bigger if true growth is higher than
measured growth. Conversely, our estimates give a more prominent role to creative
destruction with its attendant business stealing.
A natural question is how statistical offices should alter their methodology in
light of our results, presuming our estimates are sound. The market share approach
would be hard to implement without a major expansion of BLS data collection to
include market shares for entering, surviving, and exiting products in all sectors. The
indirect inference approach is even less conducive to high frequency analysis. A fea-
sible compromise might be for the BLS to impute quality growth for disappearing
products based on its direct quality adjustments for those surviving products that
have been innovated upon.30
Our missing growth estimates have other implications which deserve to be
explored further. First, ideas may be getting harder to find, but not as quickly as offi-
cial statistics suggest if missing growth is sizable and relatively stable. This would
have ramifications for the production of ideas and future growth (Gordon 2012,
Bloom et al. 2018). Second, the US Federal Reserve might wish to raise its inflation
target to come closer to achieving quality-adjusted price stability. Third, a higher
fraction of children may enjoy a better quality of life than their parents (Chetty et al.
2017). Fourth, as stressed by the Boskin Commission, US tax brackets and Social
Security benefits may rise too steeply since they are indexed to measured inflation,
the inverse of measured growth.
29
Economists at the BLS and BEA recently estimated that quality bias from health and ICT alone was about
40 basis points per year from 2000 to 2015 (Groshen et al. 2017).
30
Erickson and Pakes (2011) suggest that, for those categories in which data are available to do hedonics, the
BLS could improve upon the imputation method by using hedonic estimation that corrects for both the selection
bias associated with exit and time-varying unmeasured characteristics.
Table 11—Estimated Parameters and Results with Indirect Inference
Parameter Definition 1983–1993 1993–2003 2003–2013

d
λ CD arrival rate 0.014 0.011 0.010
λi OI arrival rate (if survive) 0.024 0.027 0.027
λn NV arrival rate 0.004 0.002 0.002
γd, γi Step size of CD, OI 1.106 1.125 1.074
γn Step size of NV 0.328 0.482 0.366
Measured growth per year (ppt) 1.66 2.29 1.32
Missing growth (ppt) 1.25 1.13 0.60
True growth per year (ppt) 2.91 3.42 1.92
Percent of missing growth from CD 79.4 79.7 80.8
Percent of growth missed 43.0 32.9 31.2
Notes: Results from running the algorithm from Garcia-Macia, Hsieh, and Klenow (2018) on three samples: 1983–
1993, 1993–2003, and 2003–2013. We adapt their algorithm to incorporate missing growth as explained in the text.
Appendix A. The Indirect Inference Method
In this section we provide an alternative, indirect inference approach to quantify

missing growth. It allows us to, in principle, separate the two sources of missing
growth: creative destruction and variety expansion. The method relies on first esti-
mating the arrival rates and step sizes in the model of Section I and then calculating
missing growth based on formulas derived from the model. To estimate the arrival
rates and step sizes, we adapt the algorithm developed in Garcia-Macia, Hsieh,
and Klenow (2018)—henceforth, GHK.
A. The Original GHK Algorithm
GHK’s algorithm uses indirect inference to estimate the step size and arrival rate
of three types of innovation: Own innovation (OI), Creative Destruction (CD), and
New Varieties (NV). GHK estimate these parameters to fit aggregate TFP growth;
the mean, minimum (one worker), and standard deviation of employment across
firms; the share of employment in young firms (firms less than five years old); the
overall job creation and destruction rates; the share of job creation from firms that
grew by less 1 log point (three-fold) over a five-year period; employment share by
age; exit rate by size; and the growth rate in the number of firms (which is equal to
the growth rate of employment in the model).31 They calculate these moments in the
LBD for 1983–1993, 1993–2003, and 2003–2013, respectively.32
With their parameter estimates in hand, GHK decompose growth into contri-
butions from new varieties, incumbent innovation on their own products, creative
destruction by incumbents, and creative destruction by entering firms.
31
GHK assume each variety carries an overhead cost. Firms choose to retire a variety if the expected profits
from that variety do not cover the overhead cost. GHK calibrate the overhead cost to match minimum employment
in the data.
32
Their algorithm matches model steady-state moments to data moments, and therefore does not produce
annual estimates.
B. How We Differ from GHK
Our model in Section I differs from the GHK model in several respects. GHK
keep track of firms with multiple products, and estimate rates of creative destruction
and new variety creation separately for entrants and incumbents. GHK also endog-
enize product exit: firms drop products whose quality relative to the average quality
is below a certain cutoff. And rather than assuming a fixed step size for innovations,
in GHK quality innovations are drawn from a Pareto distribution, where the same
Pareto shape parameter (and hence the same average step size) is assumed for qual-
ity innovations from incumbents’ own innovation and from innovations involving
creative destruction. Finally, in GHK new varieties are drawn from a scaled version
of the existing quality distribution rather than massing at a single point relative to
the existing distribution.
As a result of these differences, GHK obtain an expression for true productivity
growth which is somewhat different from that in our Section I. Using our notation,
true productivity growth g ≡ Ct+1 / Ct  −  1in GHK is
(A1) 1 + g = [(1 − δoψ){[λi(1 − λe,d  − λi,d) + (λe,d  + λi,d)](γ  σ−1

i  − 1) + 1} 
_
  1    

( )
+  λi,n  + λe,n γ  n  ]  ,
σ−1 σ−1
where δ odenotes the share of products in the previous period whose quality falls
below the obsolescence cutoff; ψis the average quality of those below-cutoff prod-
ucts relative to the average quality;33 λiis the share of products that are not obsolete,
did not experience creative destruction, and did experience an innovation by the
incumbent producer; λe ,dis the share of non-obsolete products with entrant cre-
ative destruction; λi ,dis the share of non-obsolete products with incumbent creative
destruction; λ i,n  + λe,nis the mass of new varieties from incumbents and entrants
relative to the mass of products in the previous period; and γ iand γ
dare the average
step sizes of own innovation and creative destruction, respectively. As in GHK, we
assume that the two step sizes are the same, which is why only γ iappears in equa-
tion (A1). Finally, γnis the average quality of a new variety relative to the average
quality of varieties produced in the previous period. The term 1 − δ0 ψadjusts for
the endogenous loss of varieties due to obsolescence.
The equation for measured growth g  ˆ in our modified GHK model is the same as
in our Section I:
 _     

(1 + λi(γ  i    −  1)) 
1
(A2) 1 + gˆ   = σ−1

.
σ−1
Recall that we assume the BLS accurately measures the arrival rate and the average
step size of incumbents’ own innovations. To adapt the GHK methodology to our

∫q<
   q– t, q∈Ωt  t  ( j) dj
q σ−1
__________
δo = ∫ q  
33
( j)<q– t, q( j)∈Ωt 
1 djand ψδo =   
    
. Ωtdenotes the set of products in t.
 
∫q∈
   q t  ( j) dj
Ωt 
σ−1
model with missing growth, we make the following changes to the original GHK
algorithm:34
(i) We choose parameters so that (A2) matches the observed growth rates:
1.66 percent for 1983–1993, 2.29 percent for 1993–2003, and 1.32 percent
for 2003–2013, according to the BLS.
(ii) We set the combined unconditional arrival rates of OI and CD to the cumula-
tive rate of CPI non-comparable substitutions over five years.
Key advantages of this indirect inference method include the following: (i) we
need not assume that creative destruction and new product varieties only come from
new plants (the inference is on firm-level data and allows for multi-product firms);
incumbent plants may also produce CD or NV innovations; (ii) we can decompose
missing growth into its CD and NV components using the arrival rates and step sizes
of the various kinds of innovations; and (iii) we allow for the possibility of products
disappearing because of obsolescence.
C. Results from Indirect Inference
Table 11 defines the parameters and displays their estimated values for each of
the three samples: 1983–1993, 1993–2003, and 2003–2013. The bottom panel of
Table 11 reports the resulting estimates of measured, true, and missing growth.
Missing growth is larger under this alternative approach than under the market share
approach for the first two sample periods: 1.25 percentage points per year (ver-
sus 0.52 when using the market share approach) for the 1983–1993 period, and
1.13 percentage points per year (versus 0.48) for the 1993–2003 period. For the
last sample period, 2003–2013, the missing growth estimates from the indirect
inference method are closer to those from the market share approach (0.60 percent
versus 0.65). Under the indirect inference approach, the fraction of total produc-
tivity growth that is missed is comparable across the three periods at around one-
third of true growth. Just like in the market share approach, in the indirect inference
approach we do not find that missing growth accelerated when measured growth fell
sharply in the last interval.
As mentioned, an advantage of the indirect inference approach over the market
share approach is that here we can decompose missing growth into its new varieties
(NV) and creative destruction (CD) components. We calculate missing growth from
creative destruction by taking the difference between measured productivity growth
and the productivity growth that results when we set the total arrival rate for new
varieties (λi,n  + λe,n) equal to zero. We find that vast majority of the missing growth
is due to creative destruction: around 80 percent in all three periods.
34
See our online Appendix E for a more details description of the changes we made.
D. Comparison to the Market Share Approach
As stressed, the market share approach uses plant-level data (assuming no added
products per plant, and focusing attention on plants that are at least five years old),
whereas the indirect inference approach uses firm-level data. The market share
approach assumes that creative destruction only occurs through new plants. The
indirect inference method allows for creative destruction by existing plants as well.
This may be why we found larger average missing growth in this second quantifica-
tion than in the market share approach.
The falling entry rate of new firms over the past three decades (“declining dyna-
mism”) may explain why missing growth declines across the three periods in the
indirect inference approach, which again uses firm-level data. The market share
approach with plant-level exhibited no such decline. But when, as a robustness
check, we applied the market share approach to gross entry of firms, we did obtain
a sharp decline in missing growth across periods (Table 10).
As already mentioned above, indirect inference did not require that only entrant
plants create new varieties or generate creative destruction; and this method allowed
us to split overall missing growth into its NV and CD components. On the other
hand, the market share approach is simple, requires fewer model assumptions, and
is less data demanding.
REFERENCES
Aghion, Philippe, Ufuk Akcigit, and Peter Howitt. 2014. “What Do We Learn from Schumpeterian
Growth Theory?” In Handbook of Economic Growth, Vol. 2, edited by Philippe Aghion and Steven
Durlauf, 515–63. Amsterdam: Elsevier.
Aghion, Philippe, Antonin Bergeaud, Timo Boppart, Peter J. Klenow, and Huiyu Li. 2019. “Missing
Growth from Creative Destruction: Dataset.” American Economic Review. https://doi.org/10.1257/
aer.20171745.
Aghion, Philippe, and Peter Howitt. 1992. “A Model of Growth through Creative Destruction.” Econo-
metrica 60 (2): 323–51.
Akcigit, Ufuk, and William R. Kerr. 2018. “Growth through Heterogeneous Innovations.” Journal of
Political Economy 126 (4): 1374–443.
Arkolakis, Costas, Arnaud Costinot, and Andrés Rodríguez-Clare. 2012. “New Trade Models, Same
Old Gains?” American Economic Review 102 (1): 94–130.
Atkeson, Andrew, and Ariel Burstein. 2008. “Pricing-to-Market, Trade Costs, and International Rela-
tive Prices.” American Economic Review 98 (5): 1998–2031.
Atkeson, Andrew, and Ariel Burstein. Forthcoming. “Aggregate Implications of Innovation Policy.” Jour-
nal of Political Economy.
Bernard, Andrew B., Stephen J. Redding, and Peter K. Schott. 2011. “Multiproduct Firms and Trade
Liberalization.” Quarterly Journal of Economics 126 (3): 1271–318.
Bils, Mark. 2009. “Do Higher Prices for New Goods Reflect Quality Growth or Inflation?” Quarterly
Journal of Economics 124 (2): 637–75.
Bils, Mark, and Peter J. Klenow. 2001. “Quantifying Quality Growth.” American Economic Review
91 (4): 1006–30.
Bils, Mark, and Peter J. Klenow. 2004. “Some Evidence on the Importance of Sticky Prices.” Journal
of Political Economy 112 (5): 947–85.
Bloom, Nicholas, Charles I. Jones, John Van Reenen, and Michael Webb. 2018. “Are Ideas Getting
Harder to Find?” Unpublished.
Boskin, Michael, Ellen R. Dullberger, Robert J. Gordon, Zvi Griliches, and Dale Jorgenson. 1996.
Final Report of the Advisory Commission to Study the Consumer Price Index. Washington, DC: US
Senate Committee on Finance.
Broda, Christian, and David Weinstein. 2006. “Globalization and the Gains from Variety.” Quarterly
Journal of Economics 121 (2): 541–85.
Broda, Christian, and David Weinstein. 2010. “Product Creation and Destruction: Evidence and Price
Implications.” American Economic Review 100 (3): 691–723.
Byrne, David M., John G. Fernald, and Marshall B. Reinsdorf. 2016. “Does the United States Have
a Productivity Slowdown or a Measurement Problem?” Brookings Papers on Economic Activity
(Spring): 109–82.
Byrne, David M., Stephen D. Oliner, and Daniel E. Sichel. 2015. “How Fast Are Semiconductor Prices
Falling?” NBER Working Paper 21074.
Chetty, Raj, David Grusky, Maximilian Hell, Nathaniel Hendren, Robert Manduca, and Jimmy
Narang. 2017. “The Fading American Dream: Trends in Absolute Income Mobility since 1940.”
Science 356 (6336): 398–406.
Decker, Ryan, John Haltiwanger, Ron Jarmin, and Javier Miranda. 2014. “The Role of Entrepre-
neurship in US Job Creation and Economic Dynamism.” Journal of Economic Perspectives 28
(3): 3–24.
Dolfen, Paul, Liran Einav, Peter J. Klenow, Benjamin Klopack, Jonathan Levin, Larry Levin, and
Wayne Best. 2018. “Assessing the Gains from E-Commerce.” Unpublished.
Erickson, Tim, and Ariel Pakes. 2011. “An Experimental Component Index for the CPI: From Annual
Computer Data to Monthly Data on Other Goods.” American Economic Review 101 (5): 1707–38.
Feenstra, Robert C. 1994. “New Product Varieties and the Measurement of International Prices.”
American Economic Review 84 (1): 157–77.
Foster, Lucia, John Haltiwanger, and Cornell J. Krizan. 2006. “Market Selection, Reallocation, and
Restructuring in the U.S. Retail Trade Sector in the 1990s.” Review of Economics and Statistics 88
(4): 748–58.
Garcia-Macia, Daniel, Chang-Tai Hsieh, and Peter J. Klenow. 2018. “How Destructive Is Innovation?”
Unpublished.
Gordon, Robert J. 2012. “Is U.S. Economic Growth Over? Faltering Innovation Confronts the Six
Headwinds.” NBER Working Paper 18315.
Groshen, Erica L., Brian C. Moyer, Ana M. Aizcorbe, Ralph Bradley, and David M. Friedman. 2017.
“How Government Statistics Adjust for Potential Biases from Quality Change and New Goods in
an Age of Digital Technologies: A View from the Trenches.” Journal of Economic Perspectives 31
(2): 187–210.
Hall, Robert E. 2018. “New Evidence on the Markup of Prices over Marginal Costs and the Role of
Mega-Firms in the US Economy.” NBER Working Paper 24574.
Haltiwanger, John, Ron S. Jarmin, and Javier Miranda. 2013. “Who Creates Jobs? Small versus Large
versus Young.” Review of Economics and Statistics 95 (2): 347–61.
Hanner, Daniel, Daniel Hosken, Luke M. Olson, and Loren K. Smith. 2011. Dynamics in a Mature
Industry: Entry, Exit, and Growth of Big-Box Retailers. Washington, DC: Bureau of Economics,
US Federal Trade Commission.
Hausman, Jerry, and Ephraim Leibtag. 2009. “CPI Bias from Supercenters: Does the BLS Know that
Wal-Mart Exists?” In Price Index Concepts and Measurement, edited by W. Erwin Diewert, John S.
Greenlees, and Charles R. Hulten, 203–31. Chicago: University of Chicago Press.
Hottman, Colin J., Stephen J. Redding, and David E. Weinstein. 2016. “Quantifying the Sources of
Firm Heterogeneity.” Quarterly Journal of Economics 131 (3): 1291–364.
Hsieh, Chang-Tai, Nicholas Li, Ralph Ossa, and Mu-Jeung Yang. 2016. “Accounting for the New
Gains from Trade Liberalization.” NBER Working Paper 22069.
Jarmin, Ronald S., Shawn D. Klimek, and Javier Miranda. 2009. “The Role of Retail Chains: National,
Regional and Industry Results.” In Producer Dynamics: New Evidence from Micro Data, edited by
Timothy Dunne, J. Bradford Jensen, and Mark J. Roberts, 237–62. Chicago: University of Chicago
Press.
Jones, Charles I. 1995. “Time Series Tests of Endogenous Growth Models.” Quarterly Journal of Eco-
nomics 110 (2): 495–525.
Kortum, Samuel S. 1997. “Research, Patenting, and Technological Change.” Econometrica 65
(6): 1389–419.
Moulton, Brent R. 2018. The Measurement of Output, Prices, and Productivity: What’s Changed since
the Boskin Commission? Washington, DC: Hutchins Center on Fiscal and Monetary Policy at the
Brookings Institution.
Moulton, Brent R., and Karin E. Moses. 1997. “Addressing the Quality Change Issue in the Consumer
Price Index.” Brookings Papers on Economic Activity (1): 305–66.
Nakamura, Emi, and Jón Steinsson. 2008. “Five Facts about Prices: A Reevaluation of Menu Cost
Models.” Quarterly Journal of Economics 123 (4): 1415–64.
Redding, Stephen J., and David E. Weinstein. 2018. “Measuring Aggregate Price Indexes with Demand
Shocks: Theory and Evidence for CES Preferences.” NBER Working Paper 22479.
Schott, Peter K. 2008. “The Relative Sophistication of Chinese Exports.” Economic Policy 23
(53): 6–49.
Syverson, Chad. 2017. “Challenges to Mismeasurement Explanations for the US Productivity Slow-
down.” Journal of Economic Perspectives 31 (2): 165–86.
US Bureau of Economic Analysis. 2014. “Principal Source Data.” In Concepts and Methods of the U.S.
National Income and Product Accounts. Washington, DC: US Department of Commerce.
US Bureau of Labor Statistics. 2015. “Producer Prices.” In Handbook of Methods, Chapter 14. Wash-
ington, DC: US Department of Labor.
US Government Accountability Office. 1999. Consumer Price Index: Impact of Commodity Analysts’
Decisionmaking Needs to Be Assessed. Washington, DC: US Government Accountability Office.

Missing Growth From Creative Destruction: American Economic Review 2019, 109 (8) : 2795-2822

Uploaded by

Copyright:

Available Formats

Missing Growth From Creative Destruction: American Economic Review 2019, 109 (8) : 2795-2822

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Missing Growth From Creative Destruction: American Economic Review 2019, 109 (8) : 2795-2822

Uploaded by

Copyright:

Available Formats

American Economic Review 2019, 109(8): 2795–2822

Missing Growth from Creative Destruction†

By Philippe Aghion, Antonin Bergeaud, Timo Boppart,

For exiting products, statistical agencies often impute inflation from

Whereas it is straightforward to compute inflation for an unchanging set of goods

Presuming it correctly evaluates the growth in nominal GDP to be 4 percent, the

This in turn implies “missing growth” in productivity amounting to

Byrne, Fernald, and Reinsdorf (2016) conclude that understatement of growth in

I. A Model of Missing Growth

In this section we relate our measure of missing growth to Feenstra (1994). We

A. Relating Missing Growth to Market Share Dynamics

where f​ ​( j, ​​{​s​I​​t​​,t−1​​​(i)​, ​s​I​​t​​,t​​​(i)​}​​i∈​It​​​​​)​​is the Sato-Vartia weight.8

See Table 1 in Groshen et al. (2017) and Table B1 in Moulton (2018).

B. Framework of Creative Destruction and Missing Growth

A representative household inelastically supplies L ​ ​units of labor each period and

​ct​(​​​ j)​ = ​l​t(​​​ j)​,​

by a monopolistically competitive producer, where ​l​​t​​​( j)​​is the amount of labor used

(4) ​​pt​​​​( j)​ = ​ _ σ ​ ​

We refer to this innovation process as creative destruction.

​N​t​​ = ​(1 + ​λ​n)​​ ​ ​N​t−1​​.​

True, Measured, and Missing Growth.—There are no investment goods so that

​N​ ​​​ ι ∈ ​(​N​t−1​​, ​Nt​​​]​.​

As we show in online Appendix B, the true growth rate in the economy is

where measured inflation ​​​π

II. Estimates of Missing Growth

A. Measuring the Market Share of Continuers

According to (9), missing growth is positive whenever the employment share of

To quantify missing growth, we need the elasticity of substitution across plants

Table 1—Average Missing Growth, 1983–2013 by Two-Digit Sectors

of aggregation. Missing growth does add a residual to Hall’s regression. Because

_ domestic continuer domestic sales + import continuer sales

dom producer dom sales

import continuer sales

Table 2—Missing Growth with Trade in Manufacturing

Table 3—Measured versus True Growth with Hall’s ​σ​

Missing Measured True % of growth

1983–2013 0.54 1.87 2.41 22.4

D. Robustness and Discussion

Elasticities of Substitution.—As a robustness check, we use the Visa data to esti-

Table 4—Average Missing Growth for Retail Trade

Table 5—Missing Growth with Different Elasticities ​σ​

Using Payroll to Measure Market Shares.—Table 6 compares missing growth

Using Revenue to Measure Market Shares.—For manufacturing only, we have

Table 6—Missing Growth with Employment versus Payroll

Table 7—Missing Growth with Revenue versus Employment

Revenue Employment Employment

Missing Growth with More Sectors.—Our baseline estimate of missing growth

Table 8—Missing Growth with Different Lags ​k​

Table 9—Missing Growth with Disaggregated Sectors

2-digit 3-digit 4-digit 5-digit

Table 9 compares our benchmark missing growth estimates to those obtained by

Table 10—Missing Growth versus Declining Dynamism (PPT per Year)

Plant level Firm level Net entry Gross entry

III. External Validation

A. Bils (2009) and Broda and Weinstein (2010)

B. Indirect Inference Approach

​λd​(​​​ ​γ​ σ−1

Table 11—Estimated Parameters and Results with Indirect Inference

Parameter Definition 1983–1993 1993–2003 2003–2013

Presuming it correctly evaluates the growth in nominal GDP to be 4 percent, the

where f ( j, {sIt,t−1(i), sIt,t(i)}i∈It)is the Sato-Vartia weight.8

A representative household inelastically supplies L units of labor each period and

ct( j) = lt( j),

by a monopolistically competitive producer, where lt( j)is the amount of labor used

(4) pt( j) =  _ σ     

Nt = (1 + λn) Nt−1.

N       ι ∈ (Nt−1, Nt].

where measured inflation π 

_ domestic continuer domestic sales + import continuer sales

Table 3—Measured versus True Growth with Hall’s σ

Table 5—Missing Growth with Different Elasticities σ

Table 8—Missing Growth with Different Lags k

λd( γ  σ−1

(A1) 1 + g = [(1 − δoψ){[λi(1 − λe,d  − λi,d) + (λe,d  + λi,d)](γ  σ−1

(A2) 1 + gˆ   = σ−1