from: ron katznelson sent: friday, june 05, 2009 4:39 pm ... · from: ron katznelson sent: friday,...
TRANSCRIPT
-------------------------------------------------
From: Ron Katznelson Sent: Friday, June 05, 2009 4:39 PM To: AC6/Comments Cc: Bahr, Robert Subject: Deferred Examination Comments
Attention: Robert W. Bahr
Please accept my late-filed comments directed to the Office's inquiry on Deferred Examination. The extra time was required to finalize a study attached to my comments. I believe that my comments and study will provide important information on this subject.
Best regards,
Ron
Ron D. Katznelson, Ph.D. President, Bi-Level Technologies Office: 760 753-0668 Mobile: 858 395-1440 [email protected] Selected-Works: http://works.bepress.com/rkatznelson/
The Hon. John J. Doll May 29, 2009 Acting Under Secretary of Commerce; Acting Director of the USPTO Mail Stop Comments-Patents Commissioner for Patents P.O.Box 1450, Alexandria, VA 22313-1450
Via e-mail to [email protected]
Re: Request for Comments and Notice of Roundtable on Deferred Examination for Patent Applications 74 Fed. Reg. 4946 (January 28, 2009).
Dear Acting Under Secretary Doll:
I wish to thank you for inviting me to participate in the USPTO’s February 12 Roundtable on Deferred Examination for Patent Applications (see webcast, agenda and attendee list). I commend you and your staff for engaging in this preliminary inquiry, which I hope will lead to a thoughtful and comprehensive rulemaking inquiry and ultimately to a rulemaking process adopting this practice.
This submission is made to complement my remarks made during the Roundtable, and to further explain the basis of my strong support for the Office’s adoption of an Examination On Request (“EOR”) procedures in order to reduce pendency and improve patent quality.
During the Roundtable, I alluded to my upcoming public policy paper on EOR and the attached slide set is a summary of my draft paper. Also included as an appendix to the slide set, you will find my draft of a second paper on EOR, which focuses solely on workload savings of EOR systems. The model and parameters analyzed in the paper predict that if the USPTO were to implement a three-year deferral EOR procedure, there would be between 15% to 25% savings in examination workload and an 8% to 11% drop in examined applications compared to the flow under current practice.
I look forward to further participation in your deliberative process on ways to reduce USPTO workload and increase patent quality.
Respectfully submitted by
Ron D. Katznelson, Ph.D. Encinitas, CA Office: (760) 753-0668 Mobile: (858) 395-1440
Attachment: PowerPoint presentation and Appendix
Exa
min
atio
n-O
n-R
eque
st
- A D
efer
red
Exa
min
atio
n
Prop
osal
for
the
U
.S. P
aten
t Off
ice
By
Ron
D. K
atzn
elso
n, P
h.D
.
Com
men
ts su
bmitt
ed to
USP
TO
May
29,
200
9
© R
on K
atzn
elso
n 1
Exa
min
atio
n O
n R
eque
st (E
OR
) -
Con
tent
¾ A
ccel
erat
ion
of p
aten
t cla
im o
bsol
esce
nce
and
the
skyr
ocke
ting
ap
plic
atio
n ba
cklo
g at
the
USP
TO c
all f
or a
dopt
ion
of E
OR
¾
Adv
anta
ges o
f EO
R
¾ I
nter
natio
nal e
xper
ienc
e w
ith E
OR
¾
An
EOR
syst
em p
ropo
sed
for t
he U
.S.
• Fo
r new
app
licat
ions
•
For e
xist
ing
appl
icat
ions
in U
SPTO
’s b
ackl
og
¾ A
naly
sis o
f the
pro
pose
d EO
R sy
stem
’s fi
t with
in th
e ex
istin
g st
atut
ory
fram
ewor
k.
¾
Add
ress
ing
conc
erns
rela
ted
to E
OR
•
Publ
ic N
otic
e D
elay
•
Late
Cla
imin
g ¾
Oth
er fu
rther
con
side
ratio
ns fo
r EO
R im
plem
enta
tion.
¾
Con
clus
ion
© R
on K
atzn
elso
n 2
Skyr
ocke
ting
appl
icat
ion
back
log
at th
e U
SPT
O c
alls
fo
r ad
optio
n of
Exa
min
atio
n O
n R
eque
st sy
stem
¾
The
USP
TO a
ccum
ulat
ed m
ore
than
1.2
mill
ion
appl
icat
ions
pen
ding
ex
amin
atio
n, th
ree
times
the
num
ber p
endi
ng a
dec
ade
ago.
¾
Inc
omin
g pa
tent
app
licat
ion
filin
g ra
te e
xcee
ds U
SPTO
’s e
xam
inat
ion
di
spos
al c
apab
ility
by
20%
.¾
The
USP
TO st
rate
gic
plan
fore
cast
s tha
t exa
min
atio
n re
sour
ce sh
ortfa
ll
com
pare
d to
inco
min
g ap
plic
atio
ns w
ill c
ontin
ue w
ell i
nto
2013
.1
¾E
xam
inat
ion
On
Req
uest
(AK
A D
efer
red
Exa
min
atio
n) c
an q
uick
ly
and
effe
ctiv
ely
reve
rse
this
20-
year
old
tren
d.
•A
ssum
ing
a 3-
year
def
erra
l sys
tem
, an
unex
amin
ed a
pplic
atio
n dr
opou
t rat
e of
8%
-11%
is e
xpec
ted.
Mor
eove
r, ev
en in
app
licat
ions
for w
hich
exa
min
atio
n w
ould
ulti
mat
ely
be re
ques
ted,
som
e or
igin
al c
laim
s wou
ld b
e w
ithdr
awn,
re
sulti
ng in
a la
rger
frac
tion
of w
ithdr
awn
clai
ms (
Cla
im D
ropo
ut R
ate)
. •
The
exam
inat
ion
wor
kloa
d sa
ving
s to
USP
TO w
ould
be
mor
e th
an th
e A
pplic
atio
n D
ropo
ut ra
te.
It w
ould
be
prop
ortio
nal t
o th
e C
laim
Dro
pout
Rat
e,
whi
ch is
exp
ecte
d to
be
in th
e ra
nge
of 1
5%-2
5%.
1.S
ee F
Y 2
009
Bud
get o
f the
USP
TO, a
t http
://w
ww
.usp
to.g
ov/w
eb/o
ffic
es/a
c/co
mp/
budg
/fy09
pbr.p
df
(Com
pare
‘UPR
App
licat
ions
File
d’ a
nd ‘U
PR D
ispo
sals
’ at p
age
19).
© R
on K
atzn
elso
n 3
EO
R G
oal:
Hal
t PT
O a
pplic
atio
n ba
cklo
g gr
owth
–
Ach
ieve
App
licat
ion
Loa
ding
Rat
io d
1 Lo
adin
g U
SPTO
's p
aten
t app
licat
ion
disp
osal
cap
abili
ty c
onsi
sten
tly fa
ils
Rat
io
to w
ithst
and
appl
icat
ion
filin
g ra
tes
0.25
0.50
0.75
1.00
1.25
Stab
le S
yste
m -
Ove
rload
ed U
nsta
ble
Syst
em -
US
PTO
FY-
09 S
trate
gic
Pla
n
Sour
ces:
USP
TO A
nnua
l Rep
orts
, UP
R a
pplic
atio
ns fi
led,
allo
wed
and
aba
ndon
ed;
US
PTO
FY-
2009
Bud
get R
eque
st
Act
ual
Num
ber o
f App
licat
ions
File
d Q
ueue
=
Load
ing
Rat
io
Num
ber o
f App
licat
ion
Dis
posa
ls
1.50
0.00
1982
19
84
1986
19
88
1990
19
92
1994
19
96
1998
20
00
2002
20
04
2006
20
08
2010
Fisc
al Y
ear
¾U
SPTO
fore
cast
s exa
min
atio
n re
sour
ces s
hortf
all o
f 15%
to 2
0% in
the
next
fis
cal y
ear.
An
EOR
pro
gram
can
mak
e up
for t
hat s
hortf
all,
ther
eby
achi
evin
g ap
plic
atio
n qu
euin
g st
abili
ty a
nd su
bseq
uent
pen
denc
y re
duct
ion.
4 ©
Ron
Kat
znel
son
EO
R ta
rget
s the
con
sequ
ence
s of p
rodu
ct li
fecy
cle
co
ntra
ctio
n an
d ac
cele
ratio
n of
cla
im o
bsol
esce
nce
¾ In
crea
se n
eed
for f
iling
Pro
visi
onal
and
Con
tinua
tion
appl
icat
ions
. • A
larg
er fr
actio
n of
orig
inal
app
licat
ions
requ
ire fo
llow
-up
late
cla
imin
g in
ord
er to
cov
er sp
ecifi
c ne
w
feat
ures
/pro
duct
s and
in o
rder
to a
ppro
pria
te e
quiv
alen
t re
turn
s fro
m in
vent
ions
. • R
esul
ts in
incr
ease
d nu
mbe
r of a
pplic
atio
ns.
¾ A
larg
er fr
actio
n of
cla
ims b
ecom
e ob
sole
te b
y th
e tim
e th
e U
SPTO
issu
es th
e pa
tent
. • T
hese
obs
olet
e ex
amin
ed a
pplic
atio
ns a
re p
rogr
essi
vely
less
lik
ely
to fe
tch
pate
nt re
new
al re
venu
es to
the
USP
TO.
• Sub
stan
tial e
xam
inat
ion
wor
kloa
d sa
ving
s cou
ld b
e ob
tain
ed
by n
ot e
xam
inin
g th
ese
obso
lete
cla
ims i
n th
e fir
st p
lace
.
© R
on K
atzn
elso
n 5
Con
trac
tion
of P
rodu
ct L
ifecy
cle
Acc
eler
ates
Pat
ent
Cla
im O
bsol
esce
nce
His
toric
al D
eclin
e in
Dur
atio
n of
Inte
rval
Pr
ior t
o C
ompe
titiv
e En
try
for I
nnov
atio
ns
100 18
80
1900
19
20
1940
19
60
1980
20
00
Inno
vatio
n In
trod
uctio
n Ye
ar
© R
on K
atzn
elso
n110 246820406080
'Monopoly' Duration (Years)
Hat
ch-
Wax
man
Act
Gen
eric
Phar
mac
eutic
als
Con
sum
er a
ndPr
oduc
er G
oods
2.9%
redu
ctio
npe
r yea
r
10.4
% re
duct
ion
per y
ear
3.8%
redu
ctio
npe
r yea
r
Hor
izon
tal b
ar s
pan
indi
cate
s av
erag
ing
perio
dov
er w
hich
dur
atio
n es
timat
es w
ere
obta
ined
Sour
ces:
C
onsu
mer
and
Pro
duce
rs G
oods
: R. A
garw
al &
M. G
ort,
Jour
nal O
f Law
& E
cono
mic
s 44
(1),
p. 1
61 (2
001)
.G
ener
ic P
harm
aceu
tical
s: J
.A. D
iMas
i & C
. Paq
uette
, Pha
rmac
oeco
nom
ics,
22,
Sup
pl. 2
, pp.
1, (
2004
).
6
Pate
nt L
ifetim
e E
rosi
on D
ue to
Pro
duct
Life
cycl
e Sh
orte
ning
Perc
ent
of p
aten
ts s
urvi
ving
afte
r re
new
al p
aym
ents
at
the
JPO
by
gran
t er
a.
Sou
rce:
Tok
yo I
nstit
ute
of
Inte
llect
ual P
rope
rty (2
006)
.
Pate
nt li
fetim
e at
the
USP
TO, E
PO
and
JPO
. H
alf-L
ife i
s th
e pa
tent
ag
e at
whi
ch 5
0% o
f th
e pa
tent
s ar
e no
t re
new
ed b
y th
eir
owne
rs.
Sou
rce:
Tril
ater
al P
aten
t Offi
ces
(200
6).
Pate
nt L
ife F
rom
Fili
ng D
ate
(Yea
rs)
Bef
ore
1970
1970
’s
1980
’s
Surv
ival
Rat
e
Hal
f-Life
8 9 10
11
12
13
14
15
EP
O (F
rom
app
licat
ion
date
)
JPO
(Fro
m g
rant
dat
e)
USP
TO
(Fro
m g
rant
dat
e)
Ren
ewal
Hal
f Life
(Y
ears
)
1996
19
98
2000
20
02
2004
20
06
Year
© R
on K
atzn
elso
n 7
Exa
min
atio
n-O
n-R
eque
st (E
OR
) Adv
anta
ges
¾T
o A
pplic
ant:
•
Perm
its d
efer
ral o
f the
pat
entin
g de
cisi
on u
ntil
the
tech
nolo
gy fe
atur
es
and
mar
ket o
ppor
tuni
ties a
re b
ette
r und
erst
ood.
•
Def
ers o
r sav
es u
ser f
ees a
nd e
ven
larg
er p
aten
t pro
secu
tion
cost
s. ¾
To
USP
TO
: •
Exam
inat
ion
wor
kloa
d sa
ving
s. •
Avo
id e
xam
inin
g cl
aim
s tha
t the
app
lican
t doe
s not
nee
d an
ymor
e.
• R
educ
tion
in th
e nu
mbe
r of R
CE
and
Con
tinua
tion
appl
icat
ions
. •
Impr
oved
qua
lity
by re
ceiv
ing
Sear
ch R
epor
ts a
nd p
rior a
rt fr
om
inte
rest
ed 3
rd p
artie
s hav
ing
spec
ial e
xper
tise
in th
e ar
t. •
Fina
ncia
l gai
ns b
y in
crea
sed
rene
wal
fee
yiel
d. E
xam
inat
ion
inve
stm
ents
wou
ld b
e m
ade
only
in p
aten
ts h
avin
g lo
nger
rene
wal
life
. ¾
To
the
Publ
ic:
•O
vera
ll pe
nden
cy re
duct
ion
and
redu
ced
pate
nt te
rm a
djus
tmen
ts.
•Ea
rlier
pub
lic n
otic
e fo
r non
-def
erre
d an
d se
cret
app
licat
ions
. •
Elim
inat
ion
of R
&D
cos
ts fo
r ‘de
sign
ing-
arou
nd’ “
clut
ter”
cla
ims t
hat
wou
ld o
ther
wis
e is
sue
unde
r the
cur
rent
syst
em.
© R
on K
atzn
elso
n 8
Inte
rnat
iona
l exp
erie
nce
at
natio
nal p
aten
t off
ices
em
ploy
ing
Exa
min
atio
n O
n R
eque
st (E
OR
) or
‘Def
erre
d E
xam
inat
ion’
sy
stem
s
© R
on K
atzn
elso
n 9
Application Ex amination Rate Exa
min
atio
n ra
te b
y pa
tent
app
licat
ion
year
and
by
ave
rage
num
ber
of c
laim
s at f
iling
that
yea
rEx
amin
atio
n R
ate
at N
atio
nal P
aten
t Offi
ces
p s10
0%
'85
'89 '
90
'92
'94
'86'
87'8
8'9
1'9
3 '9
5 '9
6 '97
'98
'99
'01 '02
'03 '0
4
'88 '8
9
'90
'91
'93
'92
'94
'95
'96 '9
7'98
'99
Sour
ces:
Exam
inat
ion
Rat
e: J
PO
Ann
ual R
epor
ts;
G. L
azar
idis
et a
l. W
orld
Pat
ent I
nfor
mat
ion
29, p
p. 3
17-3
26, (
2007
).A
vera
ge C
laim
s at
filin
g:
R.D
. Kat
znel
son,
Fed
eral
Circ
uit B
ar J
ourn
al, 1
7(1)
, pp.
1-30
, (20
07).
JPO
- 7
Year
s D
efer
ral,
FTF
Rat
ios:
0.9
1 -
0.89
JPO
- 3
Year
s D
efer
ral,
FTF
Rat
ios:
0.8
8 - 0
.87
EPO
- 19
to 2
4 M
onth
s D
efer
ral,
FTF
Rat
ios:
0.0
9
p 92%
90
%
80%
70%
70
%
60%
60%
50%
40%
30%
20%
10%
0% 0
1 2
3 4
5 6
7 8
9 10
11
1213
1415
1617
18
Ave
rage
Num
ber o
f Cla
ims
in A
pplic
atio
ns F
iled
on Y
ear a
nd a
tPat
ent O
ffice
Sho
wn
10
© R
on K
atzn
elso
n
Subs
tant
ial a
band
onm
ent u
nder
EPO
’s E
OR
follo
ws t
he S
earc
h
Rep
ort p
ublic
atio
n (t
he la
st 6
mon
ths o
f def
erra
l per
iod)
Aba
ndon
men
tAb
ando
nmen
t Rat
es P
rior t
o Ex
amin
atio
n by
Tec
h C
lust
er
Rat
e (%
) (B
ased
on
EPO
Exa
min
atio
n D
efer
ral A
vera
ge o
f 2 Y
ears
) 16
14
12
10 8 6 4 2 0
Sta
ge
Pos
t-Src
h. R
prt.
Pre
-Src
h. R
prt.
Audi
o vi
deo
med
ia
AVM
H
andl
ing
and
proc
essi
ng
HP
Bio
tech
nolo
gy
BIO
In
dust
rial c
hem
istr
y IC
C
ivil
engi
neer
ing
and
ther
mod
ynam
ics
CET
M
easu
ring
and
optic
s M
O
Com
pute
rs
CO
MP
Pure
and
app
lied
orga
nic
chem
istr
y PA
OC
El
ectr
onic
s EL
E Po
lym
ers
POL
Elec
tric
ity a
nd s
emic
ondu
ctor
s te
chno
logy
ES
C
Tele
com
s TE
L H
uman
nec
essi
ties
HN
Ve
hicl
es a
nd g
ener
al te
chno
logy
VG
T
Tech
nolo
gy C
lust
er D
esig
nato
rs
CO
MP
AVM
TE
L VG
T B
IO
CET
H
P ES
C
HN
EL
E PA
OC
PO
L M
O
IC
Tech
nolo
gy C
lust
er
Sour
ces:
G. L
azar
idis
et a
l. W
orld
Pat
ent I
nfor
mat
ion
29, p
p. 3
17-3
26, (
2007
) ©
Ron
Kat
znel
son
11
© R
on K
atzn
elso
n
2600
1600
2800 2800
36001700
3600
Fraction of Applications Abandoned Prior to Ex aminationEO
R’s
wor
kloa
d sa
ving
s at U
SPT
O w
ould
like
ly
be h
ighe
r w
here
they
are
mos
t nee
ded
Aba
ndon
men
t Rat
e Pr
ior t
o Ex
amin
atio
n by
Tec
hnol
ogy
Clu
ster
(Bas
ed o
n EP
O E
xam
inat
ion
Def
erra
l Ave
rage
of 2
Yea
rs)
16%
15%
14%
13%
12%
11%
10%
9% 8%
3 600
2100
3700
1700
2800
16 00
260 0
AV
M
BIO
C
ET
CO
MP
ELE
ESC
H
N
HP
IC
MO
PA
OC
P
OL
TEL
VG
T
Sour
ces:
Ave
rage
TC
Pen
denc
y: U
SPT
O, 2
008
Annu
al R
epor
t, p.
118
. Ex
amin
atio
n R
ate:
G. L
azar
idis
et a
l. W
orld
Pat
ent I
nfor
mat
ion
29, p
p. 3
17-3
26, (
2007
). ©
Ron
Kat
znel
son
Tech
nolo
gy�
C
ente
r �
Audi
o vi
deo
med
ia
AVM
Ha
ndlin
g an
d pr
oces
sing
H
P Bi
otec
hnol
ogy
BIO
In
dust
rial c
hem
istry
IC
Ci
vil e
ngin
eerin
g an
d th
erm
odyn
amic
s C
ET
Mea
surin
g an
d op
tics
MO
Co
mpu
ters
C
OM
P Pu
re a
nd a
pplie
d or
gani
c ch
emis
try
PAO
C
Elec
tron
ics
ELE
Po
lym
ers
POL
Elec
tric
ity a
nd s
emic
ondu
ctor
s te
chno
logy
E
SC
Tele
com
s TE
L Hu
man
nec
essi
ties
HN
Ve
hicl
es a
nd g
ener
al te
chno
logy
VG
T
Tech
nolo
gy C
lust
er D
esig
nato
rs
25
30
35
40
45
Ave
rage
Pen
denc
y at
USP
TO T
echn
olog
y C
ente
r (M
onth
s)
© R
on K
atzn
elso
n 12
A P
ropo
sal F
or P
aten
t E
xam
inat
ion-
On-
Req
uest
Sys
tem
Fo
r th
e U
nite
d St
ates
© R
on K
atzn
elso
n 13
Exa
min
atio
n-O
n-R
eque
st c
onsi
dera
tions
¾
Thi
s pro
posa
l wor
ks w
ithin
the
curr
ent s
tatu
te a
nd re
quir
es n
o ne
w le
gisl
atio
n ¾
Und
er th
e cu
rren
t sta
tuto
ry fe
e st
ruct
ure,
est
imat
ed fe
es th
at m
ay b
e de
ferr
ed
upon
a ru
le c
hang
e un
til a
Req
uest
for E
xam
inat
ion
are:
•
$540
Sea
rch
Fee
(Lar
ge E
ntity
); $2
20 E
xam
inat
ion
Fee
(Lar
ge E
ntity
); ~
$440
Ex
cess
Cla
im F
ees (
Inde
pend
ent a
nd T
otal
cla
ims f
ees,
aver
aged
on
all
appl
icat
ions
1 ).
Tota
l pot
entia
l def
erre
d #
$1,2
00.
•St
atut
ory
fee
ince
ntiv
e ch
ange
s sho
uld
be c
onsi
dere
d on
ly if
thes
e ex
istin
g in
cent
ives
pro
ve in
suff
icie
nt.
¾S
lide
11 sh
ows t
hat s
ubst
antia
l aba
ndon
men
t prio
r to
exam
inat
ion
is in
duce
d by
th
e Se
arch
Rep
ort (
SR).
Pref
erab
ly, U
.S. E
OR
shou
ld a
lso
have
an
SR fe
atur
e.
¾A
ssum
ing
a 3-
year
def
erra
l sys
tem
, an
appl
icat
ion
drop
out r
ate
of 8
%-1
1%
with
a c
laim
dro
pout
rate
of 1
5%-2
5% w
ould
be
expe
cted
. (Se
e A
ppen
dix)
. ¾
A 2
0% c
laim
dro
pout
rate
wou
ld m
ore
than
com
pens
ate
for U
SPTO
’s p
roje
cted
ex
amin
atio
n sh
ortfa
ll –
halti
ng a
nd su
bseq
uent
ly re
duci
ng p
ende
ncy.
1.
Bas
ed o
n FY
2006
dat
a pr
oduc
ed in
the
Tafa
s v. U
SPTO
(200
7) c
ase
, cor
rect
ed fo
r rec
ent f
ee c
hang
es a
nd
estim
ated
ave
rage
incr
ease
in c
laim
cou
nt si
nce
FY20
06.
© R
on K
atzn
elso
n 14
Prop
osed
EO
R S
yste
m –
New
App
licat
ions
¾
(a
) Per
mit
appl
ican
ts to
ele
ct b
y si
mpl
e ch
eckb
oxes
bet
wee
n tw
o op
tions
upo
n fil
ing
a ne
w a
pplic
atio
n:
•(i)
Pro
ceed
to e
xam
inat
ion
as in
cur
rent
rule
s or;
• (ii
) Def
er e
xam
inat
ion
until
an
expr
ess r
eque
st fo
r exa
min
atio
n is
file
d w
ithin
[3] y
ears
afte
r th
e ap
plic
atio
n fil
ing
date
(Max
imum
Def
erra
l Per
iod)
. ¾
(b
) Upo
n m
akin
g a
defe
rral
ele
ctio
n, a
n ap
plic
ant i
s req
uire
d to
exe
cute
a D
ecla
ratio
n &
N
on-E
xclu
sive
Lim
ited
Pow
er o
f Atto
rney
, for
the
purp
oses
of t
he e
x pa
rte
proc
ess
desc
ribed
bel
ow a
nd in
subs
eque
nt sl
ides
. ¾
(c
) Any
del
ay d
ue to
app
lican
t’s e
xam
inat
ion
defe
rral
wou
ld b
e su
btra
cted
from
any
pa
tent
term
ext
ensi
on c
redi
t to
the
appl
ican
t und
er 3
5 U
.S.C
. § 1
54(b
). ¾
(d
) App
licat
ions
subj
ect t
o se
crec
y el
ectio
n un
der 3
5 U
.S.C
. § 1
22(b
)(2)
(B) a
re e
xclu
ded
from
def
erra
l and
will
be
exam
ined
aut
omat
ical
ly a
s in
curr
ent r
ules
. ¾
(e
) Upo
n a
decl
arat
ion
unde
r 37
C.F
.R. §
10.
18, p
aym
ent o
f fee
s and
pur
suan
t to
the
Lim
ited
Pow
er o
f Atto
rney
, any
third
par
ty c
an (a
nony
mou
sly)
requ
est a
sear
ch a
nd/o
r ex
amin
atio
n of
a p
ublis
hed
appl
icat
ion
for w
hich
no
exam
inat
ion
requ
est w
as fi
led.
The
re
ques
ter m
ay a
lso
subm
it pr
ior a
rt in
acc
orda
nce
with
37
C.F
.R. §
1.9
9 at
that
tim
e.
¾
(f) A
pplic
atio
ns fo
r whi
ch n
o re
ques
t for
exa
min
atio
n is
file
d w
ithin
the
Max
imum
D
efer
ral P
erio
d w
ould
be
deem
ed a
band
oned
. A
ppro
pria
te st
atut
ory
§ 13
3 sa
fegu
ards
fr
om a
band
onm
ent w
ould
be
prov
ided
thro
ugh
exte
nsio
n of
tim
e pe
titio
ns a
nd th
e lik
e.
© R
on K
atzn
elso
n 15
Prop
osed
con
ditio
ns fo
r PT
O’s
acc
epta
nce
of
appl
ican
ts’ e
lect
ion
for
new
app
licat
ion
defe
rral
¾
The
app
licat
ion
is in
con
ditio
n fo
r pub
licat
ion
and
all s
tatu
tory
§11
1(a)
re
quire
men
ts a
re c
ompl
ete
exce
pt fo
r •
Paym
ent o
f Exa
min
atio
n Fe
e, E
xces
s Cla
im F
ees a
nd E
xces
s pag
es F
ee.
•Pa
ymen
t of S
earc
h Fe
e (O
ptio
nal).
•
Subm
issi
on o
f an
IDS
wou
ld b
e op
tiona
l, bu
t sub
ject
to a
n ID
S fe
e if
filed
mor
e th
an[1
4] m
onth
s afte
r the
app
licat
ion
filin
g da
te. [
This
pre
-pub
licat
ion
IDS
cond
ition
ben
efits
th
ird p
artie
s who
may
trig
ger a
n ex
amin
atio
n an
d/or
sear
ch b
ased
on
the
publ
icat
ion ]
¾
The
app
lican
t has
not
file
d a
nonp
ublic
atio
n re
ques
t und
er §
122(
b)(2
)(B
)(i),
or
has f
iled
a re
ques
t und
er §
122(
b)(2
)(B
)(ii)
to re
scin
d a
prev
ious
ly fi
led
nonp
ublic
atio
n re
ques
t. ¾
The
App
lican
t exe
cute
d an
d fil
ed w
ith th
e U
SPTO
a D
ecla
ratio
n &
Non
-E
xclu
sive
Lim
ited
Pow
er o
f Atto
rney
des
igna
ting
any
third
par
ty to
act
on
the
App
lican
t’s b
ehal
f as A
pplic
ant’s
aut
horiz
ed a
gent
for p
erfe
ctin
g th
e pa
tent
ap
plic
atio
n un
der §
111(
a) a
t any
tim
e by
: •
Req
uest
ing
exam
inat
ion;
•
Payi
ng th
e Ex
amin
atio
n, E
xces
s Cla
im a
nd E
xces
s pag
es F
ees;
and
•
Req
uest
ing
a Se
arch
and
pay
ing
the
Sear
ch F
ee, o
r sub
mitt
ing
a Se
arch
Rep
ort [
in th
e m
anne
r req
uire
d un
der A
ccel
erat
ed E
xam
inat
ion,
see
MPE
P §
708.
02(a
)(I)
(H)]
.
© R
on K
atzn
elso
n 16
Prop
osed
EO
R S
yste
m –
App
licat
ions
in b
ackl
og
¾
(a) A
n ap
plic
atio
n al
read
y pe
ndin
g in
PTO
’s b
ackl
og w
ould
be
sele
cted
as e
ligib
le fo
r def
erra
l if i
t m
eets
the
follo
win
g co
nditi
ons (
Elig
ible
App
licat
ions
): •
The
appl
icat
ion
rece
ived
no
Firs
t Act
ion
On
the
Mer
its (F
AO
M).
•Th
e ap
plic
atio
n is
not
subj
ect t
o se
crec
y el
ectio
n un
der 3
5 U
.S.C
. §12
2(b)
(2)(
B).
• Th
e ap
plic
atio
n w
as fi
led
no m
ore
than
[34]
mon
ths p
rior t
o th
e EO
R ru
le’s
eff
ectiv
e da
te.
¾
(b) P
rior t
o a
FAO
M o
n El
igib
le A
pplic
atio
ns, t
he P
TO w
ould
send
a R
efun
d El
ectio
n A
ctio
n to
al
l app
lican
ts o
f Elig
ible
App
licat
ions
, req
uirin
g af
firm
ativ
e re
spon
se w
ithin
[60]
day
s by
elec
ting
one
of th
e fo
llow
ing
optio
ns:
•(i)
Exa
min
atio
n de
ferr
al e
lect
ion
unde
r whi
ch th
e ap
plic
ant w
ould
rece
ive
a re
fund
of t
he E
xam
inat
ion
Fee.
Th
e ap
plic
ant w
ould
be
requ
ired
to m
ake
the
appl
icat
ion
read
y fo
r pub
licat
ion
and
to e
xecu
te a
Dec
lara
tion
& N
on-E
xclu
sive
Lim
ited
Pow
er o
f Atto
rney
, for
the
purp
oses
des
crib
ed b
elow
and
in su
bseq
uent
slid
es.
Exam
inat
ion
defe
rral
wou
ld b
e up
to th
e M
axim
um D
efer
ral P
erio
d af
ter t
he a
pplic
atio
n da
te.
•(ii
) Ele
ctio
n fo
r mai
ntai
ning
the
appl
icat
ion’
s exa
min
atio
n tra
ck, f
orgo
ing
the
Exam
inat
ion
Fee
refu
nd.
•Th
e ap
plic
atio
n w
ould
be
aban
done
d du
e to
failu
re to
resp
ond
to th
e R
efun
d El
ectio
n A
ctio
n af
ter t
he
stat
utor
y no
tice
and
resp
onse
per
iod.
¾
(c
) Any
del
ay d
ue to
app
lican
t’s e
xam
inat
ion
defe
rral
wou
ld b
e su
btra
cted
from
any
pat
ent t
erm
ex
tens
ion
cred
it to
the
appl
ican
t und
er 3
5 U
.S.C
. §15
4(b)
. ¾
(d
) Upo
n a
decl
arat
ion
unde
r 37
C.F
.R. §
10.
18, p
aym
ent o
f Exa
min
atio
n Fe
e an
d pu
rsua
nt to
the
Lim
ited
Pow
er o
f Atto
rney
, any
third
par
ty c
an (a
nony
mou
sly)
requ
est e
xam
inat
ion
of a
def
erre
d pu
blis
hed
appl
icat
ion
for w
hich
no
exam
inat
ion
requ
est w
as fi
led.
The
requ
este
r may
als
o su
bmit
prio
r art
in a
ccor
danc
e w
ith 3
7 C
.F.R
. § 1
.99
at th
at ti
me.
¾
(e
) App
licat
ions
for w
hich
no
requ
est f
or e
xam
inat
ion
is fi
led
with
in th
e M
axim
um D
efer
ral
Perio
d w
ould
be
deem
ed a
band
oned
. App
ropr
iate
stat
utor
y §
133
safe
guar
ds fr
om a
band
onm
ent
wou
ld b
e pr
ovid
ed th
roug
h ex
tens
ion
of ti
me
petit
ions
and
the
like.
© R
on K
atzn
elso
n 17
Prop
osed
Con
ditio
ns F
or P
TO
’s E
xtra
ctio
n of
A
pplic
atio
ns fr
om th
e ba
cklo
g qu
eue
for
Def
erra
l¾
The
app
licat
ion
has r
ecei
ved
no F
AO
M a
nd h
as b
een
pend
ing
no m
ore
than
[34]
mon
ths.
¾ T
he a
pplic
atio
n is
in c
ondi
tion
for p
ublic
atio
n an
d al
l §11
1(a)
requ
irem
ents
ar
e co
mpl
ete
and
•A
pplic
ant r
eque
sted
a re
fund
of t
he E
xam
inat
ion
Fee,
reve
rsin
g §1
11(a
) co
mpl
eten
ess.
•Su
bmis
sion
of a
n ID
S w
ould
be
optio
nal a
t the
Def
erra
l Ele
ctio
n da
te, b
ut
subj
ect t
o an
IDS
fee
if fil
ed a
fter t
he D
efer
ral E
lect
ion
date
. ¾
The
app
lican
t has
not
file
d a
nonp
ublic
atio
n re
ques
t und
er §
122(
b)(2
)(B
)(i),
or
has
file
d a
requ
est u
nder
§12
2(b)
(2)(
B)(
ii) to
resc
ind
a pr
evio
usly
file
d no
npub
licat
ion
requ
est.
¾ T
he A
pplic
ant e
xecu
ted
and
filed
with
the
USP
TO a
Dec
lara
tion
& N
on-
Exc
lusi
ve L
imite
d Po
wer
of A
ttorn
ey d
esig
natin
g an
y th
ird p
arty
to a
ct o
n th
e A
pplic
ant’s
beh
alf a
s App
lican
t’s a
utho
rized
age
nt fo
r per
fect
ing
the
pate
nt a
pplic
atio
n un
der §
111(
a) a
t any
tim
e by
: •
Req
uest
ing
exam
inat
ion
and
payi
ng th
e Ex
amin
atio
n Fe
e; a
nd
•Su
bmitt
ing
a Se
arch
Rep
ort [
in th
e m
anne
r req
uire
d un
der A
ccel
erat
ed
Exam
inat
ion,
see
MPE
P §
708.
02(a
)(I)
(H)]
.
© R
on K
atzn
elso
n 18
The
pro
pose
d E
OR
syst
em is
cab
ined
w
ithin
the
exis
ting
stat
utor
y fr
amew
ork
¾ N
o re
ach
for a
genc
y C
hevr
on d
efer
ence
is re
quire
d be
caus
e th
e pr
opos
ed
EOR
syst
em is
with
in th
e in
tent
and
cle
ar la
ngua
ge o
f the
stat
ute.
¾
USP
TO a
utho
rity
unde
r 35
U.S
.C. §
2(b)
(2) f
or p
rom
ulga
ting
the
prop
osed
EO
R sy
stem
is a
lso
supp
orte
d by
the
legi
slat
ive
hist
ory
of 3
5 U
.S.C
. § 4
1:
In 2
003,
the
USP
TO a
sked
Con
gres
s to
appr
ove
a ne
w u
ser f
ee st
ruct
ure
spec
ifica
lly to
ena
ble
the
sepa
ratio
n of
the
fili
ng d
ecis
ions
fro
m t
he s
earc
h an
d ex
amin
atio
n de
cisi
ons
that
app
lican
ts m
ake.
The
USP
TO
subm
itted
its
fee
rest
ruct
urin
g pr
opos
al to
Con
gres
s w
ith s
peci
fic la
ngua
ge in
the
Adm
inis
tratio
n’s
prop
osed
bi
ll en
title
d “U
nite
d St
ates
Pat
ent
and
Trad
emar
k Fe
e M
oder
niza
tion
Act
of 2
003 ”
.1 In
sup
port
of i
ts
prop
osed
legi
slat
ion,
the
PTO
sub
mitt
ed to
Con
gres
s a
“Pro
posa
l to
Rest
ruct
ure
Pate
nt a
nd T
rade
mar
k Fe
es
and
Prac
tices
for
Fisc
al Y
ear
2003
”2 de
taili
ng th
e us
er f
ee s
epar
atio
n an
d br
eakd
own.
In
ful
fillin
g ev
ery
aspe
ct o
f th
e U
SPTO
req
uest
, C
ongr
ess
spec
ifica
lly a
utho
rized
the
USP
TO u
nder
the
Con
solid
ated
A
ppro
pria
tions
Act
of
2004
3 to c
harg
e se
para
tely
for
the
user
fee
com
pone
nts,
ther
eby
enab
ling
USP
TO’s
im
plem
enta
tion
of s
epar
ate
refu
nds
and
Exam
inat
ion
On
Req
uest
pro
cedu
res.
For
exa
mpl
e, p
rior
to
enac
tmen
t, th
e H
ouse
Rep
ort s
peci
fical
ly id
entif
ied
the
Exam
inat
ion
Fee
as p
ayab
le la
ter t
han
the
filin
g fe
e.4
1 At h
ttp://
ww
w.u
spto
.gov
/web
/offi
ces/
com
/stra
t21/
feeb
ill.h
tm
2 At h
ttp://
web
.arc
hive
.org
/web
/200
3120
8101
011/
http
://w
ww
1.us
pto.
gov/
web
/offi
ces/
com
/stra
t21/
actio
n/sr
1fr1
.htm
3 P
ub. L
. 108
–447
, 118
Sta
t. 28
09 (2
004)
. (A
men
ds 3
5 U
.S.C
. § 4
1 in
a m
anne
r tha
t sep
arat
es a
pplic
atio
n fe
e in
to fi
ling
fee,
sear
ch
fee
and
exam
inat
ion
fee
to p
erm
it se
para
te p
aym
ents
or r
efun
ds o
f exa
min
atio
n an
d se
arch
fees
). 4
See
H. R
. Rep
. 108
-241
, at 1
6 (2
003)
(“Th
e ex
amin
atio
n fe
e, h
owev
er, m
ay b
e pa
id a
t a la
ter t
ime
if pa
id w
ithin
such
per
iod
and
unde
r suc
h co
nditi
ons (
incl
udin
g pa
ymen
t of a
surc
harg
e) a
s may
be
pres
crib
ed b
y th
e D
irec
tor .”
Em
phas
is a
dded
).
19
© R
on K
atzn
elso
n
Fitt
ing
EO
R P
ract
ice
with
in th
e ex
istin
g st
atut
ory
fram
ewor
k
– PT
O a
nd A
pplic
ants
’ use
of t
he a
utho
rize
d fe
e st
ruct
ure
¾D
efer
ral w
ould
resu
lt fr
om P
TO’s
new
rule
-bas
ed “
sche
dule
d ac
quie
scen
ce”,
trig
gere
d by
an
App
lican
t’s D
efer
ral E
lect
ion
re
ques
ting
to k
eep
an a
pplic
atio
n in
com
plet
e un
der §
111
(a):
•§
111(
a)(3
): “T
he a
pplic
atio
n m
ust b
e ac
com
pani
ed b
y th
e fe
e re
quir
edby
law
. The
fee
and
oath
may
be
subm
itted
afte
r the
spec
ifica
tion
and
any
requ
ired
draw
ing
are
subm
itted
, with
in su
ch p
erio
d an
d un
der s
uch
cond
ition
s, in
clud
ing
the
paym
ent o
f a su
rcha
rge,
as m
ay b
e pr
escr
ibed
by th
e D
irec
tor.”
•
“Fee
requ
ired
by la
w”
perti
nent
to d
efer
ral i
nclu
de §
41(
a)(3
)(A
) –
Exam
inat
ion
Fee;
§ 4
1(a)
(1)(
B) –
Exc
ess c
laim
fees
; § 4
1(a)
(1)(
G) –
Ex
cess
pag
es fe
es; a
nd §
41(
d)(1
)(A
) – S
earc
h fe
e.
¾N
ote
also
that
, whe
reas
the
stat
ute
requ
ires t
hat t
he fi
ling
fee
be
paid
“O
n fil
ing
each
app
licat
ion…
” §
41(a
)(1)
(A),
no su
ch
requ
irem
ent p
erta
ins t
o th
e ot
her f
ees l
iste
d ab
ove.
1
1. S
ee a
lso
H. R
. Rep
. 108
-241
, at 1
6 (2
003)
(“Th
e ex
amin
atio
n fe
e, h
owev
er, m
ay b
e pa
id a
t a la
ter t
ime
if pa
id w
ithin
such
per
iod
and
unde
r suc
h co
nditi
ons (
incl
udin
g pa
ymen
t of a
surc
harg
e) a
s may
be
pres
crib
ed b
y th
e D
irec
tor.”
© R
on K
atzn
elso
n 20
Fitt
ing
EO
R p
ract
ice
with
in th
e ex
istin
g st
atut
ory
fr
amew
ork
– Pa
tent
Ter
m A
djus
tmen
ts
¾ B
ecau
se P
aten
t Ter
m A
djus
tmen
ts (P
TA) d
ue to
USP
TO d
elay
s in
Firs
t Act
ion
unde
r §15
4(b)
(1)(
A) a
re k
eyed
-off
the
date
on
whi
ch a
n ap
plic
atio
n is
mad
e co
mpl
ete
unde
r §11
1(a)
, app
lican
ts
cann
ot re
ceiv
e PT
A c
redi
t for
the
defe
rral
per
iod
prio
r to
Firs
t A
ctio
n un
der c
urre
nt la
w.
See
§ 15
4(b)
(1)(
A)(
i)(I)
. ¾
Bec
ause
§ 1
54(b
)(1)
(B)(
iii) e
xclu
des f
rom
PTA
cre
dit “
any
dela
y in
the
proc
essi
ng o
f the
app
licat
ion
by th
e [U
SPTO
] req
uest
ed b
yth
e ap
plic
ant..
”, is
suan
ce o
f pat
ents
with
mor
e th
an 3
yea
rs
pend
ency
wou
ld n
ot e
ntitl
e ap
plic
ants
to P
TA c
redi
t due
to a
de
lay
attri
buta
ble
to th
eir e
lect
ion
to d
efer
exa
min
atio
n.
¾ N
ote
also
USP
TO’s
gen
eral
aut
horit
y in
§ 1
54(b
)(2)
(C)(
iii):
“The
Dire
ctor
shal
l pre
scrib
e re
gula
tions
est
ablis
hing
the
circ
umst
ance
s th
at c
onst
itute
a fa
ilure
of a
n ap
plic
ant t
o en
gage
in re
ason
able
eff
orts
to
conc
lude
pro
cess
ing
or e
xam
inat
ion
of a
n ap
plic
atio
n.”
© R
on K
atzn
elso
n 21
Fitt
ing
EO
R p
ract
ice
with
in th
e ex
istin
g st
atut
ory
fram
ewor
k –
“unp
erfe
ctin
g” b
ackl
og a
pplic
atio
ns
¾
Und
er th
e cu
rren
t sta
tute
, the
PTO
act
ually
has
no
auth
ority
to su
spen
d th
e ex
amin
atio
n of
an
app
licat
ion
that
mee
ts th
e st
atut
ory
requ
irem
ents
: “T
he D
irect
or sh
all c
ause
an
exam
inat
ion
to b
e m
ade
of th
e ap
plic
atio
n an
d th
e al
lege
d ne
w in
vent
ion;
and
if o
n su
ch e
xam
inat
ion
it ap
pear
s tha
t the
app
lican
t is e
ntitl
ed to
a p
aten
t und
er th
e la
w, t
he D
irect
or sh
all i
ssue
a
pate
nt th
eref
or.”
§ 1
31.
¾
Mor
eove
r, ha
ving
pai
d al
l fee
s and
hav
ing
com
plie
d w
ith §
111(
a), a
pplic
ants
wou
ld h
ave
very
littl
e in
cent
ive
to d
efer
exa
min
atio
n of
an
exis
ting
appl
icat
ion
in th
e ba
cklo
g.
¾
How
ever
, App
lican
t’s re
ques
t and
rece
ipt o
f a re
fund
of t
he E
xam
inat
ion
Fee
wou
ld
reve
rse
§111
(a) c
ompl
eten
ess o
f the
app
licat
ion.
The
sear
ch fe
e fo
r Elig
ible
App
licat
ions
m
ay n
ot b
e re
fund
able
bec
ause
sear
ches
mig
ht h
ave
com
men
ced
on su
ch a
pplic
atio
ns p
rior
to a
FA
OM
.1 [D
epen
ding
on
PTO
’s fi
nanc
ial c
onsi
dera
tions
, an
offe
r to
refu
nd th
e Ex
cess
Cla
im a
nd
Exce
ss P
ages
Fee
s mig
ht a
lso
be a
ppro
pria
te to
incr
ease
app
lican
ts’ i
ncen
tive
to d
efer
]. ¾
Si
nce
exam
inat
ion
wou
ld b
e de
ferr
ed a
nd th
e Ex
amin
atio
n Fe
e w
ould
be
paid
upo
n la
ter
com
men
cem
ent (
or m
ay n
ever
be
requ
ired)
, USP
TO’s
gen
eral
aut
horit
y to
issu
e re
fund
s th
eref
or a
re p
rovi
ded
by st
atut
e: “
The
Dire
ctor
may
refu
nd a
ny fe
e pa
id b
y m
ista
ke o
r any
am
ount
pai
d in
exc
ess o
f tha
t req
uire
d.”
§ 42
(d).
¾
Maj
or d
rop-
off s
avin
gs to
the
PTO
may
com
e fr
om a
pplic
ants
who
wou
ld n
ot re
spon
d to
U
SPTO
’s R
efun
d El
ectio
n A
ctio
n. A
utho
rity
for s
uch
aban
donm
ent p
rior t
o FA
OM
is
prov
ided
by
stat
ute
“[u]
pon
failu
re o
f the
app
lican
t to
pros
ecut
e th
e ap
plic
atio
n w
ithin
six
mon
ths a
fter a
ny a
ctio
n th
erei
n…”
§ 13
3.
1. H
owev
er, t
he U
SPTO
reso
lved
this
issu
e, p
erm
ittin
g su
ch re
fund
s in
its S
earc
h Fe
e R
efun
d ru
les.
See
71
Fed
Reg.
122
81 (M
ar. 1
0, 2
006)
.
© R
on K
atzn
elso
n 22
78
54
21
Prop
osed
Exa
min
atio
n-O
n-R
eque
st fo
r U
SPT
O
© R
on K
atzn
elso
n
QU
EUIN
G L
EGEN
D
56 12
11
10
4 2 1
9 3
Perio
d T
afte
r Ev
ent X
's d
ate
Exit
Tim
e Q
ueue
App
licat
ions
que
uing
inde
pend
ently
of o
ther
App
licat
ions
. E
xit
from
the
queu
e is
at
perio
d T
afte
r Eve
nt X
FIFO
Q
ueue
App
licat
ions
que
uing
on a
Firs
t-In-
Firs
t-Out
(F
IFO
) bas
is. E
xit f
rom
queu
e de
pend
s on
Ser
ver r
ate.
YES
Def
erre
d Ex
amin
atio
n Q
ueue
Publ
ishe
d A
pplic
atio
ns
Appl
icat
ions
With
SR R
eque
st
Sear
ch &
R
epor
t
Publ
ic
Appl
icat
ion
?
7 6 12
11
10
8
2 1
9 3
18 m
onth
s af
ter
prio
rity
date
Publ
icat
ion
Que
ue
Sear
ch
Que
ue
Appl
icat
ions
W/O
SR R
eque
st
Def
erre
d Ex
amin
atio
n ?
NO
Exam
inat
ion
Que
ue
YES
NO
7 5
6 1211
10
8 4
9 3
3 ye
ars
afte
r fil
ing
date
With
in s
tatu
tory
re
spon
se p
erio
d fo
llow
ing
US
PTO
no
tice
Appl
ican
t or 3
rdPa
rty
SR R
eque
st
Appl
ican
t or 3
rdPa
rty
Exam
inat
ion
Req
uest
Exam
inat
ion
Appl
icat
ion
com
plet
e pe
r §1
11(a
) ?
§122
Sec
recy
El
ectio
n
YES
NO
ABAN
DO
NED
"S
R"
= Se
arch
Rep
ort
Not
ice
to A
pplic
ant
For n
ew a
pplic
atio
ns o
nly
Extr
actio
n C
ondi
tions
(a
) Pub
lic
Appl
icat
ion
(b) A
pplic
ant
Req
uest
ed
Pre
- FAO
M
Exis
ting
Bac
klog
Q
ueue
Appl
ied
with
in la
st 3
4 m
onth
s
Def
ined
By
Effe
ctiv
e D
ate
Cut
ooff
YES Se
rver
Ser
ving
the
prec
edin
g qu
eue.
Ser
vice
tim
e de
pend
s on
reso
urce
s.S
R S
ervi
ce s
houl
d be
in
tim
e fo
r pub
licat
ion
dead
line
ABAN
DO
NED
Appl
ican
t Def
erra
l Dec
isio
n R
efun
d El
ectio
n Ac
tion
& D
efer
ral O
ptio
n N
otic
e to
App
lican
t
Appl
ican
t's R
espo
nse
No
Res
pons
e EO
R S
ELEC
TIO
N F
OR
EXI
STIN
G B
AC
KLO
G A
PPLI
CA
TIO
NS
© R
on K
atzn
elso
n 23
Del
ayed
Pub
lic N
otic
e C
once
rns
¾ T
hird
par
ties m
ay re
ques
t exa
min
atio
n in
ord
er to
pr
even
t a p
aten
tee
from
hol
ding
off
pro
secu
tion
of a
n ap
plic
atio
n th
at o
ther
s fin
d pr
oble
mat
ic a
nd in
nee
d of
ea
rly p
aten
tabi
lity
reso
lutio
n.
¾S
ecre
t app
licat
ions
wou
ld n
ot b
e de
ferr
ed.
¾ P
ublic
not
ice
argu
men
t as t
o th
e 18
mon
ths d
elay
in
third
-par
ties'
abili
ty to
requ
est e
xam
inat
ion
is
parti
cula
rly m
oot i
n vi
ew o
f the
long
er F
AO
M d
elay
s.1
¾ A
sser
tion
that
EO
R “
harm
s inn
ovat
ors d
ue to
del
ays i
n pu
blic
not
ice
of p
aten
t sco
pe”
is sp
ecul
ativ
e fo
lklo
re,
unsu
ppor
ted
by fa
cts o
r ana
lysi
s. Se
e ne
xt sl
ide.
1.
Firs
t Off
ice
Act
ion
aver
age
pend
ency
is c
urre
ntly
at 2
5.6
mon
ths,
and
grow
ing.
See
USP
TO's
FY
2008
Ann
ual R
epor
t, at
pag
e 62
.
© R
on K
atzn
elso
n 24
Giv
en su
bsta
ntia
l USP
TO
wor
kloa
d sa
ving
s of E
OR
, tho
se o
bjec
ting
to it
on
gro
unds
of “
harm
to in
nova
tors
due
to d
elay
ed p
ublic
not
ice”
mus
t bea
r th
e bu
rden
of s
how
ing
that
EO
R’s
pub
lic n
otic
e ne
t eff
ects
are
har
mfu
l
¾W
ithou
t EO
R’s
wor
kloa
d sa
ving
s, pu
blic
not
ice
by a
ll is
sued
pat
ents
wou
ld b
e de
laye
d fu
rthe
r due
to ru
naw
ay p
ende
ncy
incr
ease
s (se
e ne
xt sl
ide)
. ¾
Und
er E
OR
, eve
ry d
efer
red
appl
icat
ion
caus
es a
ll la
ter a
pplic
atio
ns to
mov
e ou
t of
turn
ahe
ad in
the
exam
inat
ion
queu
e, re
sulti
ng in
ear
lier p
ublic
not
ice
for t
hese
othe
r app
licat
ions
. A
re p
ublic
not
ice
dela
ys in
def
erre
d ap
plic
atio
ns re
ally
mor
e ha
rmfu
l tha
n th
e be
nefit
s of t
he e
arlie
r pub
lic n
otic
e in
oth
er is
sued
app
licat
ions
? ¾
Bec
ause
secr
et a
pplic
atio
ns u
nder
§12
2 el
ectio
n w
ould
not
be
elig
ible
for d
efer
ral,
they
wou
ld a
ll m
ove
ahea
d in
the
exam
inat
ion
queu
e an
d w
ould
pub
lish
earli
er
than
und
er th
e cu
rren
t exa
min
atio
n sy
stem
. Th
eref
ore,
any
pub
lic n
otic
e ha
rm
from
secr
et “
subm
arin
e” p
aten
ts w
ould
be
redu
ced
unde
r the
EO
R sy
stem
. ¾
The
EO
R sy
stem
wou
ld e
nabl
e so
me
dire
ct fi
ling
of N
onpr
ovis
iona
l def
erre
d ap
plic
atio
ns in
stea
d of
usi
ng th
e cu
rren
t pra
ctic
e of
Pro
visi
onal
app
licat
ions
. Th
e fo
rmer
type
wou
ld b
e pu
blis
hed
earl
ier,
adva
ncin
g th
e pu
blic
not
ice
by o
ne y
ear.
¾W
ithou
t EO
R, i
nnov
ator
s are
taxe
d by
hav
ing
to in
vest
R&
D in
non
-infr
ingi
ng
solu
tions
“de
sign
ing-
arou
nd”
pate
nt c
laim
s tha
t wou
ld h
ave
neve
r iss
ued
unde
r an
EOR
syst
em. T
his m
ajor
pub
lic n
otic
e ha
rm w
ould
be
elim
inat
ed u
nder
EO
R.
© R
on K
atzn
elso
n 25
Pate
nt T
erm
Adj
ustm
ent D
ue to
PT
O D
elay
s Th
ese
pend
ency
del
ays a
re ty
pica
lly in
add
ition
to
the
stat
utor
y 3-
year
pen
denc
y ba
selin
eFr
actio
n of
pat
ents
B
allo
onin
g Pa
tent
Ter
m In
crea
ses
with
term
adj
uste
dby
mor
e th
an T
U
nder
35
USC
§15
4(b)
1
01
23
45
T -
Pat
ent T
erm
Incr
ease
s D
ue T
o U
SPTO
Del
ays
(Yea
rs)
26
© R
on K
atzn
elso
n
0.00
1
0.01
0.1
April
- 20
07 G
rant
s
April
- 20
05 G
rant
s
Sou
rce:
US
PTO
Pat
ent I
ssue
Dat
a at
ftp:
//ftp
.usp
to.g
ov/p
ub/p
atda
ta/
Sept
embe
r - 2
008
Gra
nts
April
- 20
03 G
rant
s
“Lat
e C
laim
ing”
Obj
ectio
ns a
re a
‘Red
Her
ring
’¾
Obj
ectio
ns o
n th
e gr
ound
s tha
t EO
R p
erm
its “
late
cl
aim
ing”
igno
re a
ver
y si
mpl
e fa
ct th
at a
pplic
ants
may
(a
nd) d
o so
at a
ny ti
me
unde
r the
cur
rent
syst
em b
y fil
ing
new
cla
ims i
n C
ontin
uatio
ns.
• In
fact
, und
er E
OR
, the
nee
d fo
r lat
e cl
aim
ing
in c
hain
s of
Con
tinua
tions
that
kee
p ca
ses “
live”
wou
ld n
ot b
e ne
cess
ary
and
wou
ld li
kely
dim
inis
h.
¾ “L
ate
clai
min
g” c
once
rns m
ust b
e ad
dres
sed
not b
y re
ject
ing
the
EOR
syst
em b
ut b
y a
broa
der d
ebat
e on
the
pate
nt b
arga
in b
alan
ce o
f int
eres
ts.
• Thi
s bal
ance
is c
urre
ntly
stru
ck in
§12
0 by
per
mitt
ing
appl
ican
ts to
pro
tect
thei
r inv
entio
ns b
y la
te c
laim
ing
prov
ided
th
e cl
aim
s hav
e fu
ll §1
12 su
ppor
t in
the
orig
inal
dis
clos
ure
(whi
ch th
e pu
blic
rece
ives
with
out a
ny d
efer
ral).
© R
on K
atzn
elso
n 27
Oth
er a
dvan
tage
s to
the
publ
ic –
Sh
orte
r pr
ivat
e ri
ght p
erio
d
¾D
elay
due
to d
efer
ral w
ould
be
at a
pplic
ants
ex
pens
e.
• Pat
ent T
erm
Adj
ustm
ent (
PTA
) und
er §
154(
b)
wou
ld b
e w
ithhe
ld d
urin
g th
e de
ferr
al p
erio
d.
¾B
ecau
se n
on-d
efer
red
appl
icat
ions
wou
ld
rece
ive
earli
er o
ffic
e ac
tions
, and
bec
ause
de
ferr
ed a
pplic
atio
ns w
ould
rece
ive
no P
TA
cred
it, o
vera
ll PT
A c
redi
t gra
nts b
y th
e PT
O
wou
ld b
e re
duce
d su
bsta
ntia
lly.
The
Publ
ic
bene
fits f
rom
the
shor
ter “
mon
opol
y.”
© R
on K
atzn
elso
n 28
Furt
her
cons
ider
atio
ns fo
r E
OR
impl
emen
tatio
n¾
Sea
rch
Rep
ort (
SR)
•Th
e va
lue
of h
avin
g an
SR
prio
r to
the
dead
line
for r
eque
stin
g ex
amin
atio
n sh
ould
no
t be
unde
rest
imat
ed. T
here
are
, how
ever
, ine
ffic
ienc
ies i
n se
para
ting
the
sear
ch
and
exam
inat
ion
func
tions
. Fo
r exi
stin
g ba
cklo
g ap
plic
atio
ns, a
refu
nd o
f sea
rch
fees
wou
ld b
e a
subs
tant
ial U
SPTO
cas
h (b
ut n
ot re
cogn
ized
reve
nue)
liab
ility
. •
Thes
e di
sadv
anta
ges m
ust b
e ba
lanc
ed a
gain
st th
e fo
llow
ing
adva
ntag
es:
•Th
e ad
ditio
nal C
laim
Dro
pout
wor
kloa
d be
nefit
s wou
ld li
kely
exc
eed
the
sear
ch
inef
ficie
ncy
loss
(See
Slid
e 11
). •
Enab
ling
a 3rd
par
ty re
ques
ter t
o su
bmit
an S
R in
lieu
of p
ayin
g th
e se
arch
fee,
as
codi
fied
in 3
5 U
.S.C
. § 4
1(d)
(1)(
D),
wou
ld p
rovi
de th
e U
SPTO
a su
bsta
ntia
l qua
lity
boos
t bec
ause
3rd
par
ty re
ques
ters
are
ver
y fa
mili
ar w
ith th
e su
bjec
t mat
ter a
nd h
ave
a ve
sted
inte
rest
in c
ondu
ctin
g m
ost r
elev
ant s
earc
hes.
It w
ill re
invi
gora
te U
SPTO
’s
slog
an “
Qua
lity
is a
Sha
red
Res
pons
ibili
ty”.
¾
Exi
stin
g A
pplic
atio
n B
ackl
og
•Fo
r the
se p
endi
ng a
pplic
atio
ns, t
his E
OR
pro
posa
l doe
s not
cal
l for
refu
nd o
f the
Se
arch
Fee
, Exc
ess C
laim
s or P
ages
fees
, the
reby
redu
cing
ince
ntiv
es fo
r def
erra
l bu
t pro
vidi
ng a
n im
porta
nt v
ery
low
ent
ry c
osts
for p
artic
ipat
ing
3rd p
artie
s who
m
ay h
ave
man
y ap
plic
atio
ns to
dea
l with
. Is
ther
e a
bette
r way
to re
solv
e th
e te
nsio
n be
twee
n U
SPTO
’s fi
nanc
ial r
efun
d lia
bilit
y an
d de
ferr
al g
ains
? ¾
Per
mit
Def
erra
l of a
pplic
atio
ns fi
led
via
the
PCT?
© R
on K
atzn
elso
n 29
Con
clus
ion
¾EO
R p
rovi
des s
ubst
antia
l adv
anta
ges t
o
appl
ican
ts, t
he U
SPTO
and
the
publ
ic
¾A
n ef
fect
ive
EOR
can
be
impl
emen
ted
unde
r cu
rren
t law
¾
Furth
er e
xam
inat
ion
of th
e fe
atur
es o
f a w
ell
craf
ted
EOR
syst
em sh
ows t
hat P
ublic
Not
ice
conc
erns
of s
uch
EOR
are
unf
ound
ed
¾Th
e U
SPTO
shou
ld p
roce
ed w
ith a
form
al
Not
ice
Of I
nqui
ry o
n EO
R im
plem
enta
tion
as
soon
as p
ossi
ble
© R
on K
atzn
elso
n 30
DRAFT May 29, 2009.
A Simple Workload Savings Model for Patent Offices Employing Examination On Request (Deferred Examination) Systems
Ron D. Katznelson
I. INTRODUCTION
The purpose of an Examination On Request (EOR) system, or as it is often called, Deferred Examination system, is to allow patent applicants a set period between filing and examination, during which they would be allowed to withdraw with little or no penalty, or request examination of an updated claim set. The approach provides an opportunity for “second thoughts” that would eliminate commercially useless or marginal inventions by early voluntary withdrawal, either because the applicant loses interest in patenting the subject matter or because the search report casts doubt on the likelihood of patent grant. This would allow examiners to concentrate on those cases where the applicant had a serious commercial interest in obtaining a grant, as shown by their willingness to pay the fees for substantive examination. Substantial examination workload savings can be therefore realized by the patent office.
The historic, legal and public-policy aspects of EOR are beyond the scope of this paper, as it focuses only on a quantitative assessment of the potential workload savings for the U.S. Patent and Trademark Office (USPTO), should it adopt an EOR system. Section II introduces a model of patent applications and claims dropout during a deferral period. It describes empirical data from foreign patent offices employing EOR and introduces key parameters used in the model. These parameters are the commercial survival probability, the claim protection failure probability and the probability density of the total number of claims in applications filed in a given patent office. A simple mathematical model describing the relative number of applications that are ultimately examined (Examination Rate) is derived and compared with empirical data. Finally, Section III applies the model developed in Section II to the USPTO. The numerical values of the model’s parameters are evaluated based on available supportive data and are subsequently used to estimate the potential workload savings at the USPTO, as projected by the model.
II. THE MODEL
Consider the ensemble A of all patent applications filed in a given national patent office during a given year and for which a request for examination was subsequently due. Each member of the ensemble A contains a description of an invention pertaining to certain technology subject matter and one or more patent claims covering the specific features of that invention.
By the time the examination deferral period ends, say t months after the application date, an applicant must decide whether to invest further in an application to maintain it and proceed with its examination. During time t, the applicant may discover either that the technology is not sufficiently valuable (better alternatives or flaws in the technology are discovered), or the patentability of the invention comes into question. Note that in most cases, this information is unavailable (either unavailable at practical cost, or totally unknowable even at infinite cost) at the time that an applicant must make a filing decision to avoid statutory bars. Further, because of the “use it or lose it” asymmetry of patent statutes, applicants must file applications that later turn out to be commercially
1
DRAFT May 29, 2009.
non-viable. This asymmetry biases the future outcomes: information that comes in over time t will predominantly be information that tips the balance toward abandonment. Because there are costs for maintaining and continuing the prosecution of the application, some proportion of applicants will decide to abandon their applications if forced to make an invest-or-abandon choice in light of the information that develops. Alternatively, for applications that have estimated value that exceeds the estimated cost of further prosecution, the applicant will file a request for examination no later than time t.1
For simplicity, this model considers two factors to be required for a given application to move forward to examination. The first, which we will call commercial survival, is that its specific technological solution path and subject matter at the time the application was made, be regarded by its owner as non-obsolete, worth protecting and therefore having “survived” as commercially viable after time t. The second factor is claim protection, i.e., the ability of the specification to support the claims made and the ability of such claims to reasonably protect a commercial exploitation of the survivable subject matter. Several empirical studies have confirmed what one would expect from ordinary common sense, that the number of claims correlates well with applicants’ perception of commercial value and importance, and that when asserted, applications with more claims are more likely to have claims that are valid and infringed. We therefore use the number of claims in an application as a proxy for the strength of claim protection, although as discussed below, it may be difficult to truly separate indicators of the number of claims in an application from indicators on the maturity and commercial survivability of the subject matter.
In any event, like most other economic decision-making, both of these factors are subjective in nature. The tipping-point in applicants’ decisions depends on their perceptions of relative costs and benefits for continuing and requesting examination on an application. Of course, these tipping points may differ for various national patent systems even for the same inventions.
A. Dependence on Applications’ Commercial Survival
As to the first factor, we denote by St the event, and by ps(t) its probability over the ensemble A , that an application would have “survived” the commercial subject matter criteria by the examination request time t. It is evident that ps(t) is a declining function of t – the Maximum Examination Deferral period, making the event St less likely for longer deferral periods t. For example, the move at the JPO from seven-year deferral to three-year deferral remarkably increased application survival rate, because less of the abandonment-favoring information becomes available before the expiration of the shorter time t, as shown in Figure 4 below.
The timing of the patent application with respect to the invention’s development timeline appears to be of greater effect on the amount of commercially-surviving subject matter in an application. Applications filed as priority applications immediately after the invention’s conception under a First-To-File (FTF) patent priority systems are more likely to be abandoned before the examination
See e.g. H.C. Peterson and W.C. Lewis, Managerial Economics, pp. 511-512, 4th Edition, Prentice Hall PTR, (1998) (Generally accepted economic theory holds that individuals and companies will invest in an asset only when the perceived value of the expected economic benefits secured by the asset exceeds the anticipated investment required to obtain or maintain the asset, or both to obtain and maintain the asset, taking into account appropriate risk factors, anticipated rates of return, and related considerations).
2
1
25.7%
31.0%
8.0%
6.6%
11.1%
2.6%
1.5%
1.2%
0%
10%
20%
30%
40%
50%
60%
70%
After First Action After ER, Before First Action After SR, before Exam. Req. (ER) Before Search Report ("SR")
DRAFT May 29, 2009.
request time t than mature applications. Mature applications are those filed not immediately following the conception, but after the inventor took more time for reduction to practice, developing optimal embodiments and best modes of practicing the invention. Mature applications also have the benefit of experimentation and extensive review of the prior art. They are also those filed as second filings, claiming priority to an earlier filing date, such as provided by the Paris Convention or under divisional and continuation laws. For the most part, these latter applications are filed after vetting the content of the application and are therefore generally more likely to survive after deferral.
Fraction of EPO Patent Application Abandonment StagesApplicationsAbandoned (Euro-Direct filings in 1997-1999)
EPO 1st Filings EPO - 2nd Filings (Claiming Earlier Priority) Priority is determined by filing date Priority is NOT determined by filing date
Figure 1. Applications with priority dependant on filing date (under First-To-File law) are less mature and are more likely to be abandoned. Data Sources: EPO Data from G. Lazaridis et al. World Patent Information, 29, pp. 317-326, (2007). “First Action” herein means the withdrawal components in column (4) as defined in the heading of Table 2 of Lazaridis et al.
Evidently, there are stark differences under deferred examination systems, between abandonment rates of applications filed under FTF priority constraints and those of secondary filings (i.e., more mature applications). This can be seen for applications filed directly at the European Patent Office (EPO) shown in Figure 1. This figure shows that 58% of EPO-Direct applications filed under the FTF haste to establish a priority date are withdrawn prior to examination compared to only 12% of secondary filings. This rate is even slightly lower for applications entering the EPO via the PCT route (not shown), as those applications undergo an additional selection of maturity and value. Thus, the commercial survival probability ps(t) over the ensemble A of all applications in a given national patent office will depend on the relative share of applications that are FTF priority applications. We define the “FTF Ratio” of a given patent office as its share of first-filings under FTF priority in the total application pool. Applications that claim priority to an earlier date than their filing date are not FTF priority applications. For example, because applications at the Japan Patent Office (JPO) are dominated by domestic applicants filing under Japan’s first-to-file regime, the FTF Ratio at the JPO has ranged between 0.91 to 0.87, whereas the FTF Ratio at the European Patent Office (EPO) has been far lower, only 0.09.2 The FTF Ratio at the USPTO is zero because under the First-To-Invent
The estimates for the JPO were based on the relative share of domestic applications reported in the statistical appendices of the JPO Annual Reports. The EPO estimate was obtained based on the following: Lazaridis reports that during the years 1997-1999, there were 22,271 direct first-filings and 121,826 direct second-filings at the EPO. The
3
2
DRAFT May 29, 2009.
with-strong-grace-period law, the actual legal priority date of an application is earlier than its filing date and because U.S. applicants do not “race to the patent office” shortly after the conception of an invention to establish a legal priority date. Moreover, foreign applicants claim earlier priority to their foreign application and thus file only more mature applications in the U.S. Thus, the commercial survival probability ps(t) for the USPTO is expected to be higher than that of the EPO or the JPO.
Another element affecting the commercial survival probability ps(t) is the applicant’s perception of the residual value of keeping the examination option open compared to the prosecution costs. Thus, abandonment decisions depend on fee amounts, which can easily change ps(t) by several percent. This calculus is present even after requesting examination and applicants may opt-out prior to first office action at a rate that depends on available fee refund amount. This can be seen in Figure 2 for applications pending first office action at the JPO. It should be noted, however, that after-examination-request withdrawal inducements are far less efficacious than those available before a request. This is because applications subject to examination after a deferral period survived for a reason, and the facts are unlikely to change enough to warrant abandonment only months later.
Application Withdrawn at the JPO Prior to First Action Fraction of After Having Requested Examination Applications Pending FirstAction that were Withdrawn
3%
2%
1%
0%
Source: JPO Annual Reports, www.jpo.go.jp/index_e/reference_room.html Examination Fee RefundSubject to 100% RefundSubject to 50% RefundSubject to No Refund
1998 1999 2000 2001 2002 2003 2004 2005 2006 2007
YEAR
Figure 2. Increases in withdrawals prior to examination due to examination fee refund policies at the JPO. Effective October 2003, the JPO instituted application withdrawal incentives by refunding 50% of the examination fees for withdrawals prior to first action. As of ‘03, relevant JPO fees in ‘03 U.S. dollars were: Examination fees: $1,583; Excess claim fees based on average number of claims in ‘03 at the JPO: $263. The refund amount was raised to 100% of the examination fees in August 2006. Source: JPO.
EPO Annual Reports for this period show that these applications, including all indirect filings during the period totaled 244,350 applications. Thus, the EPO FTF ratio is 22,271/244,350 = 0.091. This low FTF Ratio should not be confused with that of European national patent offices, which typically receive a lot more first-filings FTF priority applications.
4
DRAFT May 29, 2009.
B. Dependence on the Number of Claims in Patent Applications
Data suggests that applications having large number of claims are less likely to be abandoned. Thus, as to the second factor required for an application to move forward for examination (claim protection), we use the number of claims in an application as an empirical attribute in the model. Let c denote the integer-valued random variable consisting of the total number of claims in an application drawn at random from the ensemble A . For simplicity and for reasons discussed below, we do not make a distinction between independent claims and dependent claims in this total count. Let f(k) be the probability density function defined over the ensemble A for the total number of claims in an application. { k} ( ) is the probability that c, the total number of claims in an P c f k application selected at random, is equal to k. We therefore have
f f
( ) 1, and ¢ ² c = E{ } k f k , (1)¦ f k c ¦ ( ) k 0 k 0
where E{Ɣ} denotes the statistical expectation over the ensemble A and where ¢c² is accordingly the average number of claims in applications of the ensemble.
It is not sufficient in our model for an application to have survived the commercial survival criteria; an applicant will only request examination if that commercial value aligns with claim protection available in the application. We denote by p the probability that a claim fails to provide the necessary claim protection criteria for the survivable subject matter to which the claim is directed, as perceived by the applicant and that such failure is statistically independent of other claims.3 We also assume that p is identical for all the claims in the application, regardless of their type. Of course, these assumptions are likely incorrect but we make them nonetheless in order to provide a simple and tractable analytic solution intended to provide only a qualitative functional insight. Thus, given an application’s commercial survival (i.e. given the event St ), an application having k claims must have had all k claims fail the claim protection criteria for it to be abandoned prior to a request for examination at t. Accordingly, the pre-examination abandonment probability conditioned on its commercial survival and on its number of claims k, is given by
kP^Abandonment S ;c k` p . (2)t
Because p <1, the conditional application abandonment probability declines exponentially with k, the number of claims in the application.
Although we “explain” above the functional dependence of the conditional abandonment probability on k through a model of statistical claim coverage power, the actual cause of this functional dependence need not be determined or resolved here. This is because our model can absorb several (perhaps more plausible) causes for such decline in probability with number of claims. For example, it may well be that the reason for higher survivability of an application is not because it has more claims but because its subject matter is more valuable, which is the cause for its large number of claims. Applicants tend to invest more in applications they perceive to be more valuable, and one of the key targets for that investment is crafting and submitting more claims. Absent any detailed
3 Note that this assumption actually pertains to the state of the disclosure as supportive of the number of claims because it is assumed that upon a request for examination in an EOR system, an applicant can replace or amend claims prior to examination, provided those are supported by the original disclosure. As we shall see below, the mere fact that the originally filed claims fail with probability p means that first-action would not be wasted on these failed claims.
5
DRAFT May 29, 2009.
causal knowledge of the economic parameters, we adopt a simplified, but empirically reasonable model, that the probability of abandonment decreases exponentially with increasing number of claims, as shown in (2). In this case, p is regarded as a parameter (that only “looks like” probability) to be determined empirically by fitting the analytical results to the available data.
Based on this exponential decline model, accounting for all applications within the full spectrum of claim count4 yields the following conditional abandonment probability over the ensemble A :
f f
P^Abandonment t ̀ ¦ p f k ( ) ¦ p f k , (3)S k k ( ) k 1 k 0
With substitution of the variable p, one can readily recognize the expression on the right-hand side of (3) as that of the statistical moment generating function5 associated with the probability density function f(k). It is given by
f
)c ^ ` ¦ f k ( )u E euc euk ( ) . (4) k 0
Using this notation and substituting u above by setting eu p , we obtain from (3) the following: f
kP^Abandonment St ̀ ¦ p f k ( ) )c �ln( ) p � (5)k 0 .
Given the application’s commercial survival, the probability that it will not be abandoned (hence, examined) is simply one minus the expression in (5). Using this fact and the relation between conditional probabilities and unconditional probabilities, we obtain the unconditional examination probability across the ensemble of all applications in the given national patent office:
P^Examination ̀ P^Examination S `P{ } ª1 ^S �P Abandonment St ̀ º¼ ps ¬1�)c �ln( ) �º¼ ps . (6)ª pt t ¬
For brevity, we have dropped the notational dependence on t from ps(t). It should be understood, however, that longer deferral periods t would generally reduce this probability value. The benefit of the simple expression in (6) is that it admits analytical solutions in cases where the moment generating function )c of the claim probability density is known and can be expressed in terms of the mean value ¢c² . For purposes of obtaining our estimates, it would be beneficial to adopt a known density function that provides a reasonable approximation fit to f(k) on the one hand, and on the other hand, has an analytically tractable moment generating function that can be expressed in terms of the average number of claims ¢ ² .c This we do in the next section.
C. Analytical Approximation of Probability Distributions for Claims in Patent Applications
Figure 3 shows empirical probability density functions for patent applications filed at the European Patent Office (“EPO”) during two eras as reported by EPO researchers.6 This author is unaware of any published results on claim number distributions in applications filed at other patent offices that
4 For formal convenience in extending the sum in (3) to include k = 0, we take f(0) = 0, i.e. that an application subject to examination contains at least one claim.
5 David Stirzaker, Elementary probability, Cambridge University Press, 2nd Edition (2003), p. 245.6 E. Archontopoulos, D. Guellec, N. Stevnsborg, B. Van Pottelsberghe de la Potterie and N. Van Zeebroeck, When
small is beautiful: Measuring the evolution and consequences of the voluminosity of patent applications at the EPO, Information Economics And Policy, 19(2), pp. 103-132, (June 2007) (See Figure 3 at 122).
6
DRAFT May 29, 2009.
employ EOR systems, although the mean value at the Japanese Patent Office (“JPO”) over the years have been available.7 As to the EPO researchers’ data, the pronounced ‘spikes’ at 10 and 20 total claims are due to the step-up of application claim fees at the EPO above 10 total claims and at the USPTO above 20 total claims respectively. Because the claim fee effects in U.S. claim populations produces a pronounced step at the 20-claim mark,8 and because many U.S. priority applications are filed unchanged at the EPO, the U.S. claim “spike” affects the EPO statistics. This is particularly pronounced in more recent years as shown in the 2000-2004 distribution, when U.S. priority-based application share at the EPO was higher than 20 years earlier.
A negative binomial probability density having two independent parameters that control its functional shape was found to provide an adequate match for our approximation purposes, particularly at the lower claim count region below the 10-claim spike.9 The shapes of such density functions are superimposed and labeled in Figure 3.
Frac
tion
of A
pplic
atio
ns
0.14
0.12
0.1
0.08
0.06
0.04
0.02
0
20020000--0044
1980-84
Negative Binomial Probability Densities(Unscaled)
EPO Claim data Source: E. Archontopoulos et al., Information Economics And Policy, 19(2), pp. 103-132, (June 2007)
0 10 20 30 40 50 + Total claims at filing
Figure 3. Probability density function of the total number of claims in European patent applications filed during 1980 - 1984 and 2000 - 2004. Each such function is matched with similarly shaped negative binomial probability density, shown with absolute probability values that are not to scale.
7 Ron D. Katznelson. "Defects In The Economic Impact Analysis Provided By The USPTO For Its New Claims and Continuation Rules. Appendix E to Amicus Curiae Brief" Tafas/GSK v. Dudas et al., U.S. District Court, Eastern District of Virginia. Case No. 1:07-cv-846, (hereinafter referred to as Tafas v. USPTO (2007)), Doc. 258-5, (January 10, 2008), at 13, available at http://works.bepress.com/rkatznelson/16/.
8 See Katznelson (2008), note 7, at 15, Figure 3. See also slide 23 in: Ron D. Katznelson. "The Perfect Storm of Patent Reform?" Fenwick & West Lecture Series Inaugural Symposium, UC Davis School of Law, Davis, CA. (Nov. 7, 2008). Available at http://works.bepress.com/rkatznelson/54/, (showing a diagonal high probability ‘ridge’ corresponding to a total of 20 claims below and parallel to the 25 “total claim limit” line in U.S. applications).
9 Several well-known discrete (integer) valued probability densities were considered for modeling these empirical distribution excluding the 10 and 20 claim ‘spikes’. Both the Poisson and the binomial distributions were found to provide inferior fit to the data, as they permit fewer degrees of freedom for fitting to the empirical densities.
7
DRAFT May 29, 2009.
The negative binomial probability density function is given by:10
(k r) r k k r 1· r k r ,H k
! ( ) k r integer ¨© k ¸¹
k 0,1,2,... . , (7)f ( ) * �H (1 � H) § � �
H (1 � H) ; for 0 � H � 1 andk *
* k integers k. For a random integer c distributed in accordance with the negative binomial density in (7), the mean value and the moment generating function are given respectively by:
where ( ) is the Gamma function that coincides with the factorial function (k-1)! at positive
rª º
¢ ² { } r 1� H ; and ) ( ) E^ `euc «H
. (8)c E c u » H c � � u «�1 (1 H) e � »¬ ¼
Because of the exponential decline in the magnitude of the weighting term pk in (5), the contributions of the f(k) terms for k in the order of 10 or more are negligible for relevant values of p. Thus, the spikes and other deviations of the model distribution from the actual distribution at large values of k have negligible effects on our model and the parameters were selected for best approximation of the distribution at lower k values.
This author found that, although the distribution in (7) is endowed with two independent parameters, it could be well matched to empirical claim distributions having different means (such as those in Figure 3) by using a single variable parameter, provided that r is selected to have a fixed value11 of r = 4. We use this parameter value in numerical calculations throughout the remainder of this paper.
D. Closed-Form Solution for Application Examination Rates
Under the assumptions made in the previous section above, ¢ ² is the only remaining free variable cupon which the probability density depends. Therefore, the expression on the left-hand side of (8) above can be used to substitute H in terms of ¢ ² into the expression for the moment generating c
10 See Stirzaker (2003), note 5. at 245; See also http://en.wikipedia.org/wiki/Negative_binomial_distribution. 11 For the special case where r is an integer, the negative binomial distribution is known as the Pascal distribution. It
is the probability distribution of a certain number of failures and successes in a series of independent and identically distributed Bernoulli trials. For k+r Bernoulli trials with success probability H, the negative binomial provides the probability of k failures and r successes, with success on the last trial. It is thus intuitively tempting to envision r as a latent variable characterizing the number of inventive concepts that patentees perceive as having successfully covered in an application, by drafting k claims. They do not “stop adding more claims” in the application so long as they perceive a “failure” to adequately cover these r latent inventive concepts. According to this heuristic view, the last instance for which applicants perceive coverage “failure” during this process is when claim k is included, at which point they stop adding more claims. Under this intuitive framework, the ability to precisely fit claim distributions using the same value of r = 4 over an historical span of decades, is remarkable. This suggests that the historical changes in the distributions and the growth in the average number of claims ¢ ² are merely due to the reduced “success” probability H that a claimcadequately covers one inventive concept. Under this intuitive framework, H can be thought of as a measure of claim breadth and that historically, claim scope had diminished over time, necessitating more claims. This notion is consistent with other evidence of historical claim-scope erosion described by this author in: “Patent Continuations, Product Lifecycle Contraction and the Patent Scope Erosion – A New Insight Into Patenting Trends”, Southern California Law Associations Intellectual Property Spring Seminar, Laguna Niguel, CA, (June 8 - 10, 2007), available at http://works.bepress.com/rkatznelson/3/.
8
DRAFT May 29, 2009.
function on the right hand side of (8). Accordingly, since we have set eu p , we rewrite (6) and arrive at the probability of examination as follows:
ª ª P{Examination}
§¨
©¨ 1� «
¬�1 (1 � �HH) p�
º»¼
r ·¸
¹¸
ps
§¨
©¨ 1� «
¬
r º»
r ·¸ ps (9)
c �1� p� � r ¼ ¸¹
The model in (9) provides an intuitive insight into the examination rate data obtained from national patent offices. The examination probability in applications at the respective office is an increasing function of ¢c² with values that asymptotically approach ps with large average number of claims. Three curves based on (9) are superimposed with broken lines on plots of actual data from the EPO and the JPO. Using the fixed value of r = 4 discussed above, the two available parameters p and ps were used for fitting the function in (9) to the data in Figure 4. Unfortunately, only the JPO 7 - year deferral data had sufficient spread to meaningfully constrain unique possible values of p and ps, yielding the empirical fit values of p = 0.36 and ps = 0.6. For the other two curves (JPO 3-year deferral and EPO variable deferral) the value of p = 0.36 was kept, resulting in ps values of 0.7 and 0.92 for the JPO and the EPO respectively. The latter are shown as the asymptotic values of examination rates on the right hand side of Figure 4.
Examination Rate at National Patent Offices ps
100%
'85 '89'90 '92
'98 '99 '94
'86'87'88 '91 '93 '95 '96'97
'01
'02 '04
'03
'88 '89
'90
'91
'93 '92
'94 '95 '96'97'98 '99
Sources: Examination Rate: JPO Annual Reports; G. Lazaridis et al. World Patent Information 29, pp. 317-326, (2007). Average Claims at filing: R.D. Katznelson, Federal Circuit Bar Journal, 17(1), pp.1-30, (2007).
EPO - 19 to 24 Months Deferral, FTF Ratios: 0.09
JPO - 7 Years Deferral, FTF Ratios: 0.91 - 0.89
JPO - 3 Years Deferral, FTF Ratios: 0.88 - 0.87
p 92% 90%
80%
70% 70%
60% 60%
50%
40%
30%
20%
10%
0%
App
licat
ion
Exam
inat
ion
Rat
e
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Average Number of Claims in Applications Filed on Year and at Patent Office Shown
Figure 4. Examination rates (Probability of Examination) at the EPO and the JPO versus the average number of claims filed at the respective patent office in the years shown by labels for each point. The deferral period at the JPO has changed from seven years to three years starting for applications filed after October 2001, shown in a separate cluster. The FTF Ratio ranges shown for the respective patent offices correspond to the estimated fraction of applications in that office that arrive as First-To-File priorities. For the EPO, the FTF ratio is based on the fraction of EPO first-filings in ’97-’99 as reported by Lazaridis. The JPO FTF Ratio was estimated by the ratio of resident filings to all filings as given in the JPO annual reports.
9
DRAFT May 29, 2009.
E. Examination Workload Savings in EOR Systems
Whereas workload savings is related to the fraction of applications abandoned prior to examination under an EOR system, accounting for the actual savings is a bit more complex than just counting dropped applications. For simplicity, we use the total number of claims in applications examined as a measure of examination workload. For simplicity, we also assume that examiner workload for searching and issuing a First Action On Merit (FAOM) is a fixed fraction a < 1 of the total work required for an application disposal. Of all N applications in the ensemble, there are two categories of applications that give rise to workload savings in EOR systems. The first category contains the applications that do not survive the commercial survival criterion. Under the model, this occurs with probability 1 – ps, irrespective of the number of claims and thus the number of claims dropped prior to first action in this category is given by
f
(1� ps ) ¦k f k N ( ) (1� p )¢ ²s c N (10) k 0
The second category is the category of applications having commercially-surviving subject matter but having claims that are dropped as useless due to failure to provide claim protection. Because we assume that each claim in such applications will be useless with probability p, independently of all other claims, the number of claims dropped prior to first action in this second category is given by
f
p p¦k f k N s ( ) p p s ¢ ² c N (11) k 0
Although the applications in this category may survive (except those dropped applications for which all k claims are useless), upon a request for examination, the dropped claims will instead be replaced or amended for presentation on first-action. Thus, the examination workload savings associated with both of these categories of dropped claims is that no first-action is taken on these claims. Therefore the FAOM savings are given by weighing the sum of the terms in (10) and (11) by the FAOM workload fraction a:
a(1� p )¢ ² N � a p p ¢ ² N a¢ ² N >1� p (1� p)@FAOM Savings s c s c c s (12)
As to examination savings under EOR beyond first-action work, the FAOM savings accounted for above with respect to the first category in (10) are the only savings expected for this category. This is because there would be no secondary office actions on applications that do not survive the commercial survival criterion even under an automatic examination system as applied at the USPTO today. In this latter case, these applications would be dropped after FAOM anyway.
In contrast, there would be claim examination work savings beyond FAOM for some applications in the second category in (11). This is because, unlike automatic examination systems that otherwise sustain full prosecution of the claims (or substitute claims) in an application with commercially-surviving subject matter, none of the claims in abandoned applications of this category would be presented under EOR. To obtain the total number of claims in abandoned applications that are in this otherwise commercially-surviving second category, we must use the probability density of the number of claims in abandoned applications. That is, we look for the conditional probability of having k claims in a survivable-subject-matter application given that it is abandoned under the model:
10
DRAFT May 29, 2009.
f k St ;Abandonment � P^c k St ;Abandonment ̀� kP^Abandonment St ; c k`
P^c k p , (13)f k
St ;̀ St ̀ )c �ln( ) �
( )pP^Abandonment
We have used Bayes’ theorem for conditional probabilities in the transition to the second line in the equation above, after which the identities in (2) and (5) were substituted in the numerator and denominator respectively. The resulting probability density in (13) for the number of claims in abandoned applications is a weighted version of the ensemble distribution f(k), with relatively more applications having fewer number of claims k, as they are more likely to be abandoned in accordance with our model. To obtain the secondary Office Action (OA) savings based on the total number of claims dropped in these abandoned applications, we note that the total number of such abandoned applications is P^Abandonment S ` p N )c �ln( ) �p p N , across which, the number of claims are t s s
distributed according to (13). We therefore sum the number of dropped claims over all values of k and multiply by 1 – a, the relative remaining examination work after FOAM, to obtain the workload savings in secondary OAs:
f
� )c �ln( ) � ¦ �Secondary OA Savings = (1 a) p p N k f k St ;Abandonment �s k 0
f f k w k w (14)(1� a p N ( ) a p N p f k a p N �) s ¦ k p f k (1 � ) s ¦ p � � ( ) (1� ) s p ¬ª)c ln( ) p �º¼ ,
k 0 k 0 wp wp ' ' ) s )c � p � where )c is the derivative of )c.(1 � a p N ln( ) ,
The last term in the first line of (14) was obtained by reversing the order of differentiation and summation and recognizing that the resulting sum is the moment generating function given in (5).
For the negative binomial distribution case, we use the moment generating function in (8), substituting H as we have for (9) and using it in (14) to obtain
r r�1 w ª r º ª r º
a p N » ¢ ² Secondary OA Savings = (1 � ) s p « =(1 � a p N) s p c « » , (15)wp r c (1� p) r � ¢ ² (1 � p)¬ � ¢ ² ¼ ¬ c ¼
By summing the FAOM savings from (12) and the secondary OA savings from (15) we obtain the total workload savings
ª ºr�1 ½° r °Total Savings ¢ ² N a>1� ps (1 � p)@� (1 � ) s «
¬ r � ¢ ² c (1� p) »¼ ¾ (16)c ® a p p
°̄ °¿
Finally, because the measure for the total examination workload currently under automatic examination system is ¦f ( )kf k N c N , the relative savings due to EOR is given by the
k 0
following fraction: r�1
Relative Savings = Total Savings /¢ ²N a>1� p (1 � p)@� (1 � ) ª
¬ r � ¢ ² c r (1� p)
º
¼ (17)c a p p s s « »
11
DRAFT May 29, 2009.
III. WORKLOAD IMPLICATIONS OF USING EXAMINATION ON REQUEST IN THE USPTO
Should the USPTO adopt an EOR system, a Maximum Deferral period of three years appears most likely to properly balance workload concerns and structural pendency concerns. In evaluating the relative workload savings expected at the USPTO as predicted by (17) above under this assumption, we examine estimated numerical values for the various parameters as follows.
Estimates for the FOAM workload fraction a.
Most of the time spent by examiners on a patent application is expended for purposes of first action. This includes reading the application, the cited prior art and searching other references. It also includes crafting claim rejection text and, in a small minority of cases, application allowance text and reasons for allowance clauses.12 Examination-corps overall USPTO data shows that examiners spend on FAOM an average of 85% of the total time spent on examination.13 Thus, we use a = 0.85 for our estimates herein.
Estimates for the Claim Protection Failure Probability p.
This is the probability that a drafted claim fails to adequately protect the commercially-surviving subject matter to which it was directed, given the available support in the specification. The empirical result found in Figure 4 for the JPO during the ’80’s and ’90’s is p = 0.36. It is difficult to obtain empirical estimates for patent offices having a large average number of claims by the method described in connection with Figure 4. For commercially-surviving subject matter, this probability measure may be estimated instead by examining the number of claims that are initially submitted in applications versus the number of claims that ultimately appear in patents granted from these applications. USPTO data shows that the total number of claims in allowed applications ranges from 80% to 90% of the total number of claims in these applications when filed.14 We therefore estimate for the USPTO the claim protection failure probability to be in the range of p = 0.1 to p = 0.2.
Estimates for the Commercial Survival Probability ps.
This value provides a measure of office-wide application commercial survival probability as a function of Maximum Deferral time in a manner that removes dependence on the average number of claims. Its complement, the subject matter dropout probability 1 - ps, is comprised of two major components: the first is related to application dropout after applicants discovers the art cited in the search report and the second is abandonment after discovery of new facts over time, either before or
12 Latest publicly available data is from FY 2005, wherein allowances and Quayle actions amounted to 16.5% of all FAOMs. See Ron D. Katznelson, Trends in USPTO Office Actions, (2007) at http://works.bepress.com/rkatznelson/57/.
13 U.S. Department of Commerce, Inspector General, USPTO Should Reassess How Examiner Goals, Performance Appraisal Plans, and The Award System Stimulate and Reward Examiner Production. Final Inspection Report No. IPE - 15722, (September 2004), available http://www.oig.doc.gov/oig/reports/2004/USPTO-IPE-15722-09-04.pdf, (See Figure 12. During fiscal years 1999-2002, an average of 18.55 hours were spent by examiners on FAOM, out of an average of 21.83 hours for application examination).
14 USPTO, Average Claims for Applications Filed, Allowed and Issued by FY of Filing, Tafas v. USPTO, (2007), Admin. Rcrd., A04371. (In fiscal years 1999-2002, the average number of claims at filing increased from 19.3 to 23.2 and the number of claims allowed in applications filed during these years increased from 17.4 to 18.5. This corresponds to a claim survival rate of 0.9 to 0.8 respectively. Note, however, that this estimate is conservatively high because original claims may be amended or replaced rather than dropped, implying higher claim protection failure probability).
12
DRAFT May 29, 2009.
after a search report. Abandonment after the search report can be the dominant component, particularly in offices having low FTF Ratio, as seen in the second-filings column of Figure 1. The second component (discovery of new facts over time) is the major factor for the dependence of ps on t, the Maximum Deferral period. So for example, the EPO deferral period for the application years shown in Figure 4 ranged between 19 to 24 months, increasing only in the last few years up to 33 months.15 The EPO asymptote of ps = 0.92 should therefore be associated with maximum deferral period of less than two years. As explained earlier, the relatively low FTF Ratio (0.09) at the EPO explains the relatively low commercial survival dropout rate as compared to the JPO. Because the FTF Ratio at the USPTO is even lower (zero), a comparable deferral period at the USPTO would likely result in higher survival probability than that at the EPO. However, because we assume a three-year maximal deferral, the survival probability is harder to pin down. Thus, obtaining bounds from other evidence and estimating the expected decline rate per year of ps is desirable.
USPTO FIRST ACTION STATISTICS Average First Action % Abandoned After PendencyFAOM (Months) 16% 30
15% 25
14% 20
13% 15
12% 10
11% 5
10% 0
Pend. = 1.90 t - 3787.35
Aband. = 0.0035 t - 6.8624
2000 2001 2002 2003 2004 2005 2006 2007 2008 Fiscal Year
FAOM Average Pendency (From USPTO Annual Reports)
% Abandoned After FAOM (From Tafas v. USPTO , 2007, Admin. Rcrd., A07170)
Figure 5. Abandonment after First Action On Merit (FAOM) and average FAOM pendency at the USPTO. Note that longer FAOM average pendency corresponds to increased application abandonment after FAOM. See text.
For that purpose, Figure 5 shows relevant FAOM statistics at the USPTO where application survival after first action is highly related to ps. This is because under current USPTO practice, there is no mechanism or incentive for applicants to drop out prior to examination. Express abandonment prior to examination entails no upside to the applicant, as no refund is received, and the tangible downside is absorbing the expense of tracking, writing, sending and docketing an abandonment request letter. Therefore, application survival rate after FAOM can serve as a lower bound estimate for ps in the USPTO, if appropriately adjusted for the longer deferral time compared to FAOM pendency.
Examination deferral in the EPO terminates 6 months after the publication of the search report. According to the Trilateral Patent Statistics Report (http://www.trilateral.net/statistics/tsr.html), EPO search reports have been issued after average pendencies that increased from 14.7 months for 1995 search reports to 17.2 months for 1998 and to 27.3 months for such reports issued in 2001. Thus, taking into account the reports’ delay spread, during the application years for which the asymptote is a good fit in Figure 4 , deferral was estimated to range from 13+6 months to 18+6 months. The lower examination rates seen in the last few application years ending in 1999, for which average deferral is estimated to have been 27.3+6 = 33.3 months based on 2001 issued report pendencies, is therefore not surprising.
13
15
DRAFT May 29, 2009.
A FAOM can be regarded as delivering a search report and an office action on the merits. As such, abandonment after FAOM includes a dropout component that is not contained in 1 - ps. However, the information imparted to the applicant in this component and in the search report is not a function of time and thus cannot cause the increase of dropout rate with FAOM pendency as seen in Figure 5. We therefore conclude that the variation is due only to a reduction in ps. Based on the increase in abandonment rate after-FAOM in 2001-2004, during which average FAOM pendency increased by 5.7 months, we estimate that within this parameter range, the decrement in ps was about 2.2% per pendency year.16 This rate is remarkably similar to that found at the JPO, when it shifted from a seven-year to a three-year deferral rule. As seen in the two related asymptotes in Figure 4, a four-year decrease in deferral period at the JPO, caused ps to increase from 0.6 to 0.7, corresponding to a change of 2.5% per deferral year. With less precision and certainty, one finds in Figure 4 a similar rate of decrement at the EPO during the transient rise in search report pendency (and therefore deferral) of application years 1995 to 1998.
Using the above estimates, extrapolating the abandonment rates in Figure 5 to a FAOM pendency (deferral) of 36 months, results in an abandonment rate of about 17%. Similarly, extrapolating the EPO data of 8% dropout at an average of 21 months deferral to 36 months, results in about 12% dropout. Because in these extrapolations, the 17% FAOM figure contains an additional component (first examination action), and because the 12% EPO figure corresponds to an office with non-zero FTF ratio, we must converge on a lower application dropout rate for the USPTO. We estimate it will be in the range of 8% - 11%, yielding an estimated range of ps = 0.89 to ps = 0.92.
Total Workload Savings Estimates
As explained above, workload savings at the USPTO under EOR will not be limited to abandoned applications. Even upon requesting examination of surviving applications three years after filing, applicants will withdraw failed claims that would otherwise receive a FAOM. In evaluating the total workload savings, we note that for USPTO parameters and the large average number of claims, the rightmost term in (17) is extremely small and is therefore negligible compared to the first two terms. Thus, based on the parameter ranges estimated herein, the relative savings at the USPTO are essentially given by
0.25 as the high-end estimate Relative Savings = a>1� ps (1 � p)@ ®
¯ 0.15 as the low-end estimate (18)
While a portion of these savings would come from processing 8% - 11% fewer applications, other savings would also come from avoiding unnecessary FAOMS in surviving applications, thereby reducing the average actions per disposal.17
16 By using regression analysis over the periods 2001-2004 and using the two slopes thus found, we find that an increase of 5.7 months (1.9 u 3) in average FAOM pendency corresponds to an increase of 1.05% (0.0035 u 3) in abandonment rate after FAOM. Thus, we estimate that in this parameter range, the reduction in ps is about 2.2% per deferral year.
17 USPTO, Actions Per Disposal, Tafas v. USPTO (2007), Admin. Rcrd. A07131. (The average number of actions per disposal in FY 2005 was reported as 2.7). This measure treats actions in Requests for Continued Examination (RCE) as actions in a separate application and therefore it underestimates the number of actions on a serialized patent application. It is anticipated that by saving FAOM on obsolete claims, one would make better use of the two available actions per disposal, reducing the need for further actions and RCEs.
14
DRAFT May 29, 2009.
Search Reports
The workload savings computed above are predicated on applicants having timely information from search reports to induce abandonment when appropriate. Currently, the USPTO does not issue search reports in national applications before its first action. However, deferral by three years will make foreign and local PCT search reports for counterpart applications readily available to applicants and USPTO examiners before the time comes to request examination. Given this international work-sharing opportunity, a small fraction of applications would actually require USPTO search work prior to requests for examination.18 Given that search workload is only a fraction of the FAOM work, to the extent that the USPTO adopts rules for EOR that involve generating early search reports for this minority of applications (as it should), the workload savings may actually be slightly less than that projected in (18).
IV. CONCLUSION
A simple model of workload savings associated with Examination On Request procedure was developed and compared to actual data from foreign patent offices. The model’s parameters were estimated as they pertain to the USPTO. With these estimated parameters, the model projects that if the USPTO were to implement a three-year deferral EOR procedure, there would be between 15% to 25% savings in examination workload and that the dropout rate of applications that will never be examined will be in the range of 8% to 11%.
V. ACKNOWLEDGEMENT
The Author wishes to thank Peter Hingley and George Lazaridis of the European Patent Office for their support in interpreting, breaking down and clarifying the EPO data used in this study.
Many applications would not require a search report. All RCEs (about 25% of application in FY 08) do not require new searches. About half of the applications filed with the USPTO are of foreign origin, mostly having been filed a year earlier abroad. U.S. residents who also file abroad via the PCT (about 55,000 in FY 08) obtain a search report on counterpart applications. All such foreign or domestic PCT search reports should be available by the time a request for examination is filed, leaving less than 25% of applications without early search reports.
15
18