YT YouTube2Text

YouTube2Text → Transcripts → ECON 20110 Fall 2013 - Lecture 12: Intellectual Property

ECON 20110 Fall 2013 - Lecture 12: Intellectual Property — Transcript

by Glen Weyl · 9,814 words · 1,845 segments · language en · Watch on YouTube

Open in transcriber Summarize with AI Download SRT/VTT

Full transcript

0:00okay
0:01so
0:04um
0:05today we're going to be talking about
0:07intellectual property
0:10um
0:11which is the right to produce as a
0:13monopolist for a period of time
0:17and this takes many forms copyrights
0:20uh yeah
0:55okay
0:56um
0:57this includes copyrights and patents
0:59which are the two most
1:01uh well-known formal forms of
1:04intellectual property
1:06as well as trade secrets which firms
1:09might keep
1:10which are an alternative to formal
1:12intellectual property protection secrets
1:14about how to make their product
1:16as well as what i call authenticity
1:18rents which are not formal intellectual
1:20property but where consumers might be
1:22willing to pay more for the original
1:25version of something
1:26uh even though they could get just as
1:28good of a fake version art is a classic
1:30example of that but a lot of fashion is
1:32that way as well
1:35and intellectual property i think is one
1:36of the oldest forms of
1:38uh economic policy
1:41so um i have a quote
1:44that i found with a classicist that's
1:46published on the back cover of one of
1:47the issues of the journal of political
1:49economy the in-house journal here at
1:50chicago
1:52and it um is from
1:55a greek historian saying that in seventh
1:57century bc greeks
1:59they used to offer
2:01um that if you came up with a really
2:03great uh dish you would be the only
2:06person who would have the right to cook
2:07that dish for the next year
2:12and today we're going to learn about the
2:14trade-offs
2:16involved in
2:17intellectual property and we'll relate
2:19it a bit to the punishment of crime
2:22so we'll start by talking about what
2:24monopoly profits incentivize
2:27about the creative destruction
2:31that accompanies dynamic markets and the
2:34role that plays in economic development
2:37will relate this to the static and
2:39dynamic effects of crime
2:41and then talk about the importance of
2:43incentives
2:44to motivate innovation
2:46as well as the harms associated with
2:48intellectual property and that will lead
2:51us to a trade-off between innovation
2:53incentives and the distortions that
2:55accompany
2:57intellectual property we'll then use
2:59that to develop a model of optimal
3:01intellectual property
3:03and go through some comparative statics
3:05of that as well as the lessons that it
3:06teaches for policy
3:08and then briefly discuss some of the
3:09broader perspectives on issues related
3:12to intellectual property other than this
3:14basic uh simple trade-off
3:16okay
3:17so thus far we've viewed monopoly
3:19profits as being just a transfer
3:23that is we thought that the harm of
3:24monopoly was not the profits that the
3:26firm earned but the dead weight loss
3:29and um
3:32uh
3:33while
3:34a redistributive motive might lead us to
3:37oppose these types of
3:39transfers
3:41i think a more important factor in
3:43thinking about the status of the
3:45monopoly profit is the profits is the
3:48encouragement that monopoly profits give
3:51to certain activities now sometimes the
3:53monopoly profit really is just a rent to
3:55whoever ends up collecting it
3:57but
3:59often monopoly profits actually
4:02act as a way to incentivize people to
4:04undertake activities that they wouldn't
4:06undertake in the absence of that rent
4:09and
4:10these activities could be pure rent
4:12seeking
4:13in which case they would be wasteful
4:16like lobbying for protection and in this
4:19case
4:20the monopoly profits are themselves
4:23a social
4:25waste because rather than just being a
4:27rent a transfer to the monopolist they
4:30uh end up getting dissipated people end
4:32up spending enough to try to achieve
4:34them through lobbying uh and if that
4:36lobbying is wasteful then um
4:39then
4:41market power
4:42is is harmful
4:45and that makes monopoly much more
4:46harmful than the standard analysis
4:49because not only the deadweight loss but
4:50also also the profit need to be counted
4:53as waste
4:54on the other hand monopoly
4:56rents could encourage the creation of a
4:59new market and because the monopolist
5:01who creates that new market only gets
5:04the benefit of the profit and not of the
5:06consumer surplus that that new market
5:08also creates
5:10creating those markets will itself have
5:12a entrepreneurial externality
5:15and
5:16that will make the monopoly profits not
5:19just uh a transfer but actually a
5:22benefit because of the consumer surplus
5:24that they bring about yeah so yeah why
5:26is the profits are definitely lost is it
5:28just the money they spend on men seeking
5:31well yeah the point is that the money
5:33that they spend on rent seeking will
5:35generate some profits for them but those
5:38profits won't compensate for all you
5:40know they'll compensate for the uh waste
5:43that they created but that waste is is
5:46social
5:47loss to society so the waste of social
5:50profits otherwise
5:51they would have been but the fact the
5:53thing is that their existence encouraged
5:55this wasteful activity
5:57so i mean that's a more general thing
5:59like if people are getting rewarded for
6:01something that doesn't have much social
6:03value you should be worried because then
6:05they're going to spend a lot of time
6:06trying to get that reward and that's
6:08going to turn that transfer into waste
6:14i think that's
6:15one reason why a lot of people are
6:16concerned about wall street
6:18is that you know all the money that you
6:20can make there if you don't think it's
6:21all doing that much social value is
6:24going to encourage a lot of people to go
6:26into that area who could otherwise be
6:27doing productive activities
6:29um
6:31and this makes monopoly uh far better
6:34than the standard analysis if it's
6:36incentivizing the creation of new
6:38markets
6:39okay
6:40and all this i think can be crucial in
6:43many cases because the
6:45profits that the monopolist creates can
6:47be much larger than dead weight loss
6:49that can be true in many classic
6:51monopoly settings when the
6:54profit square is bigger than the
6:56deadweight loss triangle
6:58but more importantly
7:00when we turn to oligopoly
7:03and there um isn't
7:05uh as much
7:07uh
7:08deadweight loss
7:10because um the price gets lowered and
7:13the deadweight loss is a triangle right
7:15so as the price goes towards the
7:17efficient price the deadweight loss
7:19triangle is going to be very small but
7:20the profit area is still significant
7:22because it's a
7:24rectangle rather than a triangle
7:29so imagine that we were to lower this
7:31price down the deadweight loss would
7:34start to shrink significantly but the
7:37profit area would still uh create a
7:40significant would still be significant
7:43okay
7:44so we've talked before about some of the
7:47ways of obtaining monopoly
7:50and
7:52let me refresh you on some of those so
7:54one is protection by a government
7:57for example licensure
7:59chartered corporations in history like
8:01the east india company was chartered to
8:04be monopoly by the government
8:08and these can be useful to some extent
8:11for for a country
8:13uh if it's exporting or importing from a
8:16foreign country because they actually
8:18want to exploit that other country to
8:22gain as much value as they can from that
8:24other country
8:26and and then putting a monopoly in
8:28charge of it might accomplish that goal
8:32um private or illegal violence is
8:34another way of obtaining monopoly so
8:36mafias control lots of industries
8:38especially illegal industries and also
8:40sort of semi-legal industries like
8:41casinos
8:43and exclusionary
8:45practices like we discussed in the beer
8:48market in peru can be another
8:50way of gaining
8:52monopoly power another form is illegal
8:55cartels and legal industries
8:57and anti-trust laws are designed to
8:59prevent these
9:01um but they didn't really exist in
9:02europe prior to the 1980s and in many
9:05developing countries they
9:08are weak
9:09and even in countries where they exist
9:12uh cartels can also can still form so
9:14for example there's a movie uh with matt
9:17damon that i'm trying to remember what
9:18it's called maybe something like the
9:19insider what gasland
9:23the inform the informant it might be the
9:25informant yeah
9:27and it's about the vitamin cartel uh in
9:30the united
9:35was something about states
9:37was about corn yeah the uh i thought it
9:39was about the vitamin c but it might
9:40have been the corn cartel anyways
9:43it's about some cartel you
9:45you can figure it out maybe online yeah
9:48um mergers are another way of forming uh
9:52monopolies
9:53governments try to review these and to
9:55stop them from forming monopolies but
9:59some countries
10:00don't i mean they don't get everything
10:02and also in some countries there is no
10:05uh
10:06merger policy so for example in peru the
10:08way that the
10:10cartel basically or legal cartel
10:13of beer formed there was that americans
10:17they opened up to investment with the
10:19united states and the united states
10:20companies in particular sab miller came
10:22and bought up all the brands of beer and
10:24peru and they had no merger control so
10:26that was perfectly legal
10:28intellectual property is another uh form
10:32that we'll talk about in just a minute
10:34okay
10:35so
10:36each of these forms of monopoly and the
10:38profits that come along with them
10:39encourage
10:41some uh activity
10:43uh beyond just what whatever distortion
10:46the monopoly creates uh in static terms
10:50and uh mike
10:52uh what um
10:55what are some of the things that were
10:57would be encouraged by each of these
10:59types of
11:00monopoly we talked about before uh you
11:03would like have exclusive access to
11:05markets maybe if you if you could do a
11:07thing like buy out all of your companies
11:11but what would okay so that's the
11:13consequence but the fact that you can
11:15make profits off of that what types of
11:17activities does that encourage in each
11:19of the examples um that we talked about
11:21innovation
11:23to make products
11:25in the case of intellectual property
11:28what about in some of the other cases we
11:29talked about
11:30um
11:32like how about the uh the illegal
11:34markets case
11:41i guess it would encourage things like
11:43sabotage or like private violence
11:44against other companies yeah exactly so
11:48that's uh
11:49that's
11:50um a good example of something
11:53problematic that can be encouraged by
11:56the existence of monopoly profits
12:04protection
12:06by the government
12:08can encourage lobbying which can lead to
12:11you know expenditures maybe wasteful
12:14expenditures on lobbying cozying up to
12:16politicians
12:17though it's not completely obvious that
12:19that's like the most terrible thing to
12:21encourage because giving money to
12:22politicians might allow the
12:25people to pay them less and if you think
12:27politicians are underpaid because and we
12:29don't get like high quality enough
12:31people going into politics
12:32uh that might actually be a good thing
12:34though maybe the types of people who are
12:36willing to accept money that way are not
12:37the types of people we want to be
12:38politicians
12:40and
12:42additionally sometimes companies in
12:44order to like lobby and get support from
12:46the government
12:48don't just want to pay the politicians
12:50but want to get support from the public
12:52and they may do charitable things
12:54as a result of that so for example a t
12:57had a lot
12:58monopoly for a long time in order to
13:00maintain that monopoly
13:03they
13:04um
13:06they did a lot of public relations work
13:09a lot of which was driven through the
13:10research they did at bell labs
13:13and that research i think added a lot a
13:15lot of nobel prizes came out of that so
13:17that might be a important social benefit
13:20that was encouraged by their desire to
13:22maintain their monopoly
13:24private or illegal violence is usually
13:26very harmful
13:28as we just discussed
13:30illegal cartels might enforce might
13:33encourage expenditures on illegally
13:35enforcing them which might be wasteful
13:38but mostly what they encourage is
13:39similar to what a merger would encourage
13:43which could include diseconomies of
13:45scale
13:46because firms might merge in order to
13:48exert their market power even if they
13:50were more efficient operating separately
13:53it might also encourage what's called
13:56entry for buyout i think this was last
13:58year's problem set but the idea is that
14:00firms might enter the industry even
14:03though they don't have a better product
14:05just so that they could get bought out
14:06by existing firms if you make that
14:09illegal if you make mergers between
14:10closely substitutable firms illegal then
14:12that firm never would have entered and
14:14that would have saved the money that it
14:15cost that firm to enter so that's an uh
14:18potential
14:19problem that uh merge allowing mergers
14:21could encourage yeah
14:24if entry for bio is
14:27so easy wouldn't the prospect of
14:30entrance kind of deter the exertion of
14:33too much market power in the first place
14:35yeah but maybe you want market power to
14:37be exerted in the first place
14:39because
14:40um
14:41the firm innovated in order to create a
14:44product right so you actually wanted the
14:46initial firm to have a lot of profits so
14:49that it could reward their innovation so
14:52that's like a double reason why entry
14:54for buyout might be bad it most might be
14:57bad both because it's wasteful for the
14:59price of the entry
15:01but also
15:02because um you want the initial firm to
15:05do better so like a good example of that
15:07is um
15:10so there's this policy that a lot of
15:13drug companies
15:15pay
15:16um
15:17people who have developed competitors to
15:19their drugs to delay entering the market
15:22even after
15:24the
15:27even after the patent has expired
15:29because they want to have extra life on
15:31their patent effectively
15:33and the supreme court ruled that illegal
15:36under the antitrust laws but that might
15:38actually benefit the drug companies
15:40because now there will be less incentive
15:42to develop an alternative generic just
15:44to get paid off by the drug companies
15:48you see what i mean
15:49but if there is some sort
15:52of
15:53side payment that's possible to that
15:55would actually deter the entry it's that
15:58seems it seems that might imply that
16:01it's more socially in efficient for one
16:04company to just produce
16:07uh
16:09because it's the the new entrant needs
16:12to get a signed payment at least this
16:14big
16:15well no that's not true because the
16:17thing is like it might lead to more
16:19profits but that doesn't mean it's more
16:21exposed efficient
16:23that's true
16:26um
16:28yeah
16:29yeah
16:30the information yeah thank you
16:34they add to animal feed okay and
16:36including high fructose corn syrup i see
16:39but is it is it like a vitamin type
16:41additive
16:43what it's an amino acid
16:45but is it like to give micronutrients or
16:48i'm not sure it's nutrients for like
16:50kettle feed yeah yeah so it is some form
16:52of vitamins but it's vitamins and like
16:54you know for cattle not for people yeah
16:57yeah
16:58um
17:00okay
17:01so
17:02uh intellectual uh property which is
17:06what we're going to talk about for the
17:07rest of the
17:10lecture
17:11has a lot of beneficial things that it
17:13encourages and that's what we'll
17:14emphasize but it can also encourage more
17:16costly things
17:18like trying to get a patent just before
17:21somebody else is going to file for the
17:23same thing
17:24just uh to beat them which doesn't do
17:27any
17:28doesn't create any social value it just
17:31um brings the thing slightly forward and
17:33allows you to get the credit
17:35so that that's not such a beneficial
17:37thing that intellectual property
17:38encourages yeah christian is not like
17:40very similar to the entry for buyout
17:42thing like if you get cat controls that
17:44have no
17:45intention of actually producing the
17:47product but just want to be
17:49either one to win a lawsuit or one yeah
17:51i mean i think that uh racing the racing
17:54dynamics is extremely similar to the
17:55entry for bio type dynamic you want to
17:58try to do
17:59you want to discourage copycat type
18:01stuff especially if it's preempting the
18:03other person
18:05um
18:07that's also related to a lot of the
18:09things we talked about in financial
18:10markets that are problematic where i try
18:12to like you know get the information
18:13just a minute sooner than you so that i
18:16can
18:16cash in
18:18um okay so joseph schumpeter emphasized
18:21these dynamics he said that you know the
18:24most important part of capitalism is
18:26innovation and progress and the static
18:28distortions associated with monopoly are
18:30fairly small
18:31whereas the creative
18:33power of capitalism all the new things
18:35that it's invented has really
18:37transformed the world
18:40now
18:41i think you need to be a little bit
18:42careful about this because of course the
18:44long run benefits
18:46uh once you discount them
18:48might not be quite as overwhelmingly
18:50large as it looks but you know for the
18:52most part i think that
18:54the innovation has been much more
18:56important
18:57and any reduction in innovation probably
18:59would have made much bigger long-term
19:01difference than the small static
19:03distortions
19:04in the short term
19:07and as a result he said the most
19:10important thing about capitalism to make
19:12it function well is to have a healthy
19:14competitive process
19:16and how did he describe
19:18that competitive process wesley's not
19:20here i think he dropped out of the class
19:21right
19:22yeah um
19:25does anyone else want to describe what
19:27trumpeter said about committed fraud
19:29bart um
19:30so
19:32it's like you're competing against time
19:34frames as well yeah so against your
19:36future competitors
19:38and pricing based on that competition as
19:41well
19:42yeah so he said that you know the
19:44biggest threat to a monopolist is not
19:46like that somebody else offers the same
19:48product for a cheaper price but that
19:50someone else comes along with a product
19:52that's so much better that the
19:53monopolist that it's just not even a
19:55relevant product anymore right
19:57um
19:58and
19:59uh
20:00so the the real
20:02goal should be to encourage this sort of
20:04creative destruction where people are
20:06constantly coming up with products that
20:07wipe away old industries
20:11and that in order for companies to be
20:13able to afford the r d that they need to
20:15do in order to do that type of creative
20:17destruction it may actually require
20:18concentration
20:20and that innovation
20:24is encouraged by the hope of getting at
20:27least temporary monopoly power and
20:29therefore the focus of industrial policy
20:31should really be on encouraging the
20:33creation of new products rather than on
20:36just minimizing the static distortions
20:38that are often the focus
20:39at least have been the focus for us for
20:41the last few lectures
20:43and i think the schumpeterian
20:44competitive process resonates well with
20:47high-tech industries
20:49like you know facebook
20:52really didn't offer like a lower price
20:55but swept away all the earlier
20:57generations of firms apple has
21:00repeatedly created products
21:03that um
21:04have really revolutionized the industry
21:07and it doesn't do it just by having
21:08incrementally higher quality or
21:10incrementally lower price but when it's
21:13been really successful by having a
21:15completely different product though
21:16recently they haven't really been doing
21:18that they've been doing these
21:19incremental innovations which are you
21:21know is something that people are afraid
21:23of in terms of apple's prospects for the
21:25future
21:26google similarly has sort of busted open
21:29many different segments of the market it
21:31was a late comer actually to the search
21:33market
21:34and then within a few years completely
21:36dominated it
21:37and has brought you know
21:40sort of revolutionized the maps part of
21:42the market revolutionized
21:45email many of these different areas
21:48and
21:50mobile
21:51operating systems etc
21:54and so you know
21:56there's a very dominant position that
21:58google now has in the market as a result
22:00of this
22:01but many people would say that no matter
22:04what problems that brings they might not
22:06show
22:07links as accurately for rivals
22:11um those are really small relative to
22:13just like the way that google's products
22:15have totally changed the way we interact
22:17with the internet and that anything that
22:19would reduce the chance that we would
22:20have gotten google
22:22is even if it
22:24has you know creates an incremental
22:26increase in the benefits we get out of
22:28google is probably not worth it so
22:30that's sort of jumper's concern
22:33um yeah christian oh um well sorry
22:37yeah you said it resonates with high
22:38tech would it be fair to say that it
22:39resonates with all like fat tailed
22:41industries where you have an even
22:43greater drive uh to have that
22:44competitive process yeah i mean i think
22:46that tech has been the area with like
22:48the largest amounts of innovation in
22:50recent years but like you know medicine
22:52has not well
22:54new medicines
22:56have but like a lot of the process of
22:58the medical sector hasn't advanced that
23:00much
23:00so um and you know car industry hasn't
23:03advanced that much in recent years so i
23:06don't know that it resonates as much in
23:07those types of industries
23:10um
23:12i mean it's called tech for a reason
23:14right it affects lots of different areas
23:16but it's an industry that's dominated by
23:18like in new types of technologies right
23:21um
23:23smart phones
23:24uh and apps are another example of this
23:26at least a few years ago i'm not sure
23:28this is true as true anymore but like
23:30every six months there's just like
23:32totally new types of smartphones out
23:34that totally wiped away the last
23:36generation
23:39and this is
23:40i mean less clear i think in in other
23:42industries
23:44even operating systems have not advanced
23:46that much
23:48over the last 20 years for example
23:51okay
23:52so
23:53i think to show the importance of
23:55dynamics uh compared to the static
23:58distortions uh i think that a useful
24:01analogy is to crime
24:03so most criminal activity especially
24:05property crime is committed by the poor
24:07against the rich
24:09and a famous uh quote talking about this
24:12is the
24:14law in its majestic equality forbids the
24:17poor and the rich man equally from
24:19sleeping underneath the bridge
24:21or stealing a loaf of bread
24:24and
24:27as a result most property crime is
24:29actually redistributive
24:31and to a large extent entails little
24:34direct efficiency loss it just transfers
24:36from one person to the other and usually
24:38the person who's stealing needs the
24:39thing more than the person they steal it
24:41from
24:42because of declining marginal utility of
24:44wealth and thus from a static
24:46perspective crime should be legal right
24:49crime is beneficial when you look at it
24:51from a static perspective right
24:54um
24:55so i think this is sort of an absurd
24:56conclusion that forces you to examine
24:59the pres
25:00premises and the key premise here was
25:03that we were only thinking about the
25:04static in you know short-term effects of
25:08crime and not about what would be
25:11encouraged in the long term if criminal
25:13activity
25:14were not prevented so one thing would be
25:18people would spend all their time
25:19stealing rather than working and being
25:22productive
25:23which would undermine people's
25:25incentives
25:26to actually you know produce things
25:28um
25:30we
25:31uh people would
25:33very be spend a lot of money trying to
25:35prevent other people from stealing
25:37which would be wasteful
25:39and
25:41because people wouldn't be able to hold
25:42on to the fruits of their labor people
25:45uh that would probably lead to too much
25:47redistribution relative to the optimal
25:49level of redistribution which we're
25:51going to talk about in the next class
25:54now none of this shows up in a simple
25:56static analysis
25:57and therefore of course we couldn't
25:59think about crime unless we thought
26:01about both uh these short-term benefits
26:05maybe but also the long-term uh costs of
26:09allowing crime
26:10and uh schampeter would argue that this
26:13is true of industrial economics as well
26:15if we spend all of our time thinking
26:17about these static distortions right if
26:20we think of all of our time
26:22spent thinking about like whether enough
26:25of the good is being produced we're not
26:26going to think about what's really
26:27important which is the encouragement of
26:29innovations
26:32and
26:34there are other similarities actually
26:35between the analysis of crime and
26:37industrial organization in particular in
26:39crime
26:40unlike in the pigouvian context we
26:43talked about with externalities we
26:45usually have to use punishment which is
26:46costly to inflict
26:48rather than fines which are just a
26:50transfer as a result it's costly to
26:54deter crime similarly in
26:58in ip
27:00we want to encourage innovation but we
27:02do that by giving a patent which creates
27:04a deadweight loss
27:05and so it's costly to incentivize
27:08innovation as well
27:11and that means that there's a lot of
27:13similarities between the optimal amount
27:16of criminal justice enforcement that it
27:17depends on how much people respond
27:20to enforcement
27:22in the same way that we'll see
27:23intellectual optimal intellectual
27:25property depends on how much
27:28we think people respond by innovating
27:30okay
27:32um so
27:34there's a static
27:37dead weight loss harm created by
27:39intellectual property
27:42it decreases the consumption of the good
27:44that there's intellectual property over
27:47but christian what are some of the other
27:49harms that uh intellectual property
27:51might cause so like the dynamic harms
27:54that we were talking about well so like
27:55it could encourage the racing and so
27:57forth that we're talking about but also
27:59what are what are other things that you
28:00just think of when you think of problems
28:01with intellectual property or harms that
28:03it creates like all the litigation like
28:05especially
28:06the high tech like between samsung and
28:07apple
28:08yeah so that's that's very wasteful
28:10absolutely um and then other things
28:14um
28:15well like what you just flashed yeah so
28:17it's
28:18you can discourage follow-on innovations
28:20right so like a lot of ideas if they're
28:22patented people then won't innovate uh
28:25based on them because they're afraid
28:26that they're gonna have to pay so much
28:27to the initial patent holder that it's
28:29not worth it to them to innovate right
28:32um
28:34another yeah evan so again the follow-on
28:36innovation thing also applied to the
28:37company that originally innovated
28:39so yeah they have a patent they might
28:42not be as interested in developing the
28:43next thing because they know that no one
28:45can can use their product for 20 years
28:49yeah so they might actually want to
28:50commit
28:52to what's called
28:54an open
28:57an open license
28:59because that will encourage other people
29:01to make those investments
29:03uh or they might
29:05commit to saying we're never going to
29:06charge more than a certain amount for
29:08licensing in order to encourage other
29:10people to develop things around it
29:11because they think that in the long run
29:12that'll lead them to make more profits
29:14so that's a good one
29:17so
29:19litigation
29:20is costly and complex
29:23you have to figure out whether a given
29:25application
29:26of
29:27an idea is covered by the patent and
29:30companies often try to avoid a lot of
29:32these costs by
29:34striking deals about
29:36cross licensing
29:38um
29:40also patents could reduce intrinsic
29:42incentives to innovate
29:45because people if they have the
29:47opportunity to get a patent might not
29:49want to share their ideas because they
29:51want to work on them until they get them
29:53to a point where they can patent them
29:57and
29:58the common public license and open
30:00source software is often used to deal
30:01with this issue so you can rather than
30:05patenting something put it under an open
30:07license which says that anyone can use
30:09it but only if anything that they built
30:11is also under the open license so that
30:14allows people to
30:18produce something
30:19without
30:21allowing it to become patented
30:22eventually so that they can get whatever
30:25they produce
30:26adopted very widely because that's often
30:29an intrinsic motivation that people have
30:31it's like they want whatever software
30:33they develop and whatever gets built
30:34upon it to be available to a lot of
30:36people so they get credited and so they
30:37feel that their thing is being used a
30:39lot
30:40and that's becoming a growing part of
30:42software development
30:44so here's a little bit of evidence so
30:45heidi williams
30:47who is a professor at mit
30:50had a very clever
30:52way of looking at the way in which
30:54patents can discourage follow-on
30:56innovations in particular she looked at
30:59the race to
31:00discover the human genome
31:03i don't know if anyone followed that you
31:04guys were probably pretty young when
31:06that was happening but
31:08there were two
31:10groups that were doing this one was
31:12solera genomics which was a private
31:13company and every gene they discovered
31:16they patented
31:17and one was the government and every
31:19gene they discovered they put into the
31:21public domain
31:22and because these genes were like quite
31:24similar and you can find very comparable
31:26ones you can look at were the ones that
31:28were discovered by the government or the
31:29ones discovered by solera the source of
31:32more following innovations and she found
31:35that there was a significant difference
31:37um on average there were
31:40one about one less publication
31:43on the things that were discovered by
31:44solera then by the government there was
31:47about a
31:50fifth less
31:52known
31:56therapeutic
31:58innovations
31:59uh based on that
32:01uh et cetera et cetera so it it
32:03discouraged the patenting by solara
32:05discouraged uh
32:07follow-on innovations
32:10okay
32:11nonetheless getting rewards
32:15to innovation
32:16is very important
32:22at
32:26okay so if
32:28in the best case the entrepreneur
32:32uh get gets all the profits associated
32:35with the innovation but all the profits
32:37are not all of the social value because
32:40when innovation is created it not just
32:42creates profits but it also creates
32:44consumer surplus
32:46and this is called the market maker
32:47externality if someone creates a new
32:49market they only capture part of the
32:52benefit associated with that market
32:54um
32:57now this can be lessened when there's
32:59price discrimination but as we learned
33:01price discrimination is almost always
33:03imperfect and so it's very hard for an
33:06innovator to capture all the social
33:07value associated with their product now
33:10price discrimination plays a important
33:12role in the pharmaceutical industry
33:15uh and it also happens in patent
33:17licensing because there's often
33:18negotiations uh over
33:21the price that will be paid yeah tanner
33:23if there was a wake you'd probably the
33:25prices grenade yeah would innovation
33:27still be beneficial for society because
33:29they would have no external spillovers
33:32but they would still be beneficial
33:34right like the ability to innovate would
33:36still be beneficial would just all be
33:38captured by the people who
33:40did the innovation
33:42it would make innovation
33:44i mean so that's a general thing to
33:46think about all that it means when we
33:48say oh there's not enough of something
33:50like something is socially beneficial as
33:52spillovers is that the person isn't
33:54capturing all the benefit as soon as you
33:55make that person capture all the benefit
33:57then there's it no longer has these
33:59spillover benefits but it's actually
34:01good thing
34:02in the sense that you'll get the right
34:04amount of the activity when that occurs
34:07it's just like having a positive
34:08externality right
34:13okay so formal government protection can
34:17also be useful
34:18to discourage people from holding things
34:20as trade secrets because if they hold
34:21them as trade secrets nobody else gets
34:23to see the way in which the idea works
34:26and therefore other people are unable to
34:28build on that innovation and therefore
34:30uh
34:31you know allowing formal rather than uh
34:34trade secret intellectual property
34:36protection can actually encourage
34:37innovation by allowing other people to
34:39at least see what the person was doing
34:40and build upon it
34:43and that's one reason
34:44why um governments have formal
34:47intellectual property protection is that
34:49otherwise they think there would be so
34:50many trade secrets and so much hiding
34:52that it would actually reduce the rate
34:53of innovation
34:55okay so to look at the incentives that
34:58uh profits give to innovation
35:00um
35:01eric buddish and co-authors in a recent
35:03paper
35:05looked at uh medical trials um in
35:08particular new medicines and in order to
35:10get a new medicine approved you have to
35:13put it through a rigorous process of fda
35:15evaluation where you have to prove that
35:17it's actually beneficial
35:18now the more people that die
35:21the more frequently from your disease
35:23the easier it is to prove that you're
35:25reducing
35:27the
35:29the amount
35:30of
35:31um
35:33the the number of deaths if you have if
35:35you have something that's like an early
35:37stage cancer that eventually will kill
35:38people but it takes a very long time to
35:40kill people
35:41versus a late stage cancer it's going to
35:44be much quicker to get the late stage
35:46cancer drug approved because you'll be
35:47able to more quickly prove that it was
35:49actually reducing deaths
35:51right
35:52so
35:53here
35:54is a little study about this so
35:56metastatic cancers people die from very
35:59quickly
36:00localized cancers people take a lot
36:02longer to die from and regional cancers
36:04they die from an intermediate amount of
36:06time
36:07and this shows as a function of that how
36:11many clinical trials there are on each
36:13of these types of drugs and you find
36:15there's a lot more clinical trials on
36:16the metastatic ones because you get it
36:20to market more quickly and therefore you
36:21earn more profits on it for two reasons
36:23first those profits are sooner so
36:26they're not as discounted and second
36:29you have more of your patent life still
36:31left because you get your patent when
36:32you first file for the fda trials
36:36but then you have to wait for many years
36:38in order to actually start earning
36:41profits off of that and that reduces the
36:43profit
36:44that take a long time to get through the
36:46fda relative to things that get through
36:48quickly
36:50okay
36:51so um
36:54this
36:55creates a fundamental tradeoff
36:58between incentives for innovation
37:00created by ip
37:02versus the x post distortions created by
37:04ip the more i p protection
37:07the more dead weight loss we're going to
37:09have
37:10but also the more i p protection the
37:12more benefits we get out of innovations
37:17and an important thing to realize is
37:19that
37:20when we have the first little bit of
37:22protection imagine we elevate the price
37:24just a little bit above cost that hardly
37:27causes any deadweight loss because the
37:29deadweight loss triangle is a triangle
37:31right but it creates a significant
37:33amount of profits because that's a
37:34rectangle
37:37on the other hand as we get towards the
37:38monopolist optimal
37:41price we're hardly getting any more
37:43profits by raising price
37:45but we're creating quite a lot of
37:48deadweight loss because the price is
37:50already pretty high
37:52and so
37:54the optimal protection is always going
37:56to be partial we're never going to want
37:57to go all the way to the monopoly
37:59optimum because we're hardly giving him
38:00any more benefits and getting a lot of
38:01deadweight loss for that but we're not
38:03going to want to have no protection
38:05because we could get a lot of profits
38:06for only a little bit of dead weight
38:08loss
38:09and so the question is what amount of
38:12partial protection do we want to give
38:14well it's going to depend on the
38:15trade-off between the cost and the
38:17benefit so bart remind us of the costs
38:19of
38:21intellectual property
38:23um
38:26well
38:27aside from like the deadweight loss so
38:29it creates a deadweight loss by
38:30elevating prices right and what were the
38:32other things we just mentioned
38:38so like discourages innovation follow-on
38:41innovations absolutely
38:43um and tanner what are the benefits
38:46of intellectual property again
38:47incentives incentivizes people to
38:49produce your products yep that's that's
38:51the main incentive yeah and also to
38:53disclose existing uh information rather
38:55than hold it as a trade secret
38:57potentially
38:59okay so
39:00i want to just focus on the
39:02distortion from of consumption versus
39:05the incentives so imagine that we have a
39:07bunch of innovations that could be
39:08created
39:10um
39:12and imagine they have no marginal cost
39:13of production because most uh
39:17ideas can be reproduced pretty cheaply
39:20and that on average they have a demand
39:22function q of p
39:25and i'm going to just normalize this so
39:27that
39:28the quantity that's produced
39:31when the
39:34price is zero is one
39:37and the quantity that people want to buy
39:38when the price is zero is one and that
39:40they have a monopoly price of one
39:42and then the consumer surplus
39:45is the
39:47integral from whatever price
39:49is being charged to infinity of the
39:51quantity that's the normal consumer
39:52surplus formula
39:54firms will earn a profit
39:56of
39:58p times q because there's no cost
40:02and they're only going to innovate if
40:03that profit is high enough because
40:05innovation is costly to produce
40:08and julian what are some of the reasons
40:10why innovation is costly to produce um
40:13well you have to do
40:15um research and development and it might
40:17not work uh so
40:19there's
40:20like it's there's a probability that you
40:23actually produce the thing yeah
40:25absolutely so like whatever even if
40:27research and development costs don't
40:28sound that large like they're very
40:31uncertain whether the product is going
40:32to be developed and whether it's going
40:33to be profitable once it's developed so
40:35you sort of have to divide those costs
40:36by the probability that they actually
40:38succeed right
40:40and so we're going to summarize that by
40:42there being a supply of innovations
40:44s of pi
40:46okay so
40:48the total welfare created when there's a
40:50certain number of innovations is going
40:52to be the producer benefits and the
40:54consumer benefits
40:58and can somebody tell me what those
41:02two benefits are what the formula for
41:03them are
41:09yeah julian do you want to try or check
41:11yeah girlfriend should just be the
41:13consumer surplus
41:14plus the profits right well so this is
41:16the consumer surplus per product but how
41:18many products are going to be created oh
41:20um
41:24well no it's supply of the innovations
41:26right so it's s of pi times cs of
41:29p right
41:31uh
41:32and then you need the producer benefits
41:34which we can get by integrating under
41:37the supply curve of innovations right
41:39from a profit of zero to the profit that
41:41they're actually earning
41:45okay
41:49um so
41:51uh christian could you come up and take
41:52the first order condition associated
41:54with this
41:55yeah
41:56with uh with respect to p
42:01so p is like the price or the amount of
42:03intellectual property protection we're
42:04giving do let me keep this up
42:07you can erase that is everyone okay with
42:08that stuff about the grades
42:10yep
42:25all right so we have
42:47i just want to make sure i get this
42:48right
42:54deep
42:56oh yeah
43:04first order condition is product rule
43:06and chain rule
43:08so that's
43:10s prime of i
43:13p
43:15times pi
43:17prime of p
43:20times consumer surplus
43:22e
43:23plus the other term of the product rule
43:26which is s of pi
43:28p
43:31times consumer surplus from
43:35p
43:37and then for your second term that's
43:41um
43:44s
43:46prime
43:48no that's live nets remember like the
43:50leibniz role
43:51oh
43:52so
44:02that you set that equal to zero right
44:05okay
44:06um
44:07nicely done christian
44:09okay so that's what christian just wrote
44:11out
44:13um
44:14okay
44:15so
44:17let's drop um
44:19all of the arguments
44:21of this and then
44:23julian could you find some way to try to
44:26simplify this do you have any thoughts
44:35so we have
44:36pi prime times
44:39s prime
44:40of
44:46let me drop the argument of s prime
44:53yeah both of them are s so we can factor
44:55that out
44:56okay
44:57uh right so
44:58so pi prime
45:00uh
45:01s prime plus s
45:03that s prime c s plus s
45:06uh equals
45:08negative c s prime s
45:12yep so that's this right so we got
45:15pi prime times s prime times cs is equal
45:18to negative s times cs prime plus pi
45:21right is that what you're saying yeah
45:23yeah
45:25okay
45:26um
45:27okay so now notice that deadweight loss
45:31is equal to the maximum possible surplus
45:34which is the consumer surplus but
45:35there's no profits but there's no cost
45:37when the price is zero
45:39minus the consumer surplus and
45:42profits that we actually get
45:44and so
45:45negative c s prime plus
45:47pi prime is deadweight loss prime
45:50right
45:52okay so then how can we rewrite this
45:55expression tavish
45:58look up
46:02how can we use this here
46:06pi prime
46:09negative c s prime
46:11minus pi prime is equal to deadweight
46:13loss prime where is that useful in in
46:15the formula up here
46:17um
46:23on the
46:24the second of the two terms
46:26yeah what what can you replace here
46:32yeah
46:35yeah and then it cancels with a negative
46:36so you just get s times deadweight loss
46:38prime
46:42okay so if you then divide through by
46:44the supply and multiply by pi
46:47this becomes the elasticity of the
46:49supply of innovations here
46:52and you get the elasticity of supply of
46:54innovations times consumer surplus
46:56divided by profits is equal to the ratio
46:58of marginal dead weight loss to marginal
47:00consumer surplus
47:05sorry it's not marginal consumer surplus
47:07marginal profits that's a mistake it
47:10should be pi prime there
47:12okay so this is the crucial formula and
47:14here's it correctly written
47:16the elasticity of supply of innovations
47:19times the ratio of consumer surplus to
47:21profits
47:22is equal to the ratio of marginal dead
47:24weight loss to marginal profits
47:26okay so how do we think about that so
47:29here's how much innovations respond to
47:32additional profits
47:34here's the ratio of how much external
47:37benefit they're giving to how much
47:38people are capturing
47:41and here is how much
47:42social loss we have to give up to give
47:44an additional unit of profits to the
47:47firm
47:48right
47:49okay so this means that i p protection
47:52should be greater
47:53under what circumstances christian
47:58um
48:00protection should be greater so
48:04if the marginal costs of incentive are
48:06greater right
48:08then what do you mean by that
48:11so we should give more intellectual
48:13property protection this is a okay so we
48:16should give more let me start over we
48:18should have more intellectual property
48:19protection
48:20to products that will have a greater
48:22positive externality yeah where there's
48:25more consumer surplus being generated
48:27yeah and then where the elasticity
48:31of innovation supply is high yeah
48:34absolutely
48:36um
48:38and where it's less costly in terms of
48:41dead weight loss to reward
48:43for per unit of profit we give the
48:45deadweight loss is smaller
48:48okay
48:52so
48:56we
48:57it would be
48:58good
48:59to
49:00verify these basic intuitions that the
49:02higher the elasticity is the more
49:03protection should be given and the more
49:06deadweight loss there is for unit of
49:07profit the less would would be optimal
49:12okay so to do this we need to tie down
49:14the demand form a little bit and i'm
49:16going to use a very simple
49:19demand form
49:22remember that pass-through
49:24affects both the ratio of consumer
49:26surplus to profits we talked about that
49:28right and also the ratio of deadweight
49:30loss to profits
49:33the more curve demand is the higher the
49:35pass-through the more uh consumer
49:37surplus there should be for every unit
49:39of profits but also the more dead weight
49:40loss there should be forever units of
49:42profits and a simple demand form that
49:45allows us to vary around the
49:46pass-through
49:47is called the constant pass-through
49:49class this contains all the demand forms
49:52that you've seen so far
49:54and worked with in this class
49:56it includes constant elasticity linear
49:59demand
50:00and exponential demand and the form more
50:02generally is given by this
50:05so notice that if rho is equal to
50:07one-half
50:08these cancel and this becomes linear
50:09demand
50:11right if
50:13uh
50:15if we take
50:16a particular limit then we get constant
50:19elasticity demand and as rho goes to 1
50:21we get
50:24we get
50:25uh exponential demand yeah this is
50:27demand for innovation yeah this is the
50:29queue
50:30function q is demand for innovation it's
50:32the demand for a given product so like
50:34when you create an innovation a new
50:35market is created this is the demand in
50:38that
50:38market yeah ratio when row is
50:42path
50:43that's saying that half that incidence
50:45is bored by the consumer have better
50:47patients
50:50this is under monopoly so under monopoly
50:52the incidence is row over one
50:55so it's one half would be borne by the
50:56consumer and all would be born also by
50:59the monopoly because the thing is
51:01there's more than
51:02complete incidents right
51:07row would be one would be that like
51:09they're equal amounts for the for the
51:11consumers
51:13and if row is above one but it also so
51:16that makes the cs to pi ratio greater
51:19but it also makes the deadweight loss to
51:20pi ratio greater
51:22and we'll see that that second effect is
51:24actually bigger
51:26i think
51:29okay
51:30so
51:31uh
51:34cs in this form you can just integrate
51:37up and it turns out to be this
51:39deadweight loss turns out to be this
51:41etc so you can just everything is really
51:43easy to work with in this form
51:45so in this uh
51:47within this class we can just write a
51:50little mathematica code to figure out um
51:53all these things
51:54uh elast many sub studies have estimated
51:58the elasticity of innovation supply in
52:00high tech industries it appears to be
52:02about 0.5 to one
52:04but in it tends to be higher in um
52:10in
52:11uh medical industries so how do we
52:13figure this out so what you try to look
52:15for is something that would increase the
52:17returns to innovation and look at how
52:19many more innovations you get so i had a
52:21student who actually did this a few
52:22years ago at harvard what he did is he
52:24looked when verizon announced that it
52:26would allow
52:28start having the iphone rather than just
52:29at t having it and he looked at how many
52:31more iphone apps you got
52:33uh in the following months compared to
52:36what you would have gotten before that
52:37was announced
52:38and he found that it increased
52:40significantly and he related that to how
52:42much more profits they would have made
52:43how much bigger the market got for
52:45iphone apps and he got an elasticity of
52:48about 0.8 or something like that
52:51on the other hand
52:53remember that we showed those different
52:55clinical trials and that things make
52:56different amounts of profits based on
52:58the horizon
53:00of how long it takes to get things
53:03approved you can figure out how much
53:04profits difference that translates into
53:07and therefore you know read off of that
53:09graph how much the elasticity of
53:12innovation supply is there and that
53:13turns out to be closer to about three so
53:15the elasticity of innovation supply in
53:17these areas where it's much more costly
53:18to innovate like medicine compared to
53:20high-tech tend to be higher yeah bart uh
53:23is the reason for that higher elasticity
53:25of
53:26supply
53:27uh because of like
53:29the cost of typewriters
53:31well i think the main reason is just
53:33it's very costly to innovate in medicine
53:34because you have to run these fda trials
53:36you have to raise capital you have to
53:38like do a whole bunch of experimentation
53:40whereas in
53:41high tech it tends to be less costly to
53:43innovate and therefore less elastic
53:47um
53:49okay
53:50pass-through data is much harder
53:52remember the inof paper we saw should
53:55the pastor might be infinite over some
53:56ranges at least
53:58um there's a
54:00nice paper by pasenko that looks at a
54:02lot of consumer products and how changes
54:04in their cost gets passed through uh at
54:07supermarkets and they found pass-through
54:09was roughly between point two and five
54:10uh point five and two and all um
54:14i'll use that range
54:16okay so i just ran these things on
54:17mathematica so we had all these formulas
54:19right we can just plug them in
54:21uh over here
54:23right
54:24and here's what you get
54:27so first
54:28i considered a pass-through of one and i
54:31raised the elasticity of innovation
54:33supply and that led to an increase
54:35in the
54:36optimal level of ip protection
54:40right
54:40on the other hand pass through when i
54:42moved it from 0.5
54:44to
54:481
54:49sorry from 0.5 to 2. you see that that
54:52increases both curves but it actually
54:54reduces the amount of intellectual
54:56property protection because it turns out
54:57to increase the dead weight loss more
54:59than it increases the consumer surplus
55:02at least in this case
55:04okay
55:05so
55:06we get that the higher the deadweight
55:08loss to pi is the lower the optimal
55:10protection
55:11the higher the elasticity of supply is
55:13the more the optimal protection and
55:15therefore um if you look over the full
55:18range of the parameters that i was doing
55:20basically i p protection ranges from
55:22about .45 up to about
55:250.65
55:27okay so how can we relate this to
55:30practice
55:31kim how might we try to align this uh
55:35degree of intellectual property
55:37protection with something we could
55:38actually measure about the current
55:40intellectual property regime that we
55:42have
55:44we can measure it
55:45by durational
55:47right by what duration of time yeah
55:51yeah yeah that's right so
55:52how might we think about doing that
56:10well one idea is that patent life in the
56:12u.s is about 20 years right we can think
56:15there's about five percent interest rate
56:18right maybe that's you know
56:20fairly common that means that if i had a
56:23patent forever that would basically be
56:25worth 20 years of having it this year
56:28right because uh
56:30point
56:31uh
56:33um
56:34if you do the infinite sum
56:36with 0.5 it gives you 1 over
56:400.05 which is about 20
56:42right which is 20. on the other hand if
56:45i have it for only the first 20 years
56:51then
56:51that turns out to give me about .64 of
56:54what it would be if i had it forever
56:58and so that says that u.s patent
57:00protection currently is roughly 64 of
57:03what it would be if you had infinite
57:05life patents
57:08um and that says that u.s patent
57:10protection is on maybe on the upper
57:12range of where we were looking at but
57:14pretty close to the range that we were
57:16talking about which is kind of
57:17interesting
57:19but of course um we should try to take
57:21into account some of the other harms and
57:23benefits associated with patents like
57:25the discouragement of follow-on
57:26innovation
57:28uh as well as the benefits of disclosure
57:30which could effectively raise the ratio
57:31of cs to pi yeah
57:33like the output
57:34should be between 0.45 0.65 yeah what
57:37are those numbers being exactly well
57:40it's like the fraction of full
57:41protection that you're getting
57:44um
57:46so unlike u.s practice however probably
57:49patent life should differ by industry
57:51right
57:52um now in practice some of this may come
57:56from the fact may actually occur because
58:00judges are more loath to enforce patent
58:03protection in high-tech areas than in
58:06medicine
58:07and that might effectively create
58:08different patent protections
58:11but this is much less developed
58:14than
58:16our the way that
58:17in general criminal enforcement responds
58:20to the elasticity so we um are much less
58:23likely to throw a young person into jail
58:26when they commit a crime than an older
58:27person because we don't think that
58:28they're thinking it through and
58:29therefore not responding as much to that
58:32enforcement similarly intellectual
58:34property there hasn't been as much
58:35thinking through the circumstances in
58:37which there's going to be likely to be
58:39higher elasticity of innovation supply
58:41or lower and thinking about the
58:43enforcement of intellectual property
58:44laws depending on that
58:47but in principle we should want in
58:48industries with low elasticity and high
58:51dead weight loss to have less patent
58:53protection and i think a lot of people
58:55think the tech sector is that way
58:56because it's less costly to develop
58:58things and it's so hard to keep track of
59:01who's infringing or not infringing that
59:04deadweight loss is likely to be high
59:06whereas in pharmaceutical sector it's
59:08much more costly to develop and it's
59:09much easier to track whether someone's
59:11using the product or not and as a result
59:13probably the protection there should be
59:14higher so a lot of people think there
59:16should be longer patent lives than
59:18pharmaceuticals but shorter ones or
59:20weaker protection in
59:24high tech
59:26okay
59:27so while this was sort of the most basic
59:29trade-off associated with intellectual
59:31property
59:32there are many other tools to encourage
59:34innovation so for example you can give
59:36subsidies for research and development
59:38and that's extensively embedded into a
59:40series of tax credits
59:43the national science foundation
59:45other subsidies that the us government
59:47gives
59:48there are also private subsidies for ip
59:51effectively coming through this stuff
59:53that i was saying about how att had a
59:56research lab that they were funding
59:58etc
1:00:00another thing you can do is you could
1:00:02try to avoid the deadweight loss of
1:00:04innovation while still giving the
1:00:05incentives by trying to give prizes for
1:00:08innovations
1:00:10or to buy out existing innovations so
1:00:12for example if there's a really valuable
1:00:15patent but there's a big distortion
1:00:16that's going on because of
1:00:19the patent protection the government
1:00:20might offer to buy the patent from the
1:00:23company to put it into the public domain
1:00:25so people could produce it more freely
1:00:28and historically this was a really big
1:00:30thing so napoleon and the british
1:00:33government offered lots of prizes for
1:00:35innovations and the x prizes are trying
1:00:38to do that as well
1:00:41you can also give subsidies to people
1:00:43for purchasing innovative products that
1:00:45increases the demand in the market and
1:00:47therefore increases the profits the firm
1:00:49will earn
1:00:51this is a common argument for why we
1:00:53should subsidize people using like green
1:00:55technology
1:00:56it's not just because
1:00:59it's beneficial to reduce carbon but
1:01:01that you'll incentivize the creation of
1:01:03more green technologies that way
1:01:05you can also think about whether patents
1:01:07should be broad or long so we talked
1:01:10about patent length but there's also an
1:01:12issue of how broad protection how uh
1:01:15you know broadly your patent is
1:01:17interpreted right so that's another
1:01:19trade-off
1:01:20there's also lots of requirements in
1:01:22order to get a patent so in order to get
1:01:25a patent in the united states you have
1:01:26to prove that your patent that what
1:01:28you're proposing is novel that nobody's
1:01:30done it before that it's useful it has
1:01:32some practical applicability and that
1:01:34it's not obvious
1:01:36and uh those are adjusting what those
1:01:40things mean and how they're interpreted
1:01:42can also
1:01:43impact patent protection
1:01:45in some sense this is a form of patent
1:01:48breadth
1:01:49that comes through the
1:01:51requirements of the code but there's
1:01:53just genuine questions about how these
1:01:55should be interpreted there's also
1:01:56issues about how profits get divided
1:01:59when there are sequential innovations
1:02:00like imagine i innovate something and
1:02:02then you do something that builds on
1:02:03that how should our profits be split how
1:02:05should the courts uh interpret how we
1:02:08should uh do a bargain about that
1:02:11um yeah right so um
1:02:14a subsidy would be similar to like uh
1:02:16patents that vary in length by industry
1:02:18right
1:02:19because
1:02:20well no but they don't create dead
1:02:22weight loss subsidies don't create dead
1:02:24weight loss they're like a transfer
1:02:26right okay
1:02:29so they're a bit different
1:02:31okay so there are different trade-offs
1:02:33that are useful in thinking about each
1:02:35of these different tools of policy not
1:02:37just the basic ip trade-offs so for
1:02:39example when we give researcher
1:02:42development subsidies
1:02:45and
1:02:46corporate uh we could also
1:02:49uh adjust the corporate tax rate
1:02:52the benefit of doing these is they might
1:02:54allow of like lowering these might be to
1:02:57allow
1:02:58um
1:02:59internalization of the consumer
1:03:00surpluses created by the
1:03:03market
1:03:04but on the other by creating a new
1:03:07market but on the other hand it gives
1:03:09more money to people who run companies
1:03:10right
1:03:12so if you raise the corporate tax rate
1:03:14while giving subsidies for r d then you
1:03:16can
1:03:17sort of target specifically innovations
1:03:21and without redistributing too much to
1:03:23firms so that's something that people
1:03:24often advocate
1:03:27um
1:03:28my job market paper uh that's like the
1:03:30thing that you go around and present so
1:03:32that they try to offer you a job at a
1:03:34university
1:03:35um
1:03:36explore the trade-off between ip versus
1:03:39prizes
1:03:41and
1:03:42the basic idea and this there's a famous
1:03:44quote from adam smith
1:03:46saying that um
1:03:48that if the government tried to give out
1:03:50prizes for like the greatest things that
1:03:52were created in the world it couldn't do
1:03:53nearly as good of a job as it would do
1:03:55if it just gave out patents and let the
1:03:57market determine how much people get
1:03:59rewarded because the thing is you won't
1:04:01make money off of patent unless people
1:04:02actually want to use the thing but you
1:04:04will make it if they do and so the
1:04:06government might have a hard time
1:04:07figuring out how much things could
1:04:09should be rewarded if it didn't have
1:04:11the market to tell it and that's one
1:04:13reason why you might have ip over prizes
1:04:16even though prices avoid the deadweight
1:04:17loss
1:04:20there's a trade-off between how long to
1:04:22make patents and how broad to make them
1:04:25and one argument is that you might want
1:04:27to have them for a very very long time
1:04:29but very thin
1:04:31why
1:04:32because
1:04:33that allows you to minimize in each
1:04:35period the ratio of the deadweight loss
1:04:37to the profits because remember
1:04:39initially the profits grow and then
1:04:41eventually the dead weight loss is the
1:04:42main thing that's growing
1:04:46there's also a bunch of issues about the
1:04:49division between sequential innovations
1:04:51so often if there's a bunch of
1:04:53innovations they'll all get put in that
1:04:56are all important to some process they
1:04:58build on top of one another they'll
1:04:59often get put into a patent pool
1:05:02and those pools will determine how the
1:05:03revenues are split and that depends a
1:05:05lot on government regulations on
1:05:08how the courts interpret things and so
1:05:10forth for example there's now a big
1:05:12debate about what's called fair
1:05:14reasonable and non-discriminatory
1:05:16uh
1:05:18licensing which is basically when there
1:05:21are a bunch of patents that all are
1:05:22relevant to something and you put them
1:05:23all into a package people have to agree
1:05:25to charge only a reasonable rate what
1:05:28that reasonable rate is is something
1:05:30that the group determines together and
1:05:32in principle
1:05:33our analysis says that it should be
1:05:35targeted depending on how much how much
1:05:37elasticity there was to developing that
1:05:40innovation how important it was that
1:05:42they get a big reward in order to
1:05:43encourage them to have created in the
1:05:45first place
1:05:46and the interpretation of the criteria
1:05:48are still largely an open question and
1:05:50people are still trying to think about
1:05:52what are the right trade-offs to think
1:05:53about in terms of how much we should
1:05:55require to be non-obvious or not um
1:05:58novel et cetera yeah dick in regards
1:06:03yeah isn't it possible to let the market
1:06:06decide what uh
1:06:08the patent is worth and then have it
1:06:11removed by it
1:06:12yeah so if you do that though the
1:06:14problem is that people might be
1:06:17incentivized to gain that by like trying
1:06:19to make it look like there's a lot of
1:06:22demand in the first year
1:06:24and then it doesn't turn out to actually
1:06:26be as good or there could be some
1:06:27products that are late bloomers
1:06:29and other products that like have a bang
1:06:32and then peter off and you wouldn't do a
1:06:34good job distinguishing between those if
1:06:35you did it did it that way but of course
1:06:38yeah there should be some amount of that
1:06:39for sure that could definitely improve
1:06:41but there's still challenges i'm trying
1:06:43to
1:06:44try to figure
1:06:46okay great guys thanks

About this transcript

This page contains the full transcript of ECON 20110 Fall 2013 - Lecture 12: Intellectual Property by Glen Weyl, generated from the public captions YouTube serves with the video. The transcript has 9,814 words across 1,845 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to choose a caption track available on the video.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, caption language selection, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.