00:00:04
this research we are trying to do is to
00:00:07
give an operational perspective to
00:00:09
policymaking in particular for the
00:00:12
adoption of solar panels research were
00:00:15
working on in general for green
00:00:16
technology it's common to see subsidies
00:00:19
that will help the technologies take
00:00:21
dick off faster these technologies are
00:00:24
very expensive in the beginning and they
00:00:26
need some sort of public support in
00:00:28
order for them to become economically
00:00:29
viable we see a lot of policies out
00:00:34
there they have a certain target in mind
00:00:37
lakin in European in you had twenty
00:00:40
percent of renewable by 2020 California
00:00:43
was thirty-three percent he had a
00:00:45
million rooftop programs that they
00:00:47
wanted to cover a million rooftops with
00:00:49
solar panels in Germany and this is sort
00:00:52
of the inspiration we started from so
00:00:53
instead of trying to understand where
00:00:55
these policies come from we're trying to
00:00:57
understand given the policy with a
00:00:59
certain target how do you get there in
00:01:02
the most efficient whale in the cheapest
00:01:03
way for the for the taxpayers and some
00:01:09
of the things that we we found that it's
00:01:11
important to consider the feedback loops
00:01:14
in this market dynamics you have more
00:01:17
solar panels being adopted that makes it
00:01:20
cheaper for the next solar panel to be
00:01:22
sold on average that makes it a more
00:01:25
word-of-mouth being spread about tech
00:01:27
technology and a good policy maker
00:01:30
practice should incorporate these
00:01:33
feedback loops when designing the policy
00:01:39
the key takeaway from from this research
00:01:42
is that you should take into account
00:01:45
these feedback loops such as learning by
00:01:50
doing and information diffusion when
00:01:55
designing a policy and the outcome of
00:01:58
the the optimal policy that you designed
00:02:01
to achieve a certain target is not it's
00:02:04
not trivial in the sense that you would
00:02:06
not just make a lump sum of an
00:02:10
investment in the beginning and let the
00:02:12
the market figure itself out you
00:02:16
actually observe a certain pattern of
00:02:19
yes early subsidies do stimulated option
00:02:23
but they also increase the impact of
00:02:26
future subsidies right so there's a
00:02:29
certain way you would like to distribute
00:02:31
the subsidies throughout the time
00:02:33
horizon in order to reach to your target
00:02:35
in in in the most efficient way
00:02:43
so a typical economics approach to
00:02:47
understanding policymaking starts with a
00:02:49
definition of what's the system welfare
00:02:51
and we abstracted a bit from that we
00:02:55
took a we went to the next level of
00:02:59
decision-making which is okay suppose
00:03:00
that I don't know what's the welfare but
00:03:02
I just have a given target I assume that
00:03:05
this is a target of adoption that I need
00:03:08
to get there and it's it's hopefully
00:03:12
motivated on some kind of welfare
00:03:15
optimization right the society will be
00:03:17
better off by some objective measure and
00:03:20
when we developed this framework what we
00:03:23
try to do is we try to back out what
00:03:25
would be the optimal welfare a function
00:03:30
that would justify the current subsidy
00:03:32
path right historically we see in
00:03:35
Germany a certain level of subsidies
00:03:38
being given out and I would like to try
00:03:40
to understand what sort of economic
00:03:43
welfare measure would justify that given
00:03:46
historical path and we were surprised to
00:03:50
see that there is no such welfare
00:03:52
function that we justified that history
00:03:55
there is a way to show just using this
00:04:00
framework that more subsidies would have
00:04:04
been required in the beginning it was
00:04:07
stronger a subsidy policy perhaps and a
00:04:11
stronger phase out in the later stages
00:04:13
of the program that's the indications
00:04:17
that we have we we have to put a
00:04:20
qualifier on that that we don't do a flu
00:04:22
full empirical study of the German solar
00:04:25
market we have very limited access to
00:04:27
data we just use aggregate data for the
00:04:31
yearly adoptions in the average levels
00:04:35
of subsidy but it's a it's a small
00:04:39
indication that we can try to measure
00:04:41
the efficiency of a policy program and
00:04:46
characterize that it's it was somewhat
00:04:49
inefficient even without having a clear
00:04:51
measure of welfare
00:04:56
so the most practical implication is try
00:05:01
not to be myopic that's the most
00:05:04
important recommendations try to
00:05:06
understand what our future implications
00:05:09
of your policies today in the market
00:05:12
dynamics a lot of policy sometimes has
00:05:15
decided what is the level of the subsidy
00:05:18
that we are going to do here let's try
00:05:20
to give investor a certain rate of
00:05:22
return and we'll hope that that gets us
00:05:25
close to our target right but when you
00:05:30
do this over years you have to
00:05:33
understand that the early adopters will
00:05:35
create an impact on future adoption and
00:05:38
if you don't take that into account then
00:05:41
you'll you're you're either miss your
00:05:44
target or you're going to spend a lot
00:05:46
more money than you should
00:05:51
so companies need to also understand how
00:05:55
learning happens in their industry and
00:05:58
how it will affect future policy
00:06:03
decisions so one of the most problematic
00:06:08
things we have in the this area of of
00:06:12
subsidy policies is the abrupt changes
00:06:15
in subsidy policies that has something
00:06:17
to do with some of the future research
00:06:18
that I'm working on and trying to
00:06:21
understand how changes in policy making
00:06:23
affect the industry and vice versa right
00:06:26
so there's a strategic interaction
00:06:27
between how companies are behaving and
00:06:30
how policymakers will respond and vice
00:06:32
versa for consumers I think it's it's
00:06:37
important for consumers to try to
00:06:39
understand how to time the market a lot
00:06:42
of consumers from anecdotal evidence
00:06:45
we've seen that consumers are trying to
00:06:47
wait for solar panels to be cheaper and
00:06:49
they're not necessarily doing the
00:06:51
optimal thing as well from what we have
00:06:54
observed that like the rate of subsidies
00:06:57
have been decreasing faster than the
00:07:01
cost of the installation so in a way
00:07:04
timing yourself into this market is not
00:07:08
necessarily as as easy as you might
00:07:10
think
00:07:15
so we are doing a few other studies in
00:07:20
other Smart Grid applications electric
00:07:24
vehicles there's another small study we
00:07:26
do where we use the application of the
00:07:27
of the adoption of electric cars and
00:07:33
it's a very similar type of market
00:07:36
dynamics that's happening and the same
00:07:38
lessons of you need to incorporate like
00:07:42
learning by doing an information
00:07:44
diffusion are also important to be
00:07:47
considering in these situations you can
00:07:48
also think about companies trying to
00:07:51
acquire micro chair in a certain with a
00:07:54
certain product right they do they do
00:07:57
they'll have to think about it in a
00:07:59
similar framework that with what we do
00:08:00
in this paper and that's some of the the
00:08:03
extrapolation from this paper that can
00:08:05
go into other areas
00:08:11
the actual subsidy program in Germany
00:08:13
has taken a lot of flak and in the
00:08:17
recent years for overspending at the
00:08:22
time of crisis and that's something that
00:08:27
needs to be taken with a grain of salt
00:08:29
that they have developed a huge industry
00:08:32
of solar installation and they've driven
00:08:34
down the costs of installations way
00:08:37
further down than other places like
00:08:40
California for example and if you if you
00:08:42
compare how they've done it it was
00:08:46
through strong you know time timely
00:08:49
interventions in the beginning and in
00:08:52
this paper I talked about it perhaps
00:08:54
could have been even faster in the
00:08:56
beginning and slower later on and
00:08:59
they've only started to do a more
00:09:03
aggressive phase out in the last few
00:09:06
years although they had a plan phase out
00:09:09
today they could have done a little bit
00:09:11
faster and avoided some of this perhaps
00:09:16
criticism now the main criticism is that
00:09:19
these solar panels that customers bought
00:09:21
became too profitable for them and all
00:09:23
the other ratepayers in the country are
00:09:25
going to be paying huge penalty for for
00:09:28
for the adoption of of these solar
00:09:32
panels and it's it's hard to quantify
00:09:36
how successful program is it's hard to
00:09:39
quantify what are the other benefits
00:09:43
that this industry that the merchant in
00:09:45
Germany will have right an example of a
00:09:49
country that had the not a successful
00:09:53
program was Spain which started about
00:09:56
the similar time was very aggressive
00:09:58
into the into the market perhaps in some
00:10:01
cases even more aggressive than Germany
00:10:03
they had more solar irradiation so then
00:10:08
then then Germany so could potentially
00:10:11
be more efficient installed panels in
00:10:12
Spain yet the industry was growing too
00:10:17
fast and without developing a local
00:10:20
manufacturing local technology and they
00:10:23
go
00:10:23
backed out and brought the entire
00:10:26
industry to near collapse right so
00:10:29
that's that's a clear lesson of like how
00:10:33
not to do it right as opposed to what
00:10:35
the Germany did
00:10:41
in this paper we talked about these
00:10:44
feedback loops like learning by doing an
00:10:47
information diffusion and in the next
00:10:50
few papers that we're working on we
00:10:52
explore a few different parts of the
00:10:54
market for example we look at how
00:10:57
demanding certainty will affect the
00:10:59
subsidy programs we look at how policy
00:11:03
flexibility versus commitment will
00:11:06
affect the industry all right for
00:11:09
example say your policy making you have
00:11:11
a certain target we were given deadline
00:11:14
would you like to readjust your policy
00:11:16
as you approach the target deadline to
00:11:19
make sure that you you hit your target
00:11:21
or would you like to commit the longer
00:11:23
term policy and what we actually do show
00:11:25
in that situation is that you would
00:11:27
rather commit to a longer term policy
00:11:29
because by commitment you do lose some
00:11:33
flexibility in the long run but you you
00:11:36
give a signal to the to the forms in the
00:11:40
market that they need to put in their
00:11:43
fair share and they need to absorb some
00:11:46
of the risk when you when you're in your
00:11:50
policy maker and you'll hold too much
00:11:53
flexibility on your policy and you
00:11:55
readjust your policy very often you send
00:11:59
it back bad signal to the industry right
00:12:01
so that's a new future direction of the
00:12:06
research that we are currently working
00:12:07
on
00:12:22
you
