馃攳
What Happens When You Inject $11 Million into a Small Economy? - YouTube
Channel: econimate
[9]
What happens when we inject cash into
the economy? Does this translate
[15]
one-for-one into aggregate measures such as GDP? Or is there a multiplier effect?
[20]
In other words, does that initial
infusion get magnified by some factor as
[26]
it filters through the economy?
Estimating these multipliers has been of
[30]
key economic interest for many decades,
and they are crucial for evaluating
[35]
policies aimed at stimulating the
economy. This paper charts new territory
[40]
by using a large-scale experiment to
measure this key quantity. Between 2014
[46]
and 2017, the authors worked with the NGO GiveDirectly and provided one-time cash
[53]
transfers of approximately 1,000
US dollars to over 10,500
[59]
poor households in rural Kenya.
The magnitude of this intervention was
[63]
substantial. $1000 is equivalent to
75% of average annual
[69]
expenditures among recipient households, and, in total, the transfers amounted to
[74]
roughly 17% of local GDP. The
question of this paper is simple: what
[81]
happened? Can we trace the effects of
this large infusion of cash as it
[86]
percolated through the economy? This type of large-scale cash transfer program has
[91]
become more common in developing
countries, and many researchers have
[95]
studied their effects on recipients. But
what sets this paper apart is the wider
[100]
focus: the key question here is not how
the recipients themselves were impacted,
[105]
but how the ripple effects spread
through the rest of the economy. To gauge
[111]
these spillovers, the transfers were
randomized in the following way. First,
[116]
sublocations in the study area, which
was located in a rural area in western
[121]
Kenya, were randomly assigned to be
either high or low saturation. In high
[127]
saturation areas, 2/3 of villages
were randomly assigned to the Treatment
[132]
group, with the rest in the Control group. The
[135]
proportions were reversed in low
saturation areas: there, 1/3 of
[140]
villages were Treatment and 2/3
were Control. Finally, within each
[146]
Treatment village, all eligible
households received the transfer while
[150]
no households in Control villages did.
This design allows the authors to use
[156]
geographic variation in exposure to the
transfers. In addition to random
[161]
variation in their own Treatment status,
some villages happen to be in areas
[166]
where many of their neighbors receive
transfers, while others are not. This
[171]
allows the authors to study spillovers
that occur across as well as within
[175]
villages. To start, the authors show
there are large positive impacts on
[181]
households that received the transfers.
Recipients spend most of the cash they
[186]
receive, leading total expenditures to
increase by 13.5%.
[191]
This increased spending is spread across
a broad range of items, and much of it
[196]
takes place locally. So a natural first
step in tracing the spillovers is to
[202]
look at local enterprises. And the
authors find that firms located in areas
[208]
with more transfers experience large
increases in revenue, with the magnitudes
[213]
ranging from 29 to 45%. Notably, those
increases occur regardless of whether or
[221]
not the firm's owners themselves
received a transfer, and appear to be
[225]
passed on in the form of higher wages to
firm employees. This, in turn, raises the
[232]
question of how non-recipient households
are affected by the transfers. And the
[237]
striking result here is that, despite not
receiving any transfers themselves, over
[243]
the 18 months following the intervention,
these households show consumption gains
[248]
of 13% as well, an amount roughly
comparable to the immediate gains
[253]
experienced by the Treated households.
The authors also show that there is only
[258]
minimal price inflation, indicating that
these results point to real, and not just
[263]
nominal, gains for non-recipients. So, where do these gains
[269]
come from? Are non-recipients running
down their savings in order to keep up
[274]
with their neighbors' increased
consumption? Or are they receiving money
[278]
from friends who qualified for the
transfers? The authors show that neither
[282]
of these can explain the results. Instead,
higher consumption by non-recipients is
[288]
driven by higher earned income. Thus, the
chain of events is the following:
[294]
recipient households use their transfer
to buy more goods from local enterprises;
[299]
this increases firm revenues and gets
passed on in the form of higher wages;
[304]
those higher wages increase the incomes
of non-recipients employed at these
[309]
firms; and that higher income allows non-recipients to increase their own
[314]
consumption. This is the multiplier in
action, and illustrates how an initial
[320]
infusion of cash can be magnified and
multiplied many times over as it
[325]
percolates through the economy,
ultimately generating aggregate effects
[329]
that exceed to the size of the original
stimulus. In this particular setting, the
[334]
authors estimate a multiplier of 2.7: in
other words, a transfer of $1 to the
[340]
local economy gives rise to an
additional $2.70 in
[345]
increased economic activity. In
conclusion, this paper provides some of
[352]
the first experimental estimates of this
important economic parameter, and maps
[357]
out the widespread aggregate effects of
a large-scale cash transfer program. The
[363]
results show that focusing only on
recipients may provide a limited window
[367]
into these programs. Though transfers may be targeted to only a subset of
[372]
households, their effects ultimately
reverberate throughout the local economy.
[376]
As always, you can check out the full
paper and its references to other
[381]
related research. These include a macroeconomics literature on multipliers, a
[386]
development literature studying the
effects of cash transfer programs, and
[391]
finally a broader debate on the role of
randomized controlled trials in
[395]
economics and the kinds of questions
they can
[398]
help us answer.
Most Recent Videos:
You can go back to the homepage right here: Homepage