I have discussed an issue involved in consumption voucher in another post of this blog -- how the voucher should be counted in GDP. In this present post, I would like to discuss another issue involved.
One feature of these schemes is that people are not given cash or a sum deposited in bank accounts for free disposal. The vouchers must be spent within months or their values will be lost. Another feature is that people must use electronic payment devices (Octopus or an online payment app) for the scheme. Citizens who have never used any electronic payment devices (if they use only cash or cheque) need to apply for at least one such a device then.
If you are given $5000, will you spend it? If you are given a cash or money of $5000, perhaps you will save it, not spending it. But this scheme is not about cash or money. People must spend the money within months. So, it seems obvious that the scheme will generate more consumption expenditure. Why should anyone doubt that consumption may not be stimulated by such a scheme?
But there are exactly some doubts over this issue. Economists, or someone who think like an economist, do indeed have such a doubt. They think that the answers should be sought via empirical studies. Nonetheless, if the answer is obvious, empirical studies are of secondary importance. So, why the answer is not obvious?
In fact, when the first round of the scheme was announced in 2021, a newspaper columnist had used an economics concept to predict that the vouchers are useless: people's consumption expenditure will not increase due to it, and so the scheme is simply a trouble caused to citizens.
The economics concept this columnist used is "consumption smoothing". This is not a concept that every economics student has ever learned. But the idea is not difficult: people's income may go up and down but people's consumption will not fluctuated as much as income; consumption is stable over time; people will smooth out their consumption expenditure over time as they prefer to do so.
How can people smooth out their consumption when their income fluctuate? People can do this mainly because they can save more when their incomes are high, and when their incomes go down, they use up part of their saving to keep their consumption unabated. If saving is not sufficient for smoothing out consumption, external finance (borrowing) may also be used by someone.
The columnist then analyzed the effect of voucher as this. Since people will smooth out their consumption, they will not spend more even if they are given a voucher of $5000. The voucher must be spent within months but people will buy only the goods that they originally intend to buy. Without the voucher, they simply use their own income to pay for it. With the voucher, they pay it by the vouchers. But total consumption expenditure will not be affected by the voucher.
When I read this analysis made by the columnist, I doubt if this prediction will come true. Although I know the economics concept "consumption smoothing", I doubt it has a significant effect in the issue.
A recent study shows that vouchers used via one of the electronic payment devices do increase consumption expenditure: a voucher of $5000 can generate a consumption of $5400. This empirical result shows that there is no significant substitution between payment methods in the case of voucher.
If people simply replace cash payments by vouchers, their expenditure will not increase. It is simply using one payment method instead of another. But we observe an increase in expenditure, at least for one payment method. Why do people not substitute their cash payments by vouchers? Certainly more empirical investigations are needed for answering this question. However, one can still imagine why there may be a lack of significant substitution between payment methods.
First, payment methods are different. If they are the same, they are perfect substitutes. If they are not perfect substitutes, different features (advantages, disadvantages, convenience, acceptability, and cost of using them, etc) must be involved. For example, one may find it less convenient to bring too much cash out for purchase. By contrast, electronic payments are convenient for purchase even for large-value consumption.
Second, the voucher scheme requires citizens to receive and spend the $5000 via electronic payments. The agenda behind is to promote the use of electronic payments. If electronic payments are already very popular, perhaps it does not need to be promoted. Now, if quite a number of citizens may have never used electronic payments, or rarely use them, before the voucher scheme, they are more likely those who consider electronic payments are not close substitutes for cash. If they consider both payment methods equally convenient, they may have used electronic payments before. But they haven't. So, they may consider electronic payments a method quite distinct from cash. Now, they have no choice but to use electronic payments under the voucher scheme. It is likely that they don't use it to substitute cash but simply use it for buying goods that they won't buy before.
Third, for the first round of the voucher scheme, the money will be given in two phases and the money must be spent within three to five months, depending on the phases involved. Due to the urgency of spending in short term, one may spend the money on larger-value item or on extra goods that will not be purchased without the vouchers in order to use up the voucher more quickly.
Now, it is clear that the effectiveness of the voucher in stimulating consumption depends on the design of the scheme. Suppose the voucher is simply a deposit of money in citizens' bank account. This will make the voucher a very close substitute (if not perfect substitute) for cash. In a way, voucher via Octopus is also closer to cash. If it is not due to the third point above, Octopus users may not spend more.
Let's turn back to the point of "consumption smoothing". The idea can now be expressed as this. Consumption expenditure by different types are denoted C1 and C2. If C1 and C2 are perfect substitutes, then utility is affected by C1+C2. It does not matter if C1 is higher or C2 is higher, given the same C1+C2. In the case of (perfectly) consumption smoothing, any increase in C1 will be accompanied with the same amount of reduction in C2. However, if C1 and C2 are not perfect substitutes, utility is affected by C1 as well as C2 independently, not C1+C2. Using a little bit more math notations, the model is U(C1,C2) instead of U(C1+C2). The newspaper columnist above may simply jump too soon to the conclusion that the model is U(C1+C2), not U(C1,C2).
So, the key point is not consumption smoothing. The key point is why consumption smoothing can be attained. The columnist may understand the concept of consumption smoothing very well. But the columnist may fail to figure out the reason why smoothing may or may not be attained. There must be some conditions for smoothing to be attained (or not) and attained at what degree. It is not always fully attained. The difficulty of applying economics concept is that not only one need to understand the concept well, but also need to understand the reason why the concept may or may not be applicable. This is exactly a point that is repeatedly emphasized in "hi, economics" (including, not limited to, my last post on voucher and the post on decreasing returns).
Anyway, the success of the first-round voucher scheme may be a surprise to some economists (as substitution is low in this round). Now, the second round of the scheme will be launched in April. The amount is substantially increased to $10,000. Will this increase consumption as the first round did? Well, if our above analytical framework is valid, we will find that the significance of all the three points above are weakened in the second-round arrangement. You may figure out the details of the second-round arrangement and assess how the new arrangement will weaken the three features above. Given your assessments, what will you predict?
