Supplanting salaries tax with GST will not boost revenue stability
LET ME DO the plaudits first. The government's consultation paper on the proposed goods and services tax is a model of good structure and of concise, lucid writing.
Unfortunately, there are a few flaws at the heart of its reasoning, in particular a critical assumption that was never examined but should have been. This new tax may not do much at all to reduce the volatility of the existing revenue base.
Let us start at the beginning. The problem for our government is that its expenditures run at a relatively consistent level from year to year but its revenues do not. They are highly sensitive to the more cyclical elements of the economy.
There is first of all the heavy dependence on land sales. They balance out over the longer run but they go way up in good times, way down in bad times, as does the profits tax revenue. Meanwhile, only a small proportion of the population pays any salaries tax and this tax is heavily weighted to higher incomes that also go up and down wildly with economic fortunes.
Would it not be better to have a more stable revenue base and thus avoid constant swings from big fiscal deficit to surplus?
Yes, say an increasing number of governments around the world and so says our government now, too.
The obvious solution is the GST. It is widely based, easier to collect than many other forms of tax and, best of all, most people have steady spending requirements. Tax them on what they spend rather than on what they earn and revenues will become more stable.
The obvious argument against GST is that it is regressive. It puts proportionately more of the tax burden on the poor and less on the rich than our existing tax base does. The authors of the consultation paper recognise this difficulty and try to address it, although not very effectively.
But let me leave that for another day. What concerns me immediately is the government's very laudable emphasis on introducing GST only to supplant other forms of revenue, not to raise more money.
It proposes some immediate cuts in other revenue sources. Aside from compensating lower-income households for their increased tax burden, it wants to use some of the $30 billion annually, which a 5 per cent GST would bring in, to reduce such things as car taxes and fuel and alcohol duties.
At the end of this we will have about $20 billion in GST revenues left and the consultation paper at this point poses a question to us: How would you want to use the money?
It lists three possible options - (1) reduce salaries tax, (2) reduce profits tax and (3) increase government expenditure.
We can dispense with options 2 and 3 immediately. Tax the poor to help the rich pay less in corporate tax? May I live to see the day when Legco would nod this suggestion through.
Similarly, there are likely to be few votes for more government spending with the money. Leave aside the pledges that rule this out. Anyway, one of the big dangers of GST is that it is so easy to raise GST rates. Ask the taxpayers of Sweden about it. The point is to supplant other taxes, not encourage government profligacy.
It will have to be reductions in salaries tax for which we use the bulk of the money and we now come to the crucial flaw. Is income from salaries tax really so volatile in Hong Kong that supplanting it with a GST would reduce the volatility of government revenues?
The chart gives you your answer. The red line shows you salaries tax revenue as a percentage of total spending on the consolidated account since 1990. The blue line takes official figures on consumer spending in the domestic market and shows you what the equivalent figure from a 5 per cent GST would probably have been over that period.
The fact is that both lines on the chart go up and down pretty wildly. The taxable level of salaries may decline sharply in bad times but so does personal spending and they are both bad matches for government spending.
We have assumed that GST is more stable than salaries tax because other governments say it is in their own countries and we have not sufficiently questioned this assumption.
In Hong Kong it may simply prove not true.
Meanwhile, if we use the GST revenue to supplant salaries tax revenue we will have just as much volatility as we have always had in profits taxes and land sales revenues. We will not have helped ourselves much at all to achieve a stable revenue base.
Why then a GST?