(This essay was published in South China Morning Post on 15 March 2017.)
The last time I commented on the budget speech of the Financial Secretary was more than two decades ago. The principles governing the postwar Hong Kong budgetary framework have been highly predictable since the time of Cowperthwaite and Haddon-Cave. It is a fundamentally sound framework that respects liberty, fosters prosperity for the population, and need not be tempered with lightly out of ignorance or prejudice.
There is of course still room for some needed revisions to the budgetary framework. Since the 1980s, the world economy has become much more integrated and as a consequence, propertied asset values have increased significantly relative to income. Accordingly, the ratios used to govern fiscal revenues and expenditures and the balance between recurrent and capital accounts should be revisited.
In Hong Kong, the stewardship of the budget has come under increasing criticism for under-forecasting annual budget surpluses and under-spending by both the left and the right of the political spectrum. Large budget surpluses have long been an embarrassment for government, but they are a result of rapidly rising land sales and conversion values and the requirement of the Basic Law to balance budgets.
The source of underestimation of government revenue by John Tsang was due mainly to errors in forecasting land sales revenues. This year there is a surplus of HK$92.8 billion compared to an initial estimated surplus of HK$11 billion becausethe government collected over HK$60 billion more than it expected in land revenues.
There is no science for making accurate forecasts with quantitative easing in the background. Who would have confidently predicted property prices would continue to trend upwards after the 2008 financial crisis? And who would have confidently predicted property prices would refuse to turn even after the imposition of repeated punitive stamp duties by the current administration?
The new Financial Secretary Paul Chan has promised to more accurately project land revenues by basing the figure on the average proportion of land revenue to the city’s gross domestic product over the past 10 years – rather than 30 years, as Tsang did.
If prices were to peak and then begin to decrease, then a 10-year approach would appear more accurate around the turning point, but subsequently would increasingly over-predict land revenues. Tsang’s conservative approach would, by the law of averages, become more accurate as the end of the trend nears and starts to turn.
At the end of the day, every Financial Secretary has to decide whether he prefers to face the political cost of making positive errors or economic cost of committing negative ones.
The new Financial Secretary Paul Chan has also proposed to set up a tax policy unit to review the revenue base. It may look at ways to create new taxes on top of the current major sources – profits and salaries taxes, stamp duty, and land sale revenue. This is a bold goal. Judging from the outcomes of past reviews, I am not optimistic that the exercise will get very far.
Many have urged the government to broaden the tax base over the years, but this is fraught with political and economic difficulties. The government has thus responded by becoming increasingly dependent on volatile land sales revenues on the capital account to fund expenditure growth. Not surprisingly, it is less resistant to pour more money into capital projects than recurrent spending because land revenues are generated in the capital account.
A more meaningful step in Chan’s review would be to look at changing how prices on new land sales and conversions are settled. At present, the full value has to be paid up at the outset into the capital account. These land values are the present (or capitalized) values of the future rental values of the property.
If government were to introduce a new “additional rates” assessed on the future rental values, then the paid-up value for new land would be lower. This would defer income received from land sales and conversions and so would not only reduce the government’s annual surplus, but also produce a more stable recurrent revenue stream to fund recurrent expenditure in a more transparent way. This does not introduce any new taxes and can be easily collected using existing methods. It would also broaden the tax base.
Such a shift might lessen the political pressure to spend huge budget surpluses on capital projects and indirectly precipitate more careful monitoring of project performance without overstretching the capacities of managers and overseers. There would also be less pressure to hide surpluses within all sorts of opaque development funds, future funds, and the rest. End users would also benefit by the relaxation of the forbidding capital market constraint they face when purchasing housing.
In sum, this arrangement would improve the budgetary planning and management process without altering the deep budgetary framework. I am not the first person to propose this. But it is time that it be taken seriously for otherwise the sound budgetary framework of the past sixty years may be seriously at risk in today’s populist political environment. Still, any attempt to radically alter a deep budgetary framework will be highly divisive and political. Tax reforms have toppled many bold governments when they least expect it.
