The government’s proposal for a more generous means tested Old Age Living Allowance (OALA) has bogged down in the Finance Committee of the Legislative Council following filibuster tactics by certain legislators who claim the proposal is not generous enough. Some say they would only support a less stringent means test, while others are demanding a universal pension scheme.
Unfortunately, no one is objecting to the OALA on the grounds that it would lead to higher future taxes, could have negative incentive effects on economic growth, and might eventually lead to chronic fiscal deficits as populist political pressure to enhance OALA payments grows in the future. I am concerned about the rising prospects of such developments, but there is no one in the legislature who speaks for my worries. Fiscal irresponsibility in the legislature is an important concern as Hong Kong gravitates toward a more open political system.
Where does Funding Come From?
Table 1 gives the government’s estimates of future public expenditure under the proposed OALA. The estimates are based on new population projections to the year 2041 and show that additional OALA expenditures under the proposed means tests would increase from $6.2 billion in 2012 to $21.3 billion in 2041. If income and asset declarations were waived for elders aged 70 or above, the additional OALA expenditures would increase from $9.9 billion in 2012 to $26.9 billion in 2041. If it were waived for elders aged 65 or above, the additional expenditure would increase from $13.6 billion in 2012 to $35.1 billion in 2041.
| Table 1: Government Estimates of the Long-Term Financial Implications of Old Age Living Allowance (OALA) (HK$ billions) | ||||||
|
Year |
Original Social Security Allowance |
Additional OALA expenditure according to Government’s proposal |
Additional OALA expenditure |
Additional OALA expenditure |
||
|
2012 |
$7.8 |
+$6.2 |
+$9.9 |
+$13.6 |
||
|
2017 |
$9.3 |
+$7.7 |
+$11.8 |
+$16.8 |
||
|
2022 |
$11.7 |
+$9.6 |
+$14.6 |
+$21.0 |
||
|
2027 |
$14.7 |
+$11.9 |
+$18.4 |
+$26.1 |
||
|
2032 |
$17.7 |
+$13.9 |
+$22.0 |
+$30.7 |
||
|
2037 |
$20.0 |
+$15.5 |
+$25.1 |
+$34.1 |
||
|
2041 |
$21.3 |
+$16.2 |
+$26.9 |
+$35.1 |
||
Source: Finance Committee Paper (FCR-2012-13-54), 9 November 2012.
Unfortunately these estimated expenditures have not been supplemented with an explanation of how they will be financed. Table 2 shows the total population will increase from 7.13 million in 2012 to 8.47 million in 2041. During this period the potential working population of adults between the ages of 20 and 64 will be largely unchanged, going from 4.93 million in 2012 to 4.82 million in 2041. However, the total elderly non-working population aged 65 and above will increase from 0.98 million in 2012 to 2.56 million in 2041. In other words, while 5 working people are potentially supporting every elderly person in 2012, by 2041 the number will fall to 1.9 people. At the same time, the working population will also have to support young dependent children, whose numbers will be relatively stable over the period of projection, falling from 1.23 million in 2012 to 1.09 million in 2041.
Table 2: Projected Population 2012-2041 (in millions)
|
Year |
(A) Total Population |
(B) Age 0–19 |
(C) Age 20–64 |
(D) Age 65+ |
(E) (C)/(D) |
(F) (C)/(B) |
|
2012 |
7.13 |
1.23 |
4.93 |
0.98 |
5.0 |
4.0 |
|
2017 |
7.43 |
1.15 |
5.07 |
1.21 |
4.2 |
4.4 |
|
2022 |
7.72 |
1.17 |
5.03 |
1.52 |
3.3 |
4.3 |
|
2027 |
7.99 |
1.22 |
4.88 |
1.88 |
2.6 |
4.0 |
|
2032 |
8.2 |
1.16 |
4.83 |
2.21 |
2.2 |
4.2 |
|
2037 |
8.37 |
1.23 |
4.81 |
2.44 |
2.0 |
3.9 |
|
2041 |
8.47 |
1.09 |
4.82 |
2.56 |
1.9 |
4.4 |
One simple solution for financing the additional OALA expenses is to assume that government revenue will increase with economic growth. If we assume that the entire composition of government revenue remains unchanged and total revenue increases proportionately with the growing economy, then we can compute what increases will be required in the annual GDP growth rate per working population to pay for the additional OALA expenditures. Business and individual income tax rates would be unchanged under this scenario. Table 3 shows the required growth rates. If the government’s OALA proposal with means tests is adopted, the annual real GDP growth rate per working population would have to be between 1.7% and 2.4% for different years in the period 2012-2041. If the means tests were to be abolished for the elderly population aged 70 and above, then the growth rate would have to increase to between 2.6% and 3.3%. And if no means tests were applied to the population aged 65 and above, then the growth rate would have to increase to between 3.1% and 4.0%.
Table 3: Annual Real GDP Growth Rate per Working Population Required to Finance Government Projected Additional OALA Expenditures
|
Year |
Original Social Security Allowance |
Additional OALA expenditure according to Government’s proposal |
Additional OALA expenditure |
Additional OALA expenditure |
||
|
2017 |
0.0% |
1.7% |
2.6% |
3.8% |
||
|
2022 |
0.0% |
2.0% |
2.9% |
3.9% |
||
|
2027 |
0.1% |
2.2% |
3.1% |
4.0% |
||
|
2032 |
0.0% |
2.4% |
3.3% |
3.7% |
||
|
2037 |
0.0% |
2.4% |
3.2% |
3.4% |
||
|
2041 |
0.0% |
2.3% |
3.0% |
3.1% |
||
Will it be possible for Hong Kong to achieve these real GDP growth rates per working population into the future? Table 4 shows that the rate was 4.1% in the period 1961-2011, and 4.7%, 0.9% and 4.0% for the sub-periods 1961-1997, 1997-2003, and 2003-2011, respectively. If past growth rates can be sustained into the future then the proposed OALA can be financed in principle. This conclusion would apply even if we do not require means tests. But we must also assume that there are no other major competing demands for public expenditures and future OALA benefit levels would be adjusted only for inflation.
Table 4: Population, Working Population, and Real GDP (Annual Percentage Growth)
|
1961-2011 |
1961-97 |
1997-2003 |
2003-11 |
2012-41 |
|
| Real GDP per working population |
4.1 |
4.7 |
0.9 |
4.0 |
4.0* |
| Real GDP |
6.4 |
7.5 |
1.6 |
5.0 |
4.1* |
| Real GDP per capita |
4.7 |
5.4 |
0.9 |
4.4 |
3.4* |
| Total population |
1.6 |
2.0 |
0.8 |
0.5 |
0.6 |
| Working population |
2.2 |
2.7 |
0.7 |
1.0 |
0.1 |
Note: Asterisked figures are estimates assuming real GDP per working population will grow at 4.0% annually.
Inevitable Increase of Tax Rates
However, the assumption that future OALA benefit levels would only be adjusted for inflation is extremely unrealistic and optimistic. Although at present the Social Security Allowance is only adjusted for inflation to protect its real purchasing power, it is inconceivable that such an arrangement would be acceptable to an increasingly populist legislature facing a growing proportion of elderly voters. Surely OALA expenditures will have to rise over time with the standard of living.
What might this standard be? The recently established Commission on Poverty has been charged with the task of drawing a “poverty line”, and it could define it based on either an absolute or relative standard. An absolute standard would benchmark the “poverty line” against a bundle of goods and services to protect benefit levels against inflation, but unlike a relative standard it would not incorporate improvements in the standard of living when real GDP per capita rises over time. A relative standard would draw reference to its position in the distribution of income. Oxfam, for example, has issued a position paper advocating the adoption of a relative standard based on one-half of the median household income.
At present the Social Security Allowance is in theory primarily adjusted with reference to an absolute standard. The government has proposed that the OALA also be adjusted in this manner. If Hong Kong were to adopt a relative standard in the spirit of the Oxfam proposal, then the estimated OALA expenditure by government in Table 1 would become seriously underestimated. I mention all this because any decision on the definition of the “poverty line” by the Commission on Poverty will naturally apply to the adjustment of future OALA benefit levels.
Here is what would happen if the OALA were adjusted with reference to a relative standard based on per capita GDP growth. Table 5 presents my estimates of OALA expenditures using the assumption that the economy will grow by an implied annual rise in real GDP growth per working population of 4.0% for the period 2012-2041. Given the projected population structure, the real GDP per capita will therefore rise at 3.4% per annum. As a result, OALA benefit levels would be adjusted at this rate to more or less fully accommodate improvements in the standard of living.
Table 5: Estimates of the Long-Term Financial Implications of Old Age Living Allowance (OALA) (HK$ billions)
|
Year |
Original Social Security Allowance |
Additional OALA expenditure according to Government’s proposal |
Additional OALA expenditure |
Additional OALA expenditure |
|
2012 |
$7.8 |
+$6.2 |
+$9.9 |
+$13.6 |
|
2017 |
$10.7 |
+$8.5 |
+$13.3 |
+$20.3 |
|
2022 |
$15.9 |
+$12.7 |
+$19.9 |
+$29.9 |
|
2027 |
$23.5 |
+$18.7 |
+$29.6 |
+$43.4 |
|
2032 |
$33.5 |
+$26.5 |
+$42.6 |
+$58.8 |
|
2037 |
$44.8 |
+$35.3 |
+$57.5 |
+$75.2 |
|
2041 |
$54.3 |
+$42.6 |
+$69.9 |
+$89.4 |
It is apparent that these estimates are much larger than those that the government provided to the Finance Committee. In 2041, the government estimates that OALA expenditures with means tests will only be about $16.2 billion, whereas my estimates put the figure at $42.6 billion, which is 285% higher. A similar gap will exist if means tests are waived for the elderly population aged 70 or above ($26.9 billion under the government’s estimates, $69.9 billion under mine, representing a 260% difference), and if means tests are waived for all those aged 65 and above ($35.1 billion under government estimates, $89.4 billion under mine, representing a 255% difference).
A 4% annual growth rate in real GDP per working population would not be able to cover such an increase in OALA expenditure. It then would be necessary to raise additional revenues through, for example, increasing taxes on business and personal income.
If the OALA were to be financed entirely by increasing such taxes, and assuming that higher taxes will not result in negative incentive effects, then the amount of OALA induced taxes in 2041 would have to increase by 7.79% (see Table 6). This is based on the government’s proposal with means tests in place. If means tests were waived then the additional tax revenues would have to be increased by 12.78% and 16.33% depending on whether they were waived for the elderly population aged 70 and above or aged 65 and above, respectively. Whatever the option for OALA expenditures, tax rates will have to be increased. Businesses will have to pay higher rates and more working individuals will have to be drawn into the personal income tax net.
Table 6: Estimates of the Percentage Increase in Business and Personal Income Taxes Required to Fund Old Age Living Allowance (OALA)
|
Year |
Original Social Security Allowance |
Funding OALA expenditure according to Government’s proposal |
Funding OALA expenditure |
Funding OALA expenditure |
||
|
2017 |
0.37% |
4.10% |
6.40% |
9.72% |
||
|
2022 |
0.48% |
4.97% |
7.80% |
11.72% |
||
|
2027 |
0.55% |
6.01% |
9.49% |
13.93% |
||
|
2032 |
0.53% |
6.96% |
11.17% |
15.44% |
||
|
2037 |
0.48% |
7.56% |
12.33% |
16.14% |
||
|
2041 |
0.43% |
7.79% |
12.78% |
16.33% |
||
Politicians Ignore Fiscal Prudence
Given that the current tax rates on business income is 16.5% and the standard tax rate on personal income is 15%, funding the proposed OALA expenditures would require government to increase tax rates. Assuming total taxes received continue to form 9% of GDP as they do now then the tax rates will have to increase by 1.2% to 2.7% depending on whether means tests would be required and how they would be applied. These increases do not factor in the disincentive effects on investment and work. The actual increases in tax rates would likely be even higher. Moreover, competing demands for public expenditure would also likely arise due to the growing public health care expenses of an ageing population.
A generous OALA is a policy mistake of the highest order. An OALA without means testing would be a policy disaster.
There have always been social advocates for universal social pensions. Fifty years ago this was perhaps a passionate ideology held by misguided but sincere idealistic advocates. Today there is overwhelming evidence that pay-as-you-go universal social pensions saddle governments with crippling pubic debts. Those who continue to advocate such proposals can only be described as either irresponsible opportunists or defenders of the status quo. These governments are technically insolvent and morally bankrupt. Politicians procrastinate over taking the necessary steps to restore fiscal prudence. They dare not tell their constituencies the simple truth that they cannot retire at 65 and start collecting their pension checks, but have to work until 70.
Pay-as-you-go universal social pensions introduced with the best of intentions inevitably bring society to a moral cliff of irresponsibility and lies. They represent a system of spend now and tax later. It makes respectable the unholy task of burdening future generations with higher taxes to pay for the excesses of the current generation.
Since future generations are not present to vote for such largesse, they cannot exercise their choice through the ballot box. They have to rely on the current generation to act in a responsible altruistic manner to safeguard their interests. Unfortunately populist democracy has failed completely to prevent irresponsible behavior of this sort. The pay-as-you-go universal social pension is the great vehicle through which intergenerational responsibility and altruism is severed. And in so doing it renders the populist democratic political system deeply flawed.
Damage to Cross Generation Obligations
The family as the oldest civic association has been a major victim of the pay-as-you-go system of universal social pensions. Throughout history children have been an important source of material and emotional support to their parents in old age. The family as an institution has been a central vehicle by which intergenerational transfers are organized. The family organization is supported by social norms and informal incentives that approve those who fulfill intergenerational commitments and sanction those who fail to do so. The intrinsic family values are love and altruism for a good reason – they are the contractual glue that binds together different generations connected by blood.
Once the government takes over this task it competes with the family for loyalty across generations. It inevitably weakens the family as parents withdraw resources for the support of their children and vice versa. It is indeed shocking to hear some politicians say that the purpose of universal social pensions is to help the young generation so that they do not have to pay for the support of their parents.
The family has survived for thousands of years as the basic institution to organize intergenerational transfers in the absence of government. Replacing the family with the government is a very risky undertaking for society. If government fails, it fails everyone. As a decentralized arrangement, the family provides society with much greater scope to diversify the risks of intergenerational contractual failure. The histories of modern populist democracies have shown that governments are irresponsible adopted parents and untrustworthy adopted children when it comes to making intergenerational commitments stick.
There is a good reason why the government’s proposed means tested OALA can make sense for Hong Kong. Life expectancy of the elderly has increased significantly for this generation. Some of the elderly poor might not have been able to fully anticipate such a development. This was not their fault and their children might not have the resources to help them adequately. There is a legitimate case for helping them, but such a case should rest on performing means testing. There is no loss of dignity for these elderly poor and their children to conduct means tests for it is not their fault that they failed to anticipate their plight fully.
The government should make clear that it has no intention of adopting a universal social pension scheme and rebuke those who continue to advocate such irresponsible actions that harm society’s intergenerational compact. Unfortunately the government in a bid to appease populist pressure from the legislature is suggesting that the introduction of the OALA will not prevent the government from considering a universal social pension scheme in the future. This is wrong.
