Does Social Security work? By that, I mean does giving elderly individuals a government pension increase their level of income, the amount of goods the can consume, or even their happiness?
An NBER working paper by Baker, Gruber and Milligan (2009) tries to answer this question in the Canadian setting.
Background
Currently, Canadian income transfer programs to seniors make up 2.3% of GDP, but this figure is expect to rise to 3.2% of GDP by 2030. Unlike Social Security in the U.S., the pay-as-you-go (PAYGO) component of the Canadian Social Security System if fairly small. Further, population growth in Canada is higher than in Europe making the old-age income transfer programs more solvent.
The Old Age Security (OAS) program is the oldest elderly income transfer program in Canada. It was enacted in 1952. Currently, “the monthly benefit paid to individuals who fully satisfied the residency requirement was $479.83. This benefit is clawed back from higher income pensioners at a 15 percent rate, starting at incomes of $60,806 (2005). Benefits are full indexed to the CPI and fully taxable under the Income Tax Act.”
The Guaranteed Income Supplement (GIS) is a means-tested income supplement for elderly individuals with low income. Benefits are taxed back at a 50% rate. The current enefit is between $370 and $570.
The Canadian Pension Plan (CPP) and Quebec Pension Plan (QPP) are finance by contributions from employers and employees. Individuals pay a 4.95% tax on earnings from $35,00 to $41,100 and benefits are based on a measure of average earnings over the individual’s working life. Participants can claim benefits at age 60, but the benefit level is increased by 0.5% per month if the benefits are claimed at an older age.
Data
The authors aim to investigate how changes in the programs described above affect the income, consumption and happiness level of individuals in different birth cohorts. These measures are taken from survey data from Statistics Canada. Income data comes from the Survey of Consumer Finances (1971-1997) and the Survey of Labor and Income Dynamics (1998-2002). Consumption data comes from the Family Expenditure Survey (1969-1996) and the Survey of Household Spending (1997-2002). Happiness is measured by the General Social Survey.
- For more details on the Canadian Social Security System, see my post from 4 July 2006.
Methodology
If there are time trends in elderly income and consumption, how does one identify the impact of the Canadian Social Security? The authors use changes in Canadian Social Security legislation to identify this impact. The regression methodology has 3 specifications:
- Regress actual retirement benefits on the dependent variables (income, consumption and happiness).
- Partial Simulation. In this approach, the authors hold constant the earnings, capital income, and family status of the individual, but allow the retirement age to vary. Benefits are based on a fixed earnings histories across all birth cohorts, not actual earnings.
- Full Simulation. In this case, earnings, capital income, family status, and retirement age are held constant and the authors calculate simulated benefits levels based on an average earnings histories and retirement ages across all cohorts.
Results
In general, the authors find that a higher Social Security benefit increases elderly income. In the full simulation and when simulated benefits are used as an instrument for actual benefits, Social Security income benefits increase elderly income benefits dollar-for-dollar. Further, elderly income poverty decreased significantly when more generous benefits were enacted; the authors claim that 96% of the reduction in elderly poverty is due to these added benefits. However, in the partial simulation methodology, the authors find that a $1 increase in benefits leads to only a $0.55 cent increase in elderly income, thus indicating significant crowd-out.
For consumption, the authors also find that more generous income benefits increase elderly consumption levels, but not dollar for dollar. A $1 increase in benefits leads to a $0.66-$0.80 increase in consumption, thus indicating some crowd-out. Unlike for the case of elderly income poverty, more generous Social Security benefits did not affect consumption poverty. Thus, it may be the case that poor elderly individuals have other sources of consumption (e.g., purchase by family members, unreported income gifts from family members, unreported labor income) that may offset lower government supplied income benefits.
“For the very happy question, we see no sign of a statistically significant relationship between benefits and being very happy. On the other hand, there is some evidence of a decrease in reports of being unhappy or very unhappy with higher benefits in the reduced form results, but not in the IV results.”
Healthcare Economist’s Take
This paper indicates that more generous income benefits do increase income and consumption for the elderly. Social Security benefits create more crowd-out in the case of consumption than income. It is likely that consumption is a better indicator of well-being, especially since elderly savings is very low (i.e., the elderly are generally spending down their assets than building them up). While income poverty declined due to these income benefits, consumption poverty did not. Overall, I would say that these program do help increase elderly well-being, but likely not dollar for dollar. As the authors note, this paper only looks at the benefits of Social Security without taking into account the costs of raising a significant amount of revenue to pay for these government program.
I will not comment on the happiness measure. Although many people may think it is the government’s job to make individuals happier, I believe that happiness is determined on an individual level and often based on things the government can’t control (e.g., do you get along with your spouse, has their been a death in the family). Further, happiness is often measured by comparing your emotional feelings against some status quo. Thus, a poor family would be very happy to have their income increased to $80,000, but a millionaire would be very disappointed.
I think this paper makes an important contribution showing that government old age income benefits do increase income and consumption, even if there is some crowd out on the consumption side. The question for me is less whether or not there should be some government benefit, but more about how generous it should be and whether it should serve only the poor or all elderly.
Recent comments