Much ado about nothing? Demographic bulges, the productivity puzzle, and CPP reform.
The future of the Canada Pension Plan, financed on pay-as-you-go principles, is far from certain. Policymakers often express the opinion that as members of the baby boom generation approach retirement, the CPP will not be financially or politically sustainable in its current form because the costs of providing promised CPP benefits to this large cohort will be borne by a much smaller trailing cohort. The smaller cohorts may not be able to pay, or may choose not to honor, the levels of benefits promised by the current benefit formulas. These forecasts of demographically driven financial imbalances have encouraged debates over how to reform the existing pay-as-you-go CPP arrangement so that it will be sustainable. Proposed reforms for the CPP primarily involve reducing benefits provided and raising the taxes levied on labor income that finance the system in order to pre-fund future benefits. However, the underlying assumptions of these forecasts may be flawed, and the proposed reforms based on them may be misguided. In this paper we demonstrate in an overlapping generations framework how demographic factors not only affect the taxes required to finance the prescribed benefits, but may also affect wages. While there may be fewer workers supporting more retirees in the future, the incomes of those workers will likely rise substantially over the same time period, leaving them much better able to afford the required taxes.
Discussions of the impending crisis for the CPP focus on the increase in wage taxes necessary to pay the benefits promised under current CPP benefit formulas. A contribution rate of 3.6% prevailed from the introduction of the CPP in 1966 to 1986. Over the same period, the generosity of CPP benefits, and the extent of CPP coverage, increased. By 1993, the CPP contribution rate had risen to 5% and is expected to increase to 14.2% by 2030 (Government of Canada, 1996). This increase in taxes is troubling if real wages of workers trailing the baby boomers do not increase. However, if real wages rise in the future, then there will be a greater capacity to pay CPP benefits at the higher contribution rates. For example, Pesando (1993, p. 11) points out that if productivity "increases by just 1% per year, then by 2030 real income in Canada will be almost 50% higher than it is now, other things being equal. The capacity for the economy in 2030 to sustain the level of promised [CPP] pension benefits will be enhanced accordingly."(1) Thus, it is not clear that higher contribution rates for the smaller post-baby boom cohort to fund the currently "promised" Canada Pension Plan benefits mean that the pay-as-you-go arrangement is unsustainable, nor even that the following cohort will be worse off than the baby boom cohort.
Any conclusions we make about the future of a pay-as-you-go financed CPP depend crucially on what we believe about future productivity/real wage growth for another reason. Pay-as-you-go finance for social security arrangements like the CPP makes sense when the growth in real wages exceeds the real interest rate. The Federal/Provincial Information Paper on the CPP (Government of Canada, 1996, p. 19) explains that a reversal in this relationship over the last 30 years is the driving force behind the desire for reforming the CPP by raising contribution rates to pre-fund CPP benefits:
In the 1960s and 1970s, the growth in real wages and salaries was very high - 5.1% and 4.8%. It exceeded the real rate of interest. By the 1980s and continuing into the 1990s, this situation had reversed. Interest rates were higher than the growth of wages. Today, it would be imprudent to assume any change in that relationship for the foreseeable future. (emphasis added)
Several studies that link demographic forces and the slowdown in real wage growth and the rise of real interest rates provide a serious challenge to the belief that real interest rates will remain higher than real wage growth rates (e.g., Welch, 1979; Easterlin, 1980; Stapleton and Young, 1984; Berger, 1985; Yoo, 1994; Foot and Stoffman, 1996). Since 1973, real wages and labor productivity have stagnated in North America after two decades of rapid growth. Coincident with the productivity slowdown, members of the baby boom generation entered the labor force. Not only were more males entering the labor force, but women of the baby boom generation were more likely to participate in the labor market than women had been in the past. Foot and Stoffman (1996, pp. 199-200) state the intuition for the relationship between these developments:
In a society with a lot of young workers, labor is in abundant supply and is cheap... In a predominantly young country, because only a small section of the population has any savings, capital is priced high because it is in short supply... That was the situation in Canada during the two decades from the mid 1960s to mid 1980s, when the huge baby-boom generation was moving onto the labor market and becoming borrowers... In an aging society, the growth of the workforce slows down. Meanwhile, because more people are older, they have more money to invest, and the supply of capital increases. Because of demographic change, wages ultimately rise while interest rates come down.(2)
Similarly, Summers (1993) estimates that the increase in the number of old people in OECD countries over the next 40 years will result in a decline of per capita real income of about 10%(3) but will raise the productivity of the remaining workers since fewer workers will be using "the given capital stock."
Tempering these effects are the potential reductions in savings because of the pay-as-you-go arrangement itself, which transfers income from young "savers" to old "consumers." Such transfers reduce the rate of capital formation, which, in turn, slows productivity and real wage growth (Kotlikoff, 1996). Obviously the magnitude of this effect increases as the dependency ratio rises. If the slowing effect on productivity of reduced savings dominates, then, as Kotlikoff (1996) argues, young people face the double "whammy" of low wages and high tax rates. Thus, whether or not productivity and real wages rise as the baby boomers retire cannot be predicted a priori.
The goal of this paper is to examine in an overlapping generations model how demographic factors on their own can explain much of the observed changes in productivity/wage growth of the last 30 years. The authors also examine the impact of demographic factors on a pay-as-you-go financed CPP. In a closed economy setting, cohorts following the baby boom cohort have strictly higher consumption and utility in both periods of their lives than do the members of the baby boom cohort. In an open economy setting, real wages and interest rates do not adjust to changes in population size due to perfect capital mobility. In this setting, members of the large baby boom cohort have higher consumptions in both periods of their lives than do members of all other cohorts. The smaller cohort trailing the baby boom has strictly lower consumption in both periods of their lives than do members of all other cohorts. Thus, the appropriate reforms, or lack of reforms, for pay-as-you-go social security arrangements clearly depend upon whether the appropriate case is the open or closed economy case. If Canada behaves like the small open economy model, then increasing payroll taxes now to at least partially fund the CPP benefits transfers the burden of finance away from the lower income baby bust generation to the relatively higher income baby boom generation. In contrast, if Canada can be characterized by the closed economy case, then increasing payroll taxes now to keep taxes lower in future is inter-generationally regressive as the burden of CPP finance is reduced for the well off baby bust generation and passed to the lower income baby boomers.
II. AN INTERGENERATIONAL TRANSFER MODEL
This section presents an overlapping generations model that includes a simple pay-as-you-go system of social security and allows us to consider the general equilibrium dynamics of the baby-boom on the social security system. This builds on previous work that analyzes either the dynamics using a partial equilibrium framework (see Richardson, 1994; Wolfson and Murphy, 1996) or the steady-state in a general equilibrium framework (see Kotlikoff, 1979; Burbidge, 1983). The model in this paper is a special case of Burbidge's (1983) model and is much in the same spirit as Yoo (1994). Yoo, however, does not include a social security system.
Agents born at any given time t, live for two periods; t and t+1. At any time t, the population consists of [N.sub.t] young agents born at time t and [N.sub.t-1] old agents who were born in the previous period, t-1.
One good is produced in this economy. Consumers derive utility from consuming this good in both their young period and their old period. Denote the consumption of each member of generation t in period t as [c.sub.t](t) and in period t+1 by [c.sub.t](t + 1). Young consumers supply one unit of labor inelastically in period t for which they receive a wage of [w.sub.t]. Thus, their income in time t is equal to the wage rate [w.sub.t]. At time t+1, the old period, they have no labor to supply. Thus, in addition to financing consumption while they are young, consumers in their first period must also save a portion of their income, [s.sub.t](t), for consumption when they are old. Finally, young agents pay a social security contribution, [[Theta].sub.t][w.sub.t](0 [less than or equal to] 0 [less than or equal to] 1). A young agent's budget constraint at time t is:
(1) [c.sub.t](t) + [s.sub.t](t) + [[Theta].sub.t][w.sub.t] [less than or equal to] [w.sub.t]
Old consumers have two sources of income for financing consumption in the second period of life, [c.sub.t](t + 1). The first is their savings. In the second period, this capital earns a return equal to [r.sub.t+1]. Their second source of income is the social security payment denoted as [[Tau].sub.t+1]. Thus, the second period budget constraint for a consumer of generation is:
(2) [c.sub.t](t + 1) [less than or equal to] [[Tau].sub.t+1] + [r.sub.t+1][s.sub.t](t)
Each consumer chooses [c.sub.t](t), [c.sub.t](t+1), and [s.sub.t](t) to maximize a utility function that for analytical convenience is given by:
(3) U([c.sub.t](t), [c.sub.t](t + 1)) = log([c.sub.t](t)) + log([c.sub.t](t + 1))
This problem produces the first order condition:
(4) [c.sub.t](t + 1) / [c.sub.t](t) = [r.sub.t+1]
which can be solved to yield
(5) [Mathematical Expression Omitted]
Consumption in either period can now be determined by the budget constraints.
There is one firm in the economy with production technology given by:
(6) [Mathematical Expression Omitted]
where [y.sub.t] is total production at time t, [l.sub.t] is the labor used at time t, and [k.sub.t] is the total capital stock at time t. In equilibrium all factor markets clear so [l.sub.t] = [N.sub.t] and [k.sub.t] = [N.sub.t-1][s.sub.t-1](t-1). With constant returns to scale technology the capital and labor inputs are paid their marginal products so that:
(7) [Mathematical Expression Omitted]
(8) [Mathematical Expression Omitted]
Substituting in the equilibrium conditions, we have the following two difference equations:
(9) [Mathematical Expression Omitted]
(10) [Mathematical Expression Omitted]
equations (9) and (10) describe the factor markets that provide income for consumers.
C. Social Security System
The pay-as-you-go social security system in this model is an income redistribution system, In a pure pay-as-you-go plan, benefits paid at a given time t are financed through revenues at time t. Each young worker at time t pays taxes [[Theta].sub.t][w.sub.t]. Tax revenues are distributed amongst old consumers at time t as a lump sum benefit [[Tau].sub.t]. The balanced budget condition is:
(11) [N.sub.t][[Theta].sub.t][w.sub.t] = [N.sub.t-1] [[Tau].sub.t]
Since we assume that the payments [[Tau].sub.t] are set outside the model, the tax rate [[Theta].sub.t] changes from period to period to balance the pension plan's budget. The current period tax per person must satisfy the following equation:
(12) [[Theta].sub.t] = [N.sub.t-1] / [N.sub.t] [multiplied by] [[Tau].sub.t] / [w.sub.t]
Equation (12) closes the model. Solving (9), (10), and (12) for [w.sub.t], [r.sub.t], and [[Theta].sub.t] determines all of the remaining elements of the model.
D. Comparative Dynamics
This subsection examines the comparative dynamics of the model to determine the economic impacts of changes in the age distribution of the population. The objective is to determine the signs of the derivatives [Delta][[Theta].sub.t]/[Delta]D, [Delta][w.sub.t]/[Delta]D, and [Delta][r.sub.t]/[Delta]D where D = [N.sub.t-1]/[N.sub.t] is the ratio of old consumers to young consumers. The signs of [Delta][[Theta].sub.t]/[Delta]D and [Delta][w.sub.t]/[Delta]D are indeterminate. The reason for this indeterminacy is that the required change in [[Theta].sub.t] depends on whether or not [w.sub.t] increases or decreases. The change in [w.sub.t] in turn depends on the elasticity of saving with respect to interest rates. Thus, it is by no means certain that the retirement of the baby boom generation will produce the commonly anticipated pressure on CPP tax rates.
This section presents several simulations of the model. The discussion focuses on two limiting cases of our modeled economy. First, we examine a closed economy representation of the model, followed by an open economy representation (perfectly elastic supply of capital). For each scenario we present two sets of results. The first is for a no social security economy. In this case, [Tau] = 0 and [[Theta].sub.t] = 0 for all periods, and the only source of retirement income is private savings. The second case has [Tau] = 1. In this case, the social security tax rate must be determined in every period by solving equation (12).
A. Closed Economy
We simulate the model under the assumption that domestic savings provide the only source of capital in the economy. Technology is described by:
(13) [Mathematical Expression Omitted]
There is no significance to the choice of [Gamma] = 10 as it merely scales all equilibrium values of wages, interest rates, and social security taxes. [Alpha] = 0.65 is the share of output that goes to labor in equilibrium. The choice of 0.65 reflects the stylized fact that in most industrialized economies, labor's share of national income is about two-thirds. The baby boom is represented as a one-time-only increase in the population size. We present the results for the simulation when the bulge generation is 10% larger than the other generations. We have simulated the economy under alternative choices of parameter values. These results are not reported since they produced qualitatively similar results.
In simulating the model, we first determine the equilibrium interest rate from equation (10). Since this is a non-linear equation with no closed form solution, the roots are found using a Newton-Raphson numerical method. Several starting values were chosen for the algorithm, and all produced the same value for the equilibrium interest rate. After the first period, the starting value chosen for all subsequent periods is the previous period's equilibrium value.
Next, the equilibrium interest rate is substituted into equation (9) to obtain the equilibrium wage rate. The equilibrium wage rate is used in turn to obtain the social security tax rate required to generate a balanced budget for the social security system from equation (12). With the equilibrium values of [r.sub.t], [w.sub.t], and [[Theta].sub.t], we calculate savings, the size of the capital stock, and consumption for the individual generations.
The simulations are run through 100 generations. The baby-boom is generation 90 where the population size increases from 100 persons to 110. The population size of generation 91 is again 100 persons. We chose our economy's baby-boom generation to be the ninetieth to allow the simulation to settle down to a steady-state before the bulge occurs. The simulation continues after the bulge until the model returns to its steady-state.
Tables 1 and 2 present two sets of results. The first is for a no social security economy where [Theta] = 0 for all periods and the only source of retirement income is private savings. The second case has [Tau] = 1. Tables 1 and 2 report the following: [k.sub.t] is the capital stock at time t and equals the total savings of the generation born in time t-1, [[Theta].sub.t] is the social security tax rate that is required to pay a benefit of x at time t, [w.sub.t] is the wage rate at time t, [N.sub.t] is the size of the generation born at time t, [r.sub.t] is the return on capital at time t, [c.sub.t](t) is the consumption of the generation born in time t in the first period of their lives, and [c.sub.t-1](t) is the consumption of the generation born in time t-1 in the second period of their lives.
The no social security system simulation results are presented in Table 1. First, note the time path of wages. At time 89, the wage is at its steady-state value. When the bulge generation begins to work at t = 90, the relative supply of labor increases, pushing down the wage rate to 11.86 from its steady-state value of 12.26. The post-bulge generation sees their wages rise to the extent that their wages are higher than the steady state wages. The boomers, while saving less per person than do the previous generations, in total provide a higher capital stock to the post-bulge generation. Thus, for the post-bulge generation labor is relatively [TABULAR DATA FOR TABLE 1 OMITTED] more scarce than capital, leading to the higher wage rates, hence higher first period income. In turn, some of this income is saved, resulting in slightly higher than steady-state capital stocks for succeeding generations.
The return to capital rises dramatically as the bulge generation enters the work force, reflecting the relative scarcity of capital. Subsequent periods have a lower than steady-state return on capital because of the larger capital stock arising from the higher wage rate facing workers, and subsequently increased savings.
Now examine the pattern of consumption. The generation born in time 89 is the first to show the effect of the bulge generation in their consumption patterns. They are the beneficiaries of the high return on capital generated by the relative scarcity of capital in time 90. Their consumption is higher in the second period as a result of this increased return on their savings, rising to 7.02 from a steady-state value of 6.60. In contrast, the members of the bulge generation have lower than steady-state consumption in both periods of their lives. They consume 5.93 in the first period, then 6.13 in the second period. They face the "tragedy" predicted by Kotlikoff (1992). In their first period, they face low wage rates due to the relative surplus of labor. In their second period, they face low returns on their savings because of the low return on capital generated by the relative abundance of capital compared to labor.
The generations following the bulge all have higher levels of consumption than the steady-state. In the first period of their lives, this is due to increased wage rates arising from the relative scarcity of labor. In the second period, since the return on capital is actually below the steady-state return, the increase in consumption for the following generations is due to higher levels of savings from the higher incomes in the first period of life.
Table 1 indicates that both the generation immediately preceding the bulge and the generation immediately succeeding the bulge are the big winners. Both see their consumption significantly increased as a result of the movements in the relative price of labor and capital.
Table 2 shows the effect of introducing a social security system into the economy. The social security system transfers some of the income of the young generation to the old generation. The immediate effect is to reduce the amount of savings by the young and thus the capital available for the next generation to use in production, as predicted by Kotlikoff (1979). This effect makes labor relatively more plentiful compared to the no social security case, as seen by the fall in [w.sub.t] between Table 1 and Table 2. There is also a corresponding increase in the rate of return on capital since it is now relatively more scarce. These features of the economy lead to a consistent pattern of lower consumption by all cohorts in their first period of life, then higher consumption by all cohorts in their second period of life.
The required social security tax rate, [[Theta].sub.t] [TABULAR DATA FOR TABLE 2 OMITTED] shows the effect of the bulge that many commentators suggest. To pay for the social security payments to the bulge generation, the tax rate of the following smaller generation is higher. We also see the required payment fall for the bulge generation because there are more of them to pay the social security benefits of a smaller preceding generation. This result provides insight into why it was possible in Canada in the 1970s and 1980s to keep the CPP contribution rate fixed and increase the generosity of CPP benefits and the extent of CPP coverage.
Thus far it appears that the model justifies the concern over rising contribution rates. Required contribution rates do rise as the bulge generation retires. Notice however that consumption for the post-bulge generations is still higher than the steady-state levels. Notice also that for the generation born in period 92 and for subsequent generations, the contribution rate is lower than the steady-state level and consumption is higher. Regardless of the higher social security tax rate, the immediate post-bulge generation is still one of the two big winners from the bulge. As Demsetz et al. (1996) observes, higher taxes for the post-bulge generation amount to nothing more than taxing relatively well off future generations and transferring to the less well off baby boom generation.
The North American economy in the last 20 years has had high real interest rates and low real wage growth. We can see that these stylized facts are predictions of our model. Over the last 20 years, the baby boomers have entered and dominated the labor market. This corresponds to period 90 in our model, the time when the bulge generation is working. The model predicts that over this period we will see the combination of high real interest rates and lower real wages. The model also predicts that these conditions will continue until the baby boomers start to retire. Realistically, the effects of this retirement will not be felt for another 10 to 15 years.
Another stylized fact noted by Kotlikoff (1996) is that the ratio of the consumption of seniors to the consumption of workers is very high at the current time. Again, this is a prediction of our model. The generation immediately preceding the bulge generation has tremendously high consumption in their second period of life. Unlike Kotlikoff, however, we cannot blame this on any inter-generational transfer policy because the same stylized fact occurs in the model with no social security system and thus no inter-generational transfers.
B. Open Economy
This subsection discusses the simulation of the model with a perfectly elastic supply of capital. We refer to this case, with capital free to enter and leave the country, as our open economy model. The impacts of the demographic bulge are quite different from those of [TABULAR DATA FOR TABLE 3 OMITTED] the previous closed economy case.
The key feature of this model is that the return on capital, [r.sub.t], is determined in international markets and hence is exogenous for the domestic economy. The level of capital used in production must then be free to fluctuate to maintain [r.sub.t]. Instead of solving for the equilibrium interest rate, we use the exogenous interest rate to solve for the amount of capital that must be in the economy to produce an equilibrium interest rate of [r.sub.t]. The subsequent elements of the simulation remain identical to the closed economy version of the model.
One thing that is present in the open economy model that is not required in the closed economy model is savings by the rest of the world in the domestic economy. This is denoted [s.sup.r] and is the difference between the required amount of capital and the previous generation's savings.
(14) [Mathematical Expression Omitted]
In national income accounting, the quantity [Mathematical Expression Omitted] is known as the capital account surplus, or the amount of foreign investment in the domestic economy.
In the no social security system case, presented in Table 3, wage rates and consumption, do not change from year to year. This is because the capital to labor ratio in the economy remains constant since the exogenously determined interest rate is constant. In turn, wages are constant. Since the individuals in the different generations are the same, faced with the same problem they make the same allocation of resources between periods.
The bulge generation does affect the capital stock and quantity of savings required from the rest of the world. The required capital stock for period 90 rises by 10% (because the quantity of labor increases by 10%) to maintain the capital labor ratio. This requires a larger inflow of capital into the economy in period 90 compared to previous time periods. Because of the larger number of individuals saving in the boomer generation, the economy has an excess of capital in period 91, leading to a net outflow of capital in this period. In period 92 and after, the economy has a minor inflow of capital.
Table 4 presents the results for the case when a social security system is in place. As in the closed economy case, the initial effect of the transfer of income from the young to the old is to reduce the amount of domestic savings compared to the no social security case. This results in a larger capital inflow relative to the no social-security case. Again the effect of the bulge is to further increase this required capital inflow. In the subsequent period, a smaller capital inflow is seen due to the increased quantity of aggregate savings by the bulge generation in this period.
The intuition behind the difference in the two cases is seen by examining the required social-security tax rate. Individuals in the bulge generation have the same wage rate as does any other generation, but there are more individuals available to help meet the fixed social-security payment. This means that members of this generation face a lower tax rate in their first period of life and thus can save more of their income. The immediate post-bulge generation must cover social security payments to this larger generation. As a result, less saving occurs when the post-bulge generation [TABULAR DATA FOR TABLE 4 OMITTED] is young and a corresponding increase in the required capital inflow in the period occurs when this generation is old.
Now examine consumption. Unlike the closed economy case, no generation prior to the bulge generation is affected at all by the presence of the bulge. The bulge generation sees consumption in both periods higher than the steady-state values. The reason is that with the lower tax rate they are able to both consume and save more in their first period. The higher savings translate into higher consumption in the second period as well. The immediate post-bulge generation faces the mirror situation of higher tax rates forcing lower consumption and saving in the first period, followed by lower consumption in the second period due to lower savings.
As with the closed economy version of the model, we can relate these results to stylized facts of the North American economy over the last 20 years. The capital account surplus and associated current account deficit in both Canada and the United States have grown substantially over the last 20 years. In both Tables 3 and 4, this is a feature of the time period when the bulge generation is in the labor force. The model again suggests that this will be a feature of the economy until the baby boomers begin to retire in 10 to 15 years.
C. Discussion of Simulations and Qualifications
So far the analysis has been carried out in a deterministic setting. As a result, unemployment, which is a prominent characteristic of the Canadian labor market, is not addressed. In addition, the CPP is facing rising cost pressures from the increased longevity of Canadians. We briefly discuss the implications of these two factors.
Over the last 20 years, a major feature of the Canadian labor market has been the presence of high levels of unemployment. In our simulations, we impose the assumption of full employment. This assumption is more general than it appears. For the purposes of analyzing the viability of the CPP, what really matters are the total wages paid. How these wages are distributed over the labor force is of secondary importance. Our simulations summarize all labor market adjustments in price movements; that is, the wage fully adjusts so that the labor market always clears. If we allow for unemployment, we are saying that rigidities in the labor market exist and prevent prices from fully adjusting, requiring employment to adjust as well. Whether the adjustment occurs on price alone, as in our simulations, or on both price and quantity of labor services, the impact on the aggregate wages is qualitatively the same; as the capital to labor ratio falls, the marginal product of labor declines. If the market clears, the wage falls to clear the market. If the wage is not perfectly flexible, it will not fall by as much, and the remainder of the adjustment must come from a reduction in equilibrium employment.
Another source of rising cost pressures for the CPP has been the increase in life expectancy/longevity of Canadians. Retirees are claiming benefits for longer periods of time. With agents that only live for two periods and assuming the value of the CPP benefit is fixed for all time periods, we have assumed this problem away in the results in this paper. Although not reported here, we have modeled this effect as an increase in the weight on second period consumption in the utility function, coupled with an increase in CPP benefits. The major effect of this addition to the model is that savings increase, and thus first period consumption falls. The boomers are still the group with the lowest consumption in either period of their life. The relative price movements, which are the driving force behind the current results, change in magnitude, but not direction. The central message of the paper remains unchanged.
IV. IMPLICATIONS FOR THE CPP
This paper provides insight into the current debate on the reform of pay-as-you-go financed CPP. For example, several suggestions for reforming the Canada Pension Plan include increasing contributions now, in order to partially fund the retirement payments to the baby-boomers, and lowering benefits. We conclude that whether or not rising contribution rates to finance future social security benefits is desirable depends upon whether we believe the open or closed economy case is the better approximation of reality.
Both models are informative for demonstrating the sources of "stylized facts" of macro-variables over the last 20 years. The closed economy case can mimic the movements in wages and interest rates. The open economy case, as a limiting case of perfectly elastic supply of capital, cannot mimic any price movements, but it can demonstrate the emergence of capital account surpluses we have observed. If the open economy case is the better approximation of reality, then the boomer generation should not be opposed to the status quo of social security as they clearly benefit. Baby Busters should be the most opposed since they fare poorly.
If the closed economy case is the better approximation of reality, then it is not at all clear that reforms are necessary or socially desirable even though taxes will rise as the boomers retire. The following cohort will pay higher taxes, but they also have higher real incomes than do the baby boomers, and this amounts to nothing more than taxing the relatively well off "busters" to transfer to the less well off "boomers." As Summers (1993, p. 3) argues, such considerations make proposed reforms of pay-as-you-go social security arrangements difficult to evaluate:
It is not clear how this (expanded) government expenditure should be financed. A reduction of social security benefits, an increase in the social security tax rate and an increase in the legal retirement age are all being considered. But, these policies are not desirable as they impose an unfair burden on the current generation. At the other extreme, we could say that the future generation has the duty to support the old through a high social security tax rate, because they enjoy the low housing prices and high wages which are the benefits of population decrease.
World capital markets are becoming increasingly integrated, with very few barriers to the movement of capital. Using this type of reasoning, we may be tempted to disregard the closed economy case for Canada. There may be reasons, however, why the closed economy case is a more accurate description of what is happening in Canada, even if the country is technically an open economy. First, the world's dominant economy (particularly from a Canadian perspective), the United States, experienced a baby-boom. Given the influence of the U.S. economy on the Canadian economy, an open Canadian economy in this case will see interest rates rise, and the response of the domestic economy will mimic the closed economy simulations. In an extreme case, for example, an open economy facing exogenous interest rates such as those found in Tables 1 and 2 will exactly correspond to the closed economy case in all other variables as well.
Second, although our open economy case can explain the capital account surplus in Canada over the last 15 years, an alternative explanation for the Canadian surplus undermines the open economy characterization of Canada. Capital flows into Canada have partly served to finance the federal deficit rather than to finance physical capital formation, suggesting that foreign investors regard Canadian physical capital to be a fairly risky investment. With this interpretation of the capital account surplus, capital flows may be less than perfectly mobile, and an economy's "openness" is a matter of degree rather than a distinct state; domestic forces do matter in the determination of factor prices.
Even if we cannot say whether the open or closed economy case is the appropriate one, we can make some comments about the arguments of commentators on pay-as-you-go arrangements. First, under any scenario, you cannot have both boomers and busters worse off because of a pay-as-you-go system. The tragedy of the boomers, low wages and low return to savings, occurs in the closed economy case without social security. The current CPP reforms that accelerate the rate of increase in the CPP contribution rates to build up a surplus to fund future benefits are intergenerationally regressive. Members of the following cohort pay higher taxes but still have higher consumption than the boomers as a result of their higher wages.
One of the arguments against pay-as-you-go arrangements is that they were fine when the implicit rate of return to the plan, which is the rate of growth of the economy, were greater than the return to private investment. This was the case when the systems were first set in place but has not been true since the mid 1970s. This leads to recommendations such as Townley's (1981) to discard pay-as-you-go systems in favor of funded pension finance arrangements. This paper suggests that this reduction in the implied rate of return to pay-as-you-go systems may be a result of transitory factor market conditions. In the presence of adjustment costs, it may be the case that we should maintain the current system in the face of the temporary problems.
Ultimately, the pressure for reform seems excessive. Either real wages will rise with the contribution rate as the boomers retire, or any reforms to pay-as-you-go arrangements will have largely distributional consequences, and hence are more in line with political pressure rather than the alleged economic pressures that provide the main justifications for CPP reform.
1. Born (1996) makes the same point for the United States. Pesando notes that output per employee in Canada grew 1.1% per year from 1981 to 1990 and that this growth was well below the historical average of 2.5%.
2. Kotlikoff (1992) describes the tragedy of the baby boomers as follows: They bid their wages down due to highly competitive labor markets, and when they go to consume out of their savings, they lose again from a depressed return to savings as they all sell their assets at once. Easterlin (1980) argues that stagflation, high inflation and high unemployment that emerged through the 1970s, is the expected outcome when a large generation enters the labor market.
3. Summers calculates that this decline can be fully compensated for if productivity grows at 0.2% per year or if the retirement age increases by three-to-four years.
Berger, M., "The Effect of Cohort Size on Earnings Growth," Journal of Political Economy, 93:3, 1985, 561-573.
Burbidge, J., "Social Security and Savings Plans in Overlapping Generations Models," Journal of Public Economics, 21:1, 1983, 79-92.
Demsetz, Harold, Laurence J. Kotlikoff, Henning Bohn, Earl Thompson, and Andrew Samwick, "Welfare across the Generations," Contemporary Economic Policy, January 1996, 1-21.
Easterlin, R. A., Birth and Fortune: The Impact of Numbers on Personal Welfare, Basic Books, New York, 1980.
Foot, D. K., and D. Stoffman, Boom, Bust and Echo: How to Profit From the Coming Demographic Shift, Macfarlane Walter and Ross, Toronto, 1996.
Government of Canada, Federal/Provincial Information Paper on the CPP: An Information Paper for Consultations on the Canada Pension Plan, 1996.
Kotlikoff, L., "Social Security and Equilibrium Capital Intensity," The Quarterly Journal of Economics, 93:2, 1979, 233-253.
-----, Generational Accounting: Knowing Who Pays, and When, for What We Spend, Maxwell MacMillan Canada, Toronto, 1992.
Pesando, J. E., "Policy Issues: The Pension Crisis," Paper Presented at the conference titled The Impact of an Aging Population on Society, University of Calgary, March 15-16, 1993.
Richardson, D., "Social Security Options: Validation, Offset, or Return to Pay-as-You-Go," Discussion Paper 94-20, University of British Columbia, 1994.
Stapleton, D., and D. Young, "The Effects of Demographic Changes on the Distribution of Wages, 1967-1990," Journal of Human Resources, 19:2, 1984, 175-201.
Summers, L. H., "Keynote Speech: Population Problem Confronted by the World Economy - A Summary," Public Finance, 48:Supplement, 1993, 1-3.
Townley, P., "Public Choice and the Social Insurance Paradox: A Note," Canadian Journal of Economics, 14:4, 1981, 712-717.
Welch, F., "Effects of Cohort Size on Earnings Growth: A Re-examination of the Evidence," Journal of Political Economy, 87:5, 1979, S65-S97.
Wolfson, M., and B. Murphy, "Aging and the Public Sector: Retrospect and Prospect," Paper Presented to Conference on Reform of the Retirement Income System, Queen's University, 1996.
Yoo, P., "Boom or Bust? The Economic Effects of the Baby Boom," Review, Federal Reserve Bank of St. Louis, 76:5, 1994, 13-22.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||Canada Pension Plan; economic reform|
|Author:||Emery, J.C. Herbert; Rongve, Ian|
|Publication:||Contemporary Economic Policy|
|Date:||Jan 1, 1999|
|Previous Article:||Are wage increases in Chinese state industry efficient? Productivity in Nanjing's machine-building industry.|
|Next Article:||Communal duality: agricultural subsidies from TVEs.|