Increase in Life Expectancy: Macroeconomic Impact and Policy Implications 1. Issues Kyooho Kwon, Fellow It has been widely speculated that Korea s rapidly rising life expectancy is the major cause behind dwindling private consumption and households average propensity to consume. Average propensity to consume (consumption relative to income) declined to as low as 71.9% in 2015 from 77.9% in 2003, according to Statistics Korea s Household Income and Expenditure Survey. Meanwhile, life expectancy has sharply increased to record 82.4 yrs in 2014. Life Expectancy and Average Propensity to Consume 84 (Age) (%) 80 82 78 80 Life Expectancy (Left) 76 78 74 76 Average Propensity to Consume (Right) 72 74 2003 2006 2009 2012 2015 Source: Life Table, Household Income and Expenditure Survey, Statistics Korea. 70 This paper attempts to quantitatively analyze the impact of increasing life expectancy on the macroeconomy, focusing on consumption and savings, and draw on policy implications from key findings. 1
A simulation, using the general equilibrium model, was conducted to closely examine the changes in the national savings rate when individuals adjust their savings in line with increased longevity. 2. Increasing Life Expectancy and Falling Propensity to Consume Koreans continue to live longer by 0.5 yr per annum, intensifying the burden of retirement savings. Life expectancy has risen from 75.5 yrs in 2000 to 82.4 yrs in 2014; based on the assumption retirement is at 60, life expectancy has increased roughly 40%. The current low income replacement ratio and stagnant labor force participation of seniors highlight the growing need to prepare for a stable retired life. According to the Survey of Household Finances and Living Conditions, the proportion of pensions and transfer income for those over 60 accounts for a mere 29% of total income. The senior population s labor force participation rate failed to improve in proportion to increased life expectancy, despite better health, just inching up from 38.2% in 2000 to 39.7% in 2014. Theoretically, the overall decline in Korea s average propensity to consume can be construed to be the result of the rapid rise in life expectancy. The average propensity to consume dropped sharply from 0.78% in 2003 to 0.72% in 2013; 1 widespread phenomenon observed in all age groups with 2 a steeper decline in older groups. 2
Average Propensity to Consume by Householder Age 1.00 2003 2015 0.90 0.94 0.80 0.75 0.76 0.80 0.75 0.78 0.70 0.73 0.73 0.75 0.72 0.68 0.68 0.60 20s 30s 40s 50s 60s 70s Note: Author s calculation using raw data from the Household Income and Expenditure Survey. Source: Household Income and Expenditure Survey, Statistics Korea. 1 Recognizing that longer life expectancy necessitates more savings, economic agents across all age groups have responded by lowering their propensity to consume. 2 Older generations have less time to adjust to possible shocks, as such their average propensity to consume has declined at a relatively faster pace. For instance, a 1-year rise in life expectancy from 80 to 81 causes the life span of a 70-year old to increase by 11% and that of a 50-year old by 3%; meaning that the older one is, the less there is to spend within a given income bracket. A simulation of the life cycle and overlapping generations model, reflecting Korea s demographics, was conducted to quantitatively analyze the impact of increased life expectancy on savings (=100-average propensity to consume). In the short-run, the decline in the average propensity to consume resulting from increased life expectancy leads to a decrease in economic agents consumption-to-income ratio; potentially negative impact on the Korean economy. In the long-run, however, the pre-retirement labor supply will expand, boosting the accumulation of capital due to increased savings and ultimately, enhancing GDP growth. 3
Accordingly, this paper uses the life cycle and overlapping generations model to analyze the dynamic impact of increasing life expectancy on the macroeconomy. Korea s demographics and economic growth trajectory were reflected and quantificated, and the impact of increased life expectancy caused by falling mortality was analyzed (See Appendix 2. for the summary). 3. Analysis of Increasing Life Expectancy on the Macroeconomy The above simulation reveals that increased life expectancy raises the savings rate (reduced consumption) in the short-run but boosts the economy through an accumulation of capital, which encourages consumption, in the long-run. It is assumed that an increase of 0.5 yrs at a specific point in time ( ) will induce a shock and therefore, economic agents will respond by optimizing labor supply, consumption and savings. In the short-run, at the point of shock, economic agents will increase savings by 0.3%p, in consideration of the changes in life span, compared to when there is no change. Such a rise in savings leads to a drop in consumption, but as income is secured through increased labor supply, the decrement in consumption falls below the increment in savings. In the long-run, the rise in savings will mean more capital accumulation and a larger pre-retirement labor supply, resultantly heightening economic growth and consumption. There was an increase in GDP compared to a scenario where there is no change in life expectancy, which would eventually lead to higher consumption. 4
0.4 Impact of Increased Life Expectancy on Savings and Consumption (%) 0.4 (%) 0.3 Savings Rate 0.2 Consumption 0.2 0.0 0.1-0.2 0.0-0.4-0.1-5 0 5 10 15 20-0.6-5 0 5 10 15 20 Note: 1) Horizontal axis denotes time (yrs), and the point of increased life expectancy is expressed as 0. 2) Difference between savings rate and consumption in the model when compared to the scenario of no changes in life expectancy. Analysis on the impact of increased life expectancy since 2000 finds that Korea s savings and GDP growth rates increased 3.5%p and 0.4%p, in 2015. Korean s average life expectancy has increased 0.5 yr per annum since 2000. Such an increase is estimated to have caused the savings rate (as of 2015) to rise by 3.5%p, when compared to a scenario where there is no increase. Accordingly, the increase in life expectancy has contributed to raising Korea s savings rate despite the economy having just entered a mature phase, and having a rapidly aging population. Korea s GDP is estimated to rise by about 0.4%p as of 2015. Specifically, as the rise in savings heightens capital accumulation, the marginal productivity of labor increases. As a result, labor supply expands, which eventually improves GDP. The results imply that if life expectancy does not increase, Korea s potential growth may deteriorate rather quickly. 5
Another possibility is that if potential growth diminishes as the working-age population reaches its peak and total factor productivity growth slows, the increase in life expectancy may cushion the impact of the decline. Impact of Increased Life Expectancy Since 2000 4 (%p) 0.6 (%p) 3 Savings Rate 0.4 Economic Growth Rate 2 0.2 1 0 2000 2005 2010 2015 0.0 2000 2005 2010 2015 Note: Difference between savings rate and economic growth rate when compared to a scenario where there is no change in life expectancy since 2000. Meanwhile, the long-term effects may be limited if labor supply fails to respond flexibly, or the rise in savings does not lead to more investment. If pre-retired economic agents fail to flexibly adjust their working hours, it may disrupt the expansion of labor supply, inevitably diminishing the increment in GDP growth (0.3%p). Additionally, if increased savings leads to a surplus in the current account (overseas saving) instead of more investment, GDP growth (via capital accumulation) may again be curtailed. 6
4. Summary and Policy Implications The rise in the savings rate caused by increased life expectancy could have ramifications for consumption in the short-run. But, in the long-run, it could help boost GDP growth by inducing capital accumulation and expanding the labor supply, eventually having a positive impact. The phenomenon of falling propensity to consume must be construed in view of structural aspects; a rational decision made by economic agents to respond to longer longevity. Increased life expectancy can cushion the slowdown in Korea s potential growth rate in the long-run. The above confirms that consumption-boosting measures should pay more attention to structural factors, even though private consumption has been relatively stagnant. To turn increased savings into increased domestic investment, consideration must be given to achieving services industry improvements, decisive regulatory rationalization and a flexible labor market to enhance return on investments. Above all, as longer longevity acts as an incentive for the middle-aged and seniors to participate in the labor market, more active reform of the labor market needs to be implemented to create sufficient labor demand that corresponds to the supply. Flexible management of working hours, productivity-based wage peak system and vocational support for the middle-aged are good examples and could not only help alleviate concerns about hiring seniors but also allow seniors to enjoy more job opportunities. 7
Appendix 1. Net Household Savings Rate and Average Propensity to Save The average propensity to save (= 100 - average propensity to consume) in the Household Income and Expenditure Survey can be interpreted to be the savings rate of households. But, its moving pattern is markedly different from the net household savings rate in the National Accounts, which covers whole households. The average propensity to save rose from 22.1% in 2003 to 28.1% in 2015, whereas the net household savings rate edged up only recently after staying flat at around the 5% range until the mid-2000s. Average Propensity to Save and Net Household Savings Rate 30 (%) (%) 15 24 12 18 9 12 6 6 0 Average Propensity to save (Left) Revised Household Savings Rate (Left) Net Household Savings Rate (Right) 2000 2005 2010 2015 3 0 Note: 1) Average propensity to save = 100 average propensity to consume 2) Revised household savings rate is calculated by author using the distribution of national income in the National Accounts. Source: Statistics Korea, Household Income and Expenditure Survey; Bank of Korea, National Accounts, The difference is attributed to the method of measuring income and savings; households net savings rate excludes social contributions in the share of income and savings of households. 8
When social contribution is included in household income and savings and the scope of consumption is limited to final consumption expenditure, households net savings rate and their average propensity to save appear to move in a similar pattern. Definition of net household savings rate in the National Accounts can be expressed as follows: * Net household savings rate (%) = (households net saving)/(net adjusted household disposable income + adjustment for the change in net equity of households in pension funds)*100 * Households net savings = income actual consumption expenditure (final consumption expenditure + social transfers in kind) * Net adjusted household disposable income = employee compensation + households operating surplus + net property income + social benefits other than social transfers in kind + net other current transfers social contributions - current taxes on income, wealth, etc + social transfer in kind Social contributions (mainly national pension) is a part of employee compensation and also savings of households preparing for life after retirement. Accordingly, categorizing them into household income and savings can better help identify the propensity to save. Social transfer in kind includes non-profit institutions final consumption, national health insurance, free education, etc. certain government consumption expenditures and is not determined by households; hence excluded from income and consumption. Households depreciation was also included in income and consumption, which showed that although it affected the overall level somewhat, the impact became insignificant over time. Meanwhile, the fact that households use private pension programs as well as the national pension system to prepare for post-retirement suggests that social contributions do not systematically suppress consumption. In other words, even without social contributions, which is a form of forced savings, households would make the rational decision to increase the share of private pensions in response to increased life expectancy. 9
2. Summary of the Life-cycle and Over-lapping Generations Model This paper assumes a closed economy consisting of representative companies with Cobb-Douglas production technology and households of different age groups. Individual households are independent decision makers and engage in economic activities at age 20 and retire at 65. Representative companies, whose aim is to maximize profits, produce by combining the capital and labor they obtain in the competitive market with their technology. Market income and interest rates are assumed to be determined at a level where labor supply from households and labor demand from companies are equal as is household savings and companies demand for capital. The economic model is quantificated to reflect the trajectory of Korea s demographics and economic growth. Individual households consist of male and female adults who provide labor. Korea s demographics are based on the Life Table in the Population Projections for Korea. Also, this paper applies the wage differences between age groups using the Survey on Labor Conditions by Employment Type to reflect the relative difference in productivity among workers of different ages. The trajectory of the total factor productivity of representative companies is based on actual past results, and is assumed to moderate gradually. Total factor productivity was identified by deducting the contribution of labor and capital (production factors) from the realized economic growth rate, and injected into the model. The capital stock used in calculations was borrowed from the production capital stock in the National Balance Sheets, and labor input was substituted with the number of employed in the Economically Active Population Survey. 10
For the time discount factor, which determines the savings rate in the model, the investment rate from the simulated model for 1991-2010 was set to match the average GDP to investment rate in the National Accounts. Although the average investment rate was set to match the data, the changes in the rate over time were determined endogenously. 11