The Marglin-McFadden-Mishan debate on public investment criteria: Some Clarifications

This paper reviews the post 1963 debate on public investment criteria stemming largely from Marglin's work. At issue here, are the two main propositions of Marglin, namely, that the social marginal rate of time preference is the appropriate social rate of discount (SRD) for public projects, and that the social opportunity cost (SOC) of capital raised to facilitate public investment is, in general, in excess of the money cost of such investment. Both supported and challenged by different writers, this debate does not provide any clear consensus as to the current status of these propositions. The central purpose of this paper is to put the series of apparent claims and counterclaims found in the literature to an analytical test, and draw the appropriate conclusions. This we do by explicitly considering the technology, fiscal policy and savings reinvestment behavior in a simple model of maximizing the present discounted value of the marginal net consumption stream generated by the public project. We argue that given Marglin's assumptions, the criterion that the marginal social rate of time preference be the SRD is valid as has already been demonstrated by Diamond (1968) and McFadden (1972). His other conclusion that the SOC should generally exceed unit v is seen to be correct only in special cases.     

Public investment and social risk-sharing

How the government uses market rates of return to discount risky public projects is intrinsic to the efficient allocation of capital and the choice of public investment policy. This paper shows that social discount rates for risk can be higher than, lower than, or equal to private rates. Previous analyses which argue for ‘one-sided’ rules where social discount rates are always lower or equal to private rates are shown to be inappropriate in general. The paper shows in a general model that the difference between the social value of a public project and its market value is determined by the social insurance value that the project has from improving risk-sharing relative to the market. Depending on the project and degree of market imperfection, this term can be positive, negative, or zero, explaining the different possibilities for the social evaluation of risk.     

Shadow pricing of labour and capital in an economy with unemployed labour

This paper derives the shadow prices of labour and capital to be used in the public sector in a situation of unemployment. The setting considered is that of a three-good, two-period general equilibrium model. Then shadow prices are compared to their corresponding market prices and shown to closely depend on own and cross-elasticities of supply and demand for labour and investment. In the first part, a rigid wage rate is the sole source of distortion; then, a tax on capital income is introduced so that our formula for the social rate of discount can be contrasted with that of Harberger, Sandmo and Drèze.     

A dynamic heterogeneous labour demand model for German manufacturing

There is currently general agreement amongst economists that the discount rate to compute the present value of benefits and costs in the economic evaluation of public policies is defined according to the social time preference approach. However, whether this rate has to be used for the discount of non-monetary health consequences is a question for which there is no satisfactory reply. In this paper, it is argued that such a reply rests on the estimation of the relationship between the individual time preference for health and money in the contexts of private and social choice. In support of this argument an empirical analysis has been carried out in which the individuals making-up a representative sample of the population of Zaragoza (Spain) have been faced with a series of hypothetical inter-temporal choices. Their replies have implicitly revealed their temporal preference rates and have led to the conclusion that health consequences are discounted at a higher rate than monetary consequences. This would appear to be contrary to the standard practice applied in the economic valuation of health programmes.     

Optimal taxation and discount rate for public investment in a growth setting

This paper studies the optimal policy of a government which maximizes intertemporal social welfare using such instruments as taxes on interest income and wages, and debt in conjunction with public investment. In doing so, it has to face a decentralized economy where in each generation individuals and firms are free to maximize their own objectives subject to their own private constraints. The welfare function is a sum of discounted generational utilities and its maximization is handled by using dynamic programming. From the first order conditions so derived, it appears that an optimal policy of taxation and public capital accumulation is that which sets the tax rates according to Ramsey's optimal taxation structure and which equates the rate of return on public investment to the rate of social time preference.     

A Note on Discounting an Increasingly Uncertain Future

This note clarifies the roles played by the wealth and precautionary effects in determining the socially efficient discount rate for public investment projects and how the rate should vary over time. We first give a general characterization of the effects of stochastic shifts in the consumption growth rate on the magnitude of the socially efficient discount rate. We then show that increasing uncertainty in the consumption growth rate provides a natural and compelling rationale for discounting more distant future consumption at a lower rate.     

Discount rates and shadow prices for public investment

This paper extends the Harberger–Sandmo–Drèze model for public discount rates to a many-good economy. It derives a formula for discount rates which are specific to each public enterprise or agency and used by them to discount future outputs and inputs evaluated at market prices. Such an approach is shown to be more efficient than that of simply using a single rate for all public projects. It is also more practical than asking each public firm to use second-best shadow prices in their analyses of investment projects.The general results are first provided; then, simple cases are considered and numerical examples presented to help interpret our formula and analyse its main determinants.     

Performance of Australia's Ethical Funds

Australia's three major public ethical investment funds achieved mixed financial success in the seven years to 30 June 1998, though on average the funds underperformed relative to the market. For the four‐year and five‐year holding periods to 30 June 1995 and 1996 respectively, the average holding‐period returns for the three funds were less than the risk‐free rate. This is strong evidence of investors incurring a financial discount for investing ethically and, with respect to the ethical investor's utility function, it is evidence of the marginal utility increasing as the ethical attributes of assets increase.     

Debt financing of public investment: On a popular misinterpretation of “the golden rule of public sector borrowing”

In this paper I challenge the proposition that the golden rule of public sector borrowing is consistent with the principle of intertemporal allocative efficiency, in the sense that growth-enhancing public investment justifies a structural public deficit. I demonstrate that in the long run the social opportunity cost of debt-financed public investment exceeds the social opportunity cost of tax financed public investments. This result holds if the social rate of time preference is lower than the interest rate on government borrowing. Thus a benevolent government would use taxes to finance public investment. In the short run, debt financing is justified if public investment has a considerable growth effect on private consumption. This requires a corresponding initial undersupply of public capital.     

The nature of discounting

From monetary policies to the climate change problem, from the burden of private credit card debts to the evaluation of public projects, discount rate is the central issue, yet there is little clear understanding about the nature of discounting. In this paper, applying a newly developed production theory, we discuss how discount rate is related to other factors in social systems, such as risk, duration of production, fixed cost in production and market size. The relations among different factors in a social system put constraints on the ranges of discount rate that are viable in particular environments. Our findings have strong policy implications. In a world of increasing cost of extracting natural resources, the continuation of low discount rate policy will generate wide gyration of social systems that we have witnessed in recent years.     

Discounting Nordhaus

This paper evaluates Nordhaus's neoclassical complaints about the Stern Review from the vantage point of classical growth theory. Nordhaus criticizes the Stern Review because it uses a discount rate that is well below the market rate of return on capital. From the perspective of classical growth theory, Nordhaus's belief in choosing preference parameters for the social planner based on observed market rates of return is equivalent to assigning the preferences of the capitalist agents to the social planner. This equivalence is an implication of the Cambridge Theorem, which interprets the Ramsey equation as the saving function of the capitalist agents.     

Schooling choices: Preferences, discount rates, and rates of return

This paper develops an empirical model to identify the structural parameters of schooling preferences and human capital production. Our model distinguishes between consumption and investment motives with regard to schooling. The results show that both motives matter. Preferences for schooling vary with social background and ability. Children from poorer social backgrounds and of lower ability have a lower preference for schooling. The discount rate that enters the net value of lifetime income varies with social background as well. The marginal rate of return to schooling decreases with ability and schooling. On average the marginal rate of return is 7.3 per cent, which can be contrasted with a `Mincerian' rate of return equal to 4.8 per cent. This indicates that the usual OLS estimate underestimates the true rate of return. First version received: November 1997/Final version received: February 1999     

The Intergenerational State Education and Pensions

When credit markets to finance investment in human capital are missing, the competitive equilibrium allocation is inefficient. When generations overlap, this failure can be mitigated by properly designed social arrangements. We show that public financing of education and public pensions can be designed to implement an intergenerational transfer scheme supporting the complete market allocation. Neither the public financing of education nor the pension scheme we consider resemble standard ones. In our mechanism, via the public education system, the young borrow from the middle aged to invest in human capital. They pay back the debt via a social security tax, the proceedings of which finance pension payments. When the complete market allocation is achieved, the rate of return implicit in this borrowing–lending scheme should equal the market rate of return.     

Migrant Age Profiles and Long-Run Living Standards in Australia

The age profile of immigration to Australia can be altered by public policies. This article applies a method of valuing long term population paths in order to evaluate alternative immigrant age profiles based on a proxy indicator of living standards over time for Australia. Results show that the variation in value of the population path follows an inverted U-shape by immigrant's age. The strength of preference between immigrants aged 20–29 years and older working age immigrants reduces as the social discount rate increases. Higher life expectancy and lower fertility accentuate preference for younger working age immigrants.     

Self-fulfilling patience

This paper explores macroeconomic implications of investment in patience in a standard neoclassical growth model with Becker-Mulligan endogenous time preferences. The endogenous discount rate acts as a new margin for inter-temporal decisions, in addition to the standard margin that hinges on the marginal return of capital. This time preference margin alters the equilibrium dynamics and stability of the neoclassical growth model substantially. When the discount rate is positive, there may exist multiple steady states that are either saddle-point stable or unstable. When the discount rate is negative, the unique steady state is locally indeterminate due to self-fulfilling patience investment. Interestingly, the existence of the local indeterminacy does not need any externality. When the discount rate is zero, various types of bifurcations can happen, leading to rich equilibrium dynamics such as limit cycles and chaos. We show that opening up the closed economy can cause aggregate instability.     

Consensus Consumer and Intertemporal Asset Pricing with Heterogeneous Beliefs

The aim of this paper is to analyse the impact of heterogeneous beliefs in an otherwise standard competitive complete market economy. The construction of a consensus probability belief, as well as a consensus consumer, is shown to be valid modulo an aggregation bias, which takes the form of a discount factor. In classical cases, the consensus probability belief is a risk tolerance weighted average of the individual beliefs, and the discount factor is proportional to beliefs dispersion. This discount factor makes the heterogeneous beliefs setting fundamentally different from the homogeneous beliefs setting, and it is consistent with the interpretation of beliefs heterogeneity as a source of risk.
We then use our construction to rewrite in a simple way the equilibrium characteristics (market price of risk, risk premium, risk-free rate) in a heterogeneous beliefs framework and to analyse the impact of beliefs heterogeneity. Finally, we show that it is possible to construct specific parametrizations of the heterogeneous beliefs model that lead to globally higher risk premia and lower risk-free rates.     

Human capital and economic growth in an overlapping generations model

The paper describes an aggregative optimal growth model, the essential features of which are that individuals are mortal and obtain their labor skill through educational training. The process of human capital formation is described by an education function which relates the pass rate to the educational expenditure per student. Two alternative scenarios, private and public education regimes, are separately investigated. Under the decentralized education regime, risk-neutral individuals borrow to finance their education when young. Under the centralized education regime, the cost of education is financed by taxes imposed on the workers in the economy, and the central government maximizes a long-term social target function. The equilibria of both regimes are analyzed and various comparative static results derived. It is shown that educational investment in a decentralized equilibrium is higher than that in the centralized steady state. We also establish that there exists a time discount rate at which or above which the decentralized per capita consumption exceeds that of the centralized steady state whereas for time rates of discount sufficiently near the population growth rate, the above result will be reversed.     

Time Preference and Trade Imbalance

This paper presents a unified theory of trade and investment in a world where the rate of time preference varies between countries. In the framework proposed by Buiter (1981 ), we can analyze a situation wherein two countries have different rates of discount. Here, the value of the debt to income does not converge to zero but remains constant even in the long run. Furthermore, we show that the existence of less‐capital‐intensive nontradables promotes capital movements: since a more patient country incompletely specializes in less‐capital‐intensive nontradables, capital must flow out of it.     

Incentive regulation and investment: evidence from European energy utilities

This paper investigates the relationship between investment and regulatory regimes (incentive vs. rate-of-return regulation) for a sample of EU energy utilities from 1997 to 2007. We control for the effect of firm ownership and for cross-country differences in the underlying energy demand and energy supply. To deal with potential endogeneity of the regulatory regime, we apply instrumental variable methods (2SLS and GMM). Our results show that investment rate is higher under incentive regulation than under rate of return regulation. Using original data on the regulatory tools (X factor and WACC), we find that investment of incentive regulated firms appears highly sensitive to the X factor, consistent with efficiency- and profit-seeking motivations. Electric utilities investment is also sensitive to the level and change in the weighted average cost of capital (WACC). Finally, we find that the positive relationship between private control and investment is not robust to IV estimations, suggesting that in Europe regulation may have reduced the differences between private and public firms’ incentives to invest.     

Do retirement planning strategies alter the effect of time preference on retirement wealth?

An individual’s willingness to accumulate retirement wealth is influenced by their preference for intertemporal consumption. People with a strong preference for current consumption (high personal discount rate) may choose to save less and face the risk of decreased retirement preparedness. A negative relation between a high personal discount rate and retirement wealth may be reduced when individuals engage in some form of retirement planning. Using the National Longitudinal Survey of Youth, we provide evidence that respondents with a high personal discount rate accumulated 37% less retirement wealth, on average, between 2004 and 2008, when compared with respondents with a low personal discount rate. However, when retirement planning strategies were included in the model, there was no statistical difference in retirement wealth between people with high and low personal discount rates. The retirement planning strategies included calculating a retirement income need, hiring a financial planner for retirement or engaging in both of these activities.