Mariacristina De Nardi

De Nardi co-authored the article with Giulio Fella, Marike Knoef, Gonzalo Paz-Pardo and Raun Van Ooijen. It examines the size and distribution of wage shocks and the role of insurance against these shocks in the United States and the Netherlands. Unlike previous papers which explored shocks to individual earnings, De Nardi, Fella, Knoef, Paz-Pardo and Van Ooijen distinguished the mechanisms of shocks in wages and changes in hours. The authors provided data on the distribution of male wages and earnings and their household earnings and income found by analyzing distributional measures of wage changes. The results of their analysis indicated that in both countries, there was evidence of non-linearity and age dependence; high wage and earnings risk existed for people with the lowest and highest wages and earnings. Wage and earnings persistence varied significantly by age; in the Netherlands, income risk is lower for the middle range of income distribution. De Nardi, Fella, Knoef, Paz-Pardo and Van Ooijen found that the government plays a large role in reducing the risk in the Netherlands; in the United States, the role of family in reducing risk is more important. They looked into how much insurance was provided by spousal labor by comparing individual earnings to total earnings at the household level, and how much was provided by the tax and transfer systems. De Nardi, Fella, Knoef, Paz-Pardo and Van Ooijen compared both countries, concluding that the labor supply has little effect on the standard deviation of wages in the Netherlands. They found that taxes and transfers have a greater effect in the Netherlands, reducing inequality and risk in wages and earnings. In the United States, the standard deviation of all income measures was generally higher; there was evidence of volatility in hours (which increased the dispersion of wage changes), and the standard deviation of male earnings had increased. The United States spousal-labor supply played a larger role in reducing the standard deviation of male earnings on all levels of previous earnings; in the Netherlands, that role was played by the government.[9]

Poor health can shorten a person's life expectancy. This article explores the effects of health shocks over a lifetime. Co-authored with Svetlana Pashchenko and Ponpoke Porapakkarm, the article developed a structural framework to help measure the long-term effects of poor health. Considering the fact that inequality can be traced back to factors predetermined early in life, the authors created a model which can yield the short- and long-term dynamics of health and health-related inequality of economic outcomes to measure the lifetime costs of poor health. For their research, De Nardi, Pashchenko and Porapakkarm used three data sets: the Health and Retirement Study, the Medical Expenditure Panel and Survey, and the Panel Study of Income Dynamics. They found that individuals were generally willing to pay about five percent of their average income to increase their probability of being healthier by one percent. People in poor health tend to save less, and differences in health are a factor in explaining a large wealth gap.[10]

Margherita Borella, Mariacristina De Nardi, Fang Yang investigate the aggregate importance of gender and marriage. They note that most macroeconomists ignore women and marriage when setting up structural models and when calibrating them using data on men only. Borella, De Nardi, Yang question if by ignoring gender and marriage in both models and data, that implies the resulting calibration matches well the key economic aggregates. They used data from the Panel Study and Income Dynamics and the Health and Retirement Survey for the 1941–1945 cohort. In their paper, they constructed and calibrated four different economies. Economy 1 uses a standard one-gender, no marriage, lifecycle framework and like usual macro models, they only used data on men for calibration. While Economy 2 is the same model as Economy 1 but instead they calibrated using data on both women and men. Economy 3 uses the same model as both Economy 1 and 2 but now its calibrated by aggregating the data at the household level and using couples instead of singles. Finally, in Economy 4, the model single and married men and women over their whole life cycle. They found that Economy 4 did better than the other economies. Which shows that modelling marriage and gender is important for understanding key economic aggregates over the life cycle. Borella, De Nardi, Yang conclude that macroeconomists should be taking marriage and gender into account in the quantitative structural models. They state that by modelling marriage and gender explicitly would give the best results for matching the aggregates.[11]

In this article, Gene Amromin, Mariacristina De Nardi, Karl Schulze investigated whether the widening gap between the rich and poor has any direct effects on macroeconomic aggregates. They largely focus on the Great Recession. By using the Panel Study of Income Dynamics and Credit Bureau Panel Data, they compared consumption and wealth during the Great Recession. Amromin, De Nardi and Schulze argue that the role of borrowing constraints cannot be fully captured by only looking at data where the majority of households have little wealth before the recession. They state that most macroeconomic theories do this. Summarizing two papers, one by Krusell and Smith and another one by Krueger, Mitman and Perri, it is provided evidence to show that inequality can make the effects of a recession even worse. Amromin, De Nardi and Schulze conclude that different measures of household constraints have been permanently increased due to the Great Recession.[12]

Collaborating with 27 other economists, they measured the composition and magnitude of medical spending in the last three years before death. They took health care data from nine countries: Denmark, England, France, Germany, Japan, the Netherlands, Taiwan, United States and Canada(Quebec). The comparison between countries revealed that there is no direct relationship between how a country's health care service are funded and how they are provided. They found the average medical spending per capital in the last twelve months of life is about $80,000 in the United States, $60,000 in the Netherlands and Denmark and $50,000 in Germany. Medical spending in the last twelve months and last three years of life is high. De Nardi, French, et al. recorded that in the last twelve months of life, medical spendings ranged from 8.2 percent in Japan to 22.7 percent in Quebec. Since Hospital spendings make a significant portion of medical spendings in the last twelve months compared to the last three years of life, they saw the medical spendings ranging from 13.5 percent in Japan to 34.9 percent in Taiwan win the last three years of life. They concluded that all nine countries they looked into, the medical spendings at the end of life were high compared to spending at other ages. In the last twelve months of life, they saw medical spending making up most of its aggregate spending. In the United States, it was about 8.5 percent and 11.2 percent in Taiwan. The reduction of medical spending does not effect the total medical spending significantly. De Nardi, French, et al. concluded that instead there should be a reduction in the caring costs for people with chronic conditions.[13]

This paper by De Nardi, Eric French and John Bailey Jones construct a model of saving for retired singles that include a mix of different medical expenses and life expectancy. For their paper, they used the Assets and Health Dynamics of the Oldest Old dataset (AHEAD). De Nardi, French and Jones only took the data on single retired individuals for their analysis. Which consisted of 3,259 individuals, where 592 were men and 2,667 were women. For their model, they accounted for social insurance programs. Their estimated model shows that the savings of the elderly are largely due medical expenditures. De Nardi, French and Jones predict that the elderly, between the ages of 74 and 84, the median assets for those in the top permanent income quantile see a fall from $170,000 to $130,000. If all medical expenses were to be eliminated, the median asset would fall much more. We would expect it to fall from $170,000 to $80,000. Their results are mainly due to two things. First, even if the elderly had health insurance, the out-of-pocket medical and nursing home expense would still be a large amount. Second, they found that the average medical cost rises significantly with age and income. Their model predicts the average out-of-pocket medical expenses will rise from $1,100 when they are 75 to $9,200 when they are 95. They also found that, a 95 year old in the top quintile of the permanent income distribution is to spend $15,800 on medical expenses. While someone who is the same age but in the bottom quintile of the permanent income distribution is expected to spend $1,700. So medical spending largely effects savings. By properly accounting for old age expenditure on medical care and social insurance programs providing a consumption floor, it can explain the elderly's savings. De Nardi, French and Jones conclude that savings of the elderly are motivated by the medical expenditures and that social insurance affects the saving of both the income-rich and income-poor.[14]

De Nardi received a research grant from the National Science Foundation Research Grant. She is the principal investigator and conducted the research from 2003 to 2006.

De Nardi co-authored this journal article with Marco Cagetti. De Nardi and Cagetti constructed a model that analyzes the role of borrowing constraints as determinants of entrepreneurial choices and the effects it has on wealth inequality and aggregate capital accumulation. They found that more restrictive borrowing constraints created less inequality in wealth holdings but causes a reduction in average firm size, number of entrepreneurs and aggregate capital accumulation. Cagetti and De Nardi show the relationship between entrepreneurship and wealth. From their data, they explain how entrepreneurs are richer than non-entrepreneurs but business owners are much richer than self-employers. The paper notes that many entrepreneurs have potentially high rates of return but are constrained in the amount they can borrow. Since an entrepreneur wealth acts as collateral, the amount that can be borrowed is determined by how rich they are. Cagetti and De Nardi used the assumption where if there is a borrowing constraint, entrepreneurs could sell their idea to someone else that faces less constraint. They show how the main factor in determining the number of entrepreneurs, size of firm, aggregate capital accumulation and overall wealth concentration in the population is due to how stringent the borrowing constraint is and voluntary bequest. They concluded that the main factors can cause issues for policy analysis, such as subsidized loans and taxing bequest. The paper notes that taxing bequest can reduce the amount of entrepreneurial wealth, which affects the amount they can borrow. Resulting in less entrepreneurs.[15]

De Nardi collaborated with Marco Cagetti to summarize the wealth distribution and its economic models in the United States. The majority of the data they used are from the Survey of Consumer Finances. Cagetti and De Nardi explain that all quantitative models of wealth inequality make human capital exogenous and they argue that we should actually take human capital accumulation into consideration when studying savings and wealth inequality. They also point out that some areas in models of inequality can be employed and extended. Cagetti and De Nardi states that it takes a lot of work to fully understand the quantitative importance of each factor in determining wealth inequality. Though there has been great advance in the models, there is still room for improvement in these models in order to apply them to problems where inequality is a key determinant.[16]