Monday, March 2, 2020

When discounted at the risk-free rate, real Social Security wealth increased from $5.6 tn in 1989 to $42.0 tn in 2016; adjusting for systematic risk, it grew from $4.6 tn in 1989 to $34.0 tn in 2016

Catherine, Sylvain and Miller, Max and Sarin, Natasha, Social Security and Trends in Inequality (February 29, 2020). SSRN:

Abstract: Recent influential work finds large increases in inequality in the U.S., based on measures of wealth concentration that notably exclude the value of social insurance programs. This paper revisits this conclusion by incorporating Social Security retirement benefits into measures of wealth inequality. Wealth inequality has not increased in the last three decades when Social Security is accounted for. When discounted at the risk-free rate, real Social Security wealth increased substantially from $5.6 trillion in 1989 to just over $42.0 trillion in 2016. When we adjust for systematic risk coming from the covariance of Social Security returns with the market portfolio, this increase remains sizable, growing from over $4.6 trillion in 1989 to $34.0 trillion in 2016. Consequently, by 2016, Social Security wealth represented 58% of the wealth of the bottom 90% of the wealth distribution. Redistribution through programs like Social Security increases the progressivity of the economy, and it is important that our estimates of wealth concentration reflect this.

Keywords: Social Security, Inequality, Top Wealth Shares
JEL Classification: D31, E21, G51, H55, N32

CEO types: “Leaders,” who do multifunction, high-level meetings, & “managers,” who do individual meetings with core functions; firms that hire leaders perform better (it takes 3 years to see the difference)

Oriana Bandiera, Andrea Prat, Stephen Hansen, and Raffaella Sadun, "CEO Behavior and Firm Performance," Journal of Political Economy 0, no. 0 (-Not available-): 000. Feb 2020.

Abstract: We develop a new method to measure CEO behavior in large samples via a survey that collects high-frequency, high-dimensional diary data and a machine learning algorithm that estimates behavioral types. Applying this method to 1,114 CEOs in six countries reveals two types: “leaders,” who do multifunction, high-level meetings, and “managers,” who do individual meetings with core functions. Firms that hire leaders perform better, and it takes three years for a new CEO to make a difference. Structural estimates indicate that productivity differentials are due to mismatches rather than to leaders being better for all firms.

From a 2017 version (
This paper combines a new survey methodology with a machine learning algorithm to measure the behavior of CEOs in large samples. We show that CEOs di↵er in their behavior along several dimensions, and that the data can be reduced to a summary CEO index which distinguishes between “managers” –i.e. CEOs that are primarily involved with production-related activities– and leaders -i.e. CEOs that are primarily involved in communication and coordination activities.

Guided by a simple firm-CEO assignment model, we show that there is no “best practice” in
CEO behavior—that is, a behavior that is optimal for all the firms—rather, there is evidence of
horizontal di↵erentiation in CEO behavior, and significant frictions in the assignment of CEOs to firms. In our sample of manufacturing firms across six countries we estimate that 17% of firm-CEO pairs are misassigned and that misassignments are found in all regions but are more frequent in emerging economies. The consequences for productivity are large: the implied productivity loss due to di↵erential misassignment is equal to 13% of the labor productivity gap between firms in high- and middle/low-income countries in our sample.

This paper shows that an under explored dimension of managerial activity–that is, how CEOs
spend their time–is both heterogeneous across managers and firms, and correlated with firm performance. Future work could utilize our data and methodology to inform new leadership models, which incorporate more explicitly the drivers and consequences of di↵erences in CEO behavior, and in particular explore the underlying firm-CEO matching function, which is not dealt with explicitly in the current paper. Furthermore, a possible next step of this research would be to extend the data collection to the diaries of multiple managerial figures beyond the CEO. This approach would allow us to further explore whether and how managerial interactions and team behavior vary across firms and correlate with firm performance (Hambrick and Mason, 1984). These aspects of managerial behavior, which are now largely absent from our analysis, are considered to be increasingly important in the labor market (Deming (2015)), but have so far been largely unexplored from an empirical perspective. We leave these topics for further research.

Himba men prioritize mate fidelity and current reproductive partners in investment decisions, but social obligations to past and current partners and the presence of other male investors also influence decisions

Paternity confidence and social obligations explain men's allocations to romantic partners in an experimental giving game. Brooke A.Scelza, Sean P. Prall, Kathrine Starkweather. Evolution and Human Behavior, Volume 41, Issue 1, January 2020, Pages 96-103.

Abstract: Paternal care in humans is facultative, with investment decisions responsive to socioecological context. In particular, paternity confidence is thought to have a significant impact on men's provisioning. However, various aspects of the relationship a man has with his partner can also influence the way he provides for his children. Previous papers have tended to focus either on these kinds of relationship dynamics or on the impact of paternity confidence. However, these categories are often intertwined and parsing their contributions can be conceptually and methodologically difficult. To better understand how paternity confidence and relationship dynamics impact men's investment decisions, we used a series of pictorial vignettes to assess the resource allocation strategies of Himba men. We focus on three traits: mate fidelity, partner type (marital or non-marital), and relationship status (current or former). Results suggest that men prioritize mate fidelity and current reproductive partners in investment decisions, but social obligations to past and current partners and the presence of other male investors also influence decisions. Himba men appear to be balancing social norms related to marriage and fatherhood with individually-driven incentives to invest in current and more faithful partners.

4. Discussion

Our results show strong evidence for our primary predictions, which are derived from life history theory, and which are fairly intuitive. Fig. 3 demonstrates this most clearly. All else equal, men favor more faithful partners (rows 1, 10, 15), partners they have formal unions with (2, 7) and, to a lesser extent, current partners (4, 9, 11, 14). Our secondary predictions, which bring these factors together into a series of potential strategies, show greatest support for the paternal investment hypothesis, as partners with omoka children are consistently short-changed in the allocations, and the social obligation hypothesis, as wives are strongly favored over other types of partners (Fig. 1, Panel D). The best fitting model included all three primary factors: partner fidelity, relationship type and relationship status, providing additional evidence that men are employing a mixed-strategy. These experimental results emphasizing a mixed strategy are aligned with behavioral data on Himba paternal investment (Prall and Scelza, 2016). Measures of children's nutritional status, fosterage status and livestock transfers show some titration of investment toward children men believe to be omoka, but these occur largely when investments occur in the more private household domain. When investments occur in more public domains, social obligations to wives and marital offspring lead men to treat biological and non-biological children more equally.

[Fig. 3. Posterior predictions of differences between categories from the best fit model predicting food allocation. Each distribution represents the difference between the two types of women, so that distributions to the right of zero indicate the first woman listed is predicted to receive more food gifts.

The best fitting model shows that overall girlfriends and ex-wives are treated relatively equally (Fig. 1, Panel D), but this finding masks some of the interesting variation that occurs in distributions to these two partner types (Fig. 3). In the two cases where omoka status differs between girlfriends and ex-wives (rows 12 and 13), men allocate substantially more to the partner with his biological child. Paternal investment appears to be a greater motivator than either social obligations or mating effort. Similarly, when the child is not omoka in both cases, distributions to girlfriends and ex-wives are indistinguishable (row 11). Where we see a breakdown is in the scenario where both children are omoka. Here we see men favor girlfriends over ex-wives, suggesting some support for mating effort over social obligation (row 14).
The second part of our study was designed to understand whether men are titrating their allocations dependent on whether another partner (and potential provider) is present. Men's allocations to girlfriends fit this pattern, with substantially greater giving to girlfriends who are unmarried than to those who are married. There are several explanations that could account for these results, other than the need-based explanation we are positing. Men could give less to women with other partners as a form of punishment for their infidelity. Women without other partners may also be more likely to father a man's next child, so that giving more to her represents mating effort. Our data do not allow us to test between these explanations. However, opportunistic quotes recorded during the interview process invoking differential need were frequent among our participants. In a typical iteration, one man stated, “This woman has a husband. I will help this one because she doesn't have a husband and needs support.” Therefore, we conclude that need-based considerations are likely to be one important factor in men's decision-making.
A slightly different pattern emerges in Task 2B where men made allocations across wives. Those without boyfriends still received more than those with boyfriends, but the results were less consistent. Men were much more likely to distribute resources evenly across wives than across girlfriends. In the two categories where men were distributing an even number of items (for food and clinic fees) they divided them equally between wives, 40% and 81% of the time respectively, compared to only 19% and 42% of the time with girlfriends. One man reported, “I think of them as my co-wives. You have to treat them equally so that they don't get jealous and one doesn't think that I don't like her.” In the third iteration of Task 2, men were forced to make an unequal distribution. Unlike with girlfriends, where this did not pose a problem, some men were very resistant to making a choice between their wives, with one commenting “I would cut that goat in half” rather than have to choose. This again speaks to the complex interaction of factors affecting men's strategies. Social obligation may dictate that men should treat their wives (relatively) equally, but men may also be hedging their bets, keeping both wives happy may increase their overall chance of paternity in future children.
We also surmise that need was not driving men's choices in the wife scenario to the same degree that it was with informal partners. Opportunistic comments did not stress support from another partner in the way they did in the girlfriend comparison. This may be due to the difference in obligatory provisioning that exists between formal and informal partners. While a man is obligated to provide for his wives, giving to girlfriends is voluntary. Previously, our work showed that for this reason, women had a stronger preference for generosity in boyfriends, whereas in husbands, wealth mattered more (Scelza & Prall, 2018). Here we see that men bias their giving to girlfriends based on need, but factoring in the gifts of others to their wives seems secondary to concerns about keeping the peace at home and treating wives equally.

4.1. Limitations and future directions

The experimental approach used here provides some insight as to how Himba men make decisions about allocating resources across partners of different types. While the high rates of concurrency and nonpaternity in this population mean that most men encounter these types of decisions at some point in their lives, rarely would they encounter all of them at once, highlighting an advantage of the vignette approach. However, as with all studies that use hypothetical allocations, these data may not reflect the ways that men divvy up real resources, but rather how they view social expectations to allocate resources. In response to this, many economic games use cash and follow through with distributions as promised in the game; however, this approach was not feasible here. Despite this limitation, we found that the men in our study understood the complexity of the problem before them, as highlighted by opportunistic comments made throughout the task.
In addition while we thought it was important to address the ways in which men's allocations could be altered by the presence of other providers, this is neither the only nor necessarily the most important way that need might influence giving patterns. Other factors we did not consider in this study include the relative wealth or security of the woman herself, the wealth of her parents, and her number of dependents. Similarly, we did not look at how men's own wealth or security affect his provisioning strategy. Finally, while we did not detect any effects of age or marital status in our analyses (see Fig. S7), it is possible this is due to small sample size. Future studies with larger numbers of participants could address these shortcomings, using either behavioral or experimental data.
Finally, this paper focuses exclusively on men's provisioning decisions. The emphasis on men's care as facultative in the literature drove the design of this study. However, while certain aspects of maternal care are obligate in humans (most significantly pregnancy), most care by mothers is quite variable. Future work should consider how the factors studied here impact maternal provisioning decisions.