Macro Roundup

New Nick Bloom paper we highlighted in newsletter presents evidence suggesting supervision is the limiting constraint, firms with better management prices are more productive and employees have higher earnings with CEO pay rising the most, "...We use confidential Census matched employer-employee earnings data to study the relationships between pay, productivity, and management practices. We find that employees at more productive firms and firms with more structured management practices have substantially higher pay, both on average and across every percentile of the pay distribution. This pay-performance relationship is particularly strong amongst top paid executives, with a doubling of firm productivity associated with 11% more pay for the highest-paid employee (likely the CEO) compared to 4.7% for the median worker. This pay-performance link holds in public and private firms,although is almost twice as strong in public firms for the highest paid executives. Top-executive pay volatility is also strongly related to productivity and structure managed, suggesting this performance-pay relationship arises from more aggressive monitoring and incentive practices amongst top executives...."
To measure quality of management they use the Management and Organizational Practices Survey (MOPS), "...TheMOPS was issued as a supplement to ASM for survey years 2010 and 2015. All establishments that were included in the ASM samples for those years were also sent the MOPS.10 In both survey waves, the MOPS asks plant managers 16 questions about the management practices at their establishments. In particular, respondents are asked questions regarding their practices relating to performance monitoring, target setting, and incentivization of workers….we score responses to each question between 0 and 1, where zero corresponds to the least structured practices (practices that are less explicit, formal, frequent, or specific) and one corresponds to the most structured practices (practices that are more explicit, formal, frequent, or specific). We then compute an establishment’s overall structured management score as the simple mean of the scores of all completed questions.11The resulting management score is itself bounded between 0 and 1, where we interpret an establishment with a larger value as having more structured management practices….find that establishments with higher structured management scores also tend to be more productive. We aggregate across establishments to the firm level by taking employment-weighted averages of the management scores across establishments...."

Four core findings, "First, employees at more productive firms and firms with more structured management practices have substantially higher pay across all percentiles of the earnings distribution. Not only are executive earnings higher but so are earnings at every level, from the 1st percentile upwards."

Second, "... this increase in earnings is greater at higher pay levels. Thus, higher productivity and more structured management practices are also associated with higher levels of pay dispersion within firms (as well as higher pay levels as noted above).This is particularly notable at the very top end of the earnings ranks. For example, while the firm’s top earner sees a pay increase of 11% for the doubling of productivity, the 5th, 25th and 50th ranked managers see only 9.0%, 7.1% and 6.5% respectively...."

Third, "... this pay-performance link holds in public and private firms, although it is almost twice as strong in public firms for the highest paid workers. The highest paid worker (likely the CEO) sees a 16.4% pay increase in public firms but only a 9.4% pay increase in private firms for a doubling a productivity. Lower ranks, in particular employees outside the top 50 highest paid, display similar performance-pay relationships in public and private firms...."

Fourth, "...Finally, top-earner pay volatility is also strongly related to productivity and structured managed practices. One explanation is that more productive firms adopt more aggressive management practices - more intensive monitoring and aggressive performance pay schemes - which leads to both higher levels of pay but also higher volatility of pay...."

The evidence:

Productivity and pay, "...Figure 1 presents the point estimates of model (2), where the outcome is the mean log earnings of workers within a firm earnings percentile. The figure shows three striking patterns. First, all the coefficients on the y-axis are positive, showing that workers at higher-productivity firms tend to earn more across the entire earnings distribution, consistent with the higher average pay in Column (1) of Table 2. Second, consistent with Columns (2)-(5) of Table 2, this positive relationship is not equal across the earnings distribution. Instead, the upwards slope shows that the pay of higher earners is more positively correlated with productivity than the pay of lower earners. For example, a 10% increase in productivity predicts a 0.82% increase in pay for earners at the 90th percentile (the 0.086 coefficient for the 90th percentile in Figure 1) but only a 0.51% increase in pay for earners at the 10th percentile (the 0.053 coefficient for the 10th percentile in Figure 1). Third, this upward slope in the correlation between pay and productivity across the earnings percentiles is particularly large across the top percentiles: the difference in coefficients between the 100th percentile and the 90th percentile is larger than the difference in the coefficients between the 90th percentile and the 10th percentile...."

Management structure and pay across the within-firm earnings distribution, "...First, we document how management relates to average pay and within-firm inequality. Table 3 presents estimates of versions of model (1) in which we replace productivity with the firm’s management score as the key explanatory variable; this table is analogous to Table 2. Recall that management scores are normalized between 0 and 1, with a mean value of 0.67 in our sample (Table 1).As Table 3 shows, the patterns we see between productivity and pay in Table 2 are also present for management and pay. Firms with more structured management tend to have higher average pay and larger inequality. In terms of the aggregate distribution of management, these patterns are economically large. Moving from the 10th to the 90th percentile in the distribution of management is associated with an increase in average pay of 6.3% and an increase in the gap between the top earner’s pay and the median worker’s pay of 26%.23 Second, as we did for productivity, we consider how pay across the firm earnings distribution correlates with management. Analogous to Figure 1,Figure 6 presents the point estimates of version of model (2), where the outcome is the mean log earnings of workers in a given earnings percentile and we replace productivity with management as the key explanatory variable. While the standard errors are relatively large - the management sample contains only 2.5% of the observations from the main sample - we see similar patterns. Workers across the earnings distribution at firms with more structured management tend to have higher earnings, and these correlations generally increase across the distribution, with workers in the top percentile experiencing the largest correlation.In Figure A3, we focus on the very top earners at the firms; similar to Figure 2, we see positive and increasing coefficients on management for these top earners, as we move towards the top earner, although these estimates are noisy.Finally, we return to studying pay volatility. Because part of management structure is how a firm incentivizes its workers, studying the link between management and pay volatility is informative about whether the relationship between productivity and pay volatility may reflect incentives. Figure 7, analogous to Figure 5, presents point estimates of model (2) where the outcomes it the with in year standard deviation of log quarterly earnings and we replace productivity with management. We see striking patterns here: Top earners at firms with more structured management tend to have higher earnings, particularly so for the very top earners. This implies senior executives at firms with highly structured management practices - more performance monitoring and more aggressive performance bonuses - have earnings that are more tightly linked to firm-productivity, but also higher levels of pay volatility...."

“…We investigate this by estimating model (2) where the outcome is the within-year pay volatility for the top earners at firms. We measure within-year pay volatility as the standard deviation in quarterly log earnings, within a given year. Figure 5 presents these estimates. Consistent with performance based pay,earners at more productive firms are more likely to have higher within-year pay volatility, and this is particularly true for the very top earners. Because our sample restricts to full-year workers, we do not believe that the with-year pay volatility, or the relationship between volatility and productivity, is driven by top earners at more productive firms being more likely to leave the firm. Rather, because these workers are consistently employed at the firm, these patterns reflect real within-job pay variation, which may be particularly driven by the sizes of their bonuses….”

“…To further explore the role performance-based pay for top earners, we explore changes in productivity and pay within firms. We replicate Figure 2 but now include firm fixed effects, such that now the covariation of pay and productivity are interpreted as changes over time within firms. Figure 4 shows the point estimates of this version of model (2). By adding the firm fixed effects, we reduce the variation in productivity and the coefficients decrease. Recall that in Figure 2, the coefficient on the pay of the top earner, without firm fixed effects, is 0.15. In Figure 4, this coefficient falls by two-thirds to 0.05. However, the patterns are still apparent. Regardless of which of the top earners we consider at firms, when a firm’s productivity is relatively high, those earners earn more (i.e., the coefficients are positive). Further, the correlation between productivity and pay within firm increases as we consider higher earners at the firm. These patterns mean that when a firm’s productivity is relatively high, all top earners are predicted to earn more, but the predicted proportional increase are largest for workers at the very top…”

Foreign buyers manage their Treasury holding more efficiently than American investors, "...We instead observe foreigners’ Treasury portfolio at the security-level and find results that strikingly differ from the existing literature…. Foreign investors are prominent in the Treasury bond market, holding just under half of all outstanding Treasuries…Foreigners do not earn a low return on their Treasury portfolios and have higher Sharpe ratios than U.S. investors. And private foreign investors appear to be price sensitive..."
Bottomline, “…Our security-level and aggregate analyses show that private foreign investors perform quite well, earn similar yields as other investors on similar maturity bonds (which is inconsistent with bad timing), on a risk-adjusted basis perform better than U.S. investors, and, consistent with elastic demand, increase purchases and the duration of their Treasury portfolio when CIP deviations decrease. Foreign officials, with much broader objective functions, appear to be price insensitive but do not seem to pay for this, as their Sharpe ratios are higher than other investor types'. Therefore, our analysis shows no support for the notion that foreign investors have poor performance and, other than foreign officials, inelastic demand….”
Evidence, “…Results are in Table 1. Using directly observed portfolios, over the period June 2003-June 2019 private foreign investors had returns (3.77 percent per annum) that exceeded market returns (3.59 percent). The fact that private foreign investors earn higher than market returns is not consistent with the poor performance = inelastic demand notion that is used to support some convenience yield theories. U.S. private investors (4.34 percent) also had higher than market returns. And while foreign official investors' returns (3.02 percent) were lower than market returns, the 57 basis point gap is relatively small and statistically insignificant. Combining the foreign official and foreign private portfolios, we find that foreign investors' return of 3.21 percent is only 38 basis points lower than the market, and the difference is not statistically significant….”

Key evidence, “…A decline in interest rates boosts the value of industry leaders relative to followers, and this effect snowballs as the level of the interest rate becomes lower. We now explore mechanism through which industry leaders gain relative to industry followers when interest rate declines. We first explore whether the decline in economywide interest rates such as the 10 year Treasury rate has been associated with a differential decline in the cost of debt financing for industry leaders relative to followers. A stronger pass-through of lower Treasury rates into the cost of debt financing for industry leaders represents another advantage of a low interest rate environment for leaders relative to followers. As before, industry leaders are defined as the largest 5% of firms in each Fama-French industry, measured by market capitalization. We define each firms borrowing cost as the ratio of interest expense to total liabilities, which is simply average rate of interest paid on it’s liabilities. We treat the largest 5% of borrowing costs in the sample as missing in order to exclude any outliers in the data. Figure 6 plots the evolution of the median borrowing cost faced by both leaders and followers, displaying both a clear downward trend and also a stark increase in spreads since 1980. The spread in the 1980s was 31 basis points whereas the spread in the 2010s is 127 basis points.The interest rate spread between market leaders and market followers widened even though the level of interest rate fell considerably. We present estimates of the effect of a decline in interest rates as defined by the monetary policy shocks discussed in Section 3 on the cost of debt financing for industry leaders versus followers. As above, we measure borrowing costs using the average interest rate paid by the firm on all debt. We calculate this as the total interest expense divided by total debt. We exclude the top 5% observed values of borrowing costs, due to the presence of implausibly large values….”

Note this builds off their “Low Interest Rates, Market Power, and Productivity Growth” which argued that low interest rates reduce business dynamism as market leaders gain share as market followers cut investment to a greater degree then they do in face of low rates (attached)

Thomas Kroen, Ernest Liu, Atif Mian and Amir Sufi, “Falling Rates and Rising Superstars,” National Bureau Of Economic Research, October 2021, https://www.nber.org/papers/w29368

Mian and Sufi find that falling interest rates disproportionately help industry leaders, particularly in low rate environments. They find that spreads (borrowing) btw industry followers versus the leaders is stable and the borrowing capacity of leaders growths Mian characterizes it as “snowballs” as interest rates approach 0. They find that as rates go down the market value of leaders rises more than followers (superstar effect) and their borrowing cost (relative to followers) fall. That allows cheaper leverage, buybacks, capex and acquisitions making effect more pronounced. “…we find that falling interest rates boost the relative valuation of industry leaders relative to industry followers, and the relative valuation effect becomes largest as the initial interest rate approaches zero. This result is robust to the use of high frequency monetary policy shocks as a source of variation in economy-wide interest rates. A decline in the interest rate disproportionately lowers the cost of borrowing of industry leaders, who take advantage of the lower cost of borrowing to raise additional debt financing, increase leverage, repurchase shares, boost capital investment, and conduct acquisitions. All of these effects snowball as the level of the interest rate decline; that is, a decline in interest rates has a stronger effect on all of these outcomes of leaders relative to followers when the initial level of the interest rate is already low.The findings provide empirical support to the idea that extremely low interest rates may be a culprit in explaining the rise of superstar firms in the U.S. economy…

The basic dynamic, “…This paper formally investigates the empirical relationship between economy wide interest rates and the rise of superstar firms in the United States using the merged CRSP-Compustat data set from 1962 to 2019. We begin by analyzing the stock market reaction to an interest rate decline for industry leaders versus industry followers, where we define the top 5 percent of firms by value in an industry as “leaders” and the rest as “followers”. We construct a “leader portfolio” that goes long industry leaders and shorts industry followers, and we examine the portfolio’s performance in response to changes in the ten year U.S. Treasury rate (r). We find that the leader portfolio exhibits higher returns in response to a decline in r, and, more importantly, this response becomes stronger, or snowballs, when the initial r is low. We control for the price to earnings ratio as a measure of implied duration, showing that the snowballing effect is not mechanically driven by industry leaders having higher duration. This suggests that lower r impacts relative firm valuations not only through changing the discount rate but also through possible endogenous changes in expected future cash flows that favor the current industry leader – a finding consistent with the strategic competition channel above. Interest rate changes are endogenous to changes in the overall economy, and there is an obvious concern that omitted variables may be responsible for the results. For example, the interest rate decline is naturally correlated with negative news about future demand; if industry leaders have a larger option value on future demand, then the spurious correlation between the interest rate movement and expected demand will lead us to underestimate the true differential impact of the interest rate decline on industry leaders….”
Impact is stronger at lower interest rates, “…Since the impact is stronger at low levels of r, for expositional purposes we always report the predicted magnitude of our estimates at an interest rate of r = 2%. We find that a 10 basis point reduction in r when r = 2% translates into a 0.53 percentage point larger increase in the market valuation of industry leaders relative to industry followers. We investigate the possible reasons and mechanisms that lead to lower rates boosting the relative valuation of industry leaders. We find that a fall in r exhibits a stronger pass-through to the borrowing costs of industry leaders relative to industry followers, and this effect snowballs at lower r. In terms of magnitudes, a 10 basis point decline in r leads to a 14 basis point relative decline in the borrowing cost of industry leaders relative to followers at r = 2%. When the initial r is very close to the zero lower bound, a 10 basis point decline in r leads to a 24 basis point relative decline. Industry leaders take advantage of the lower cost of debt financing. There is a large relative increase in debt issued by industry leaders relative to industry followers, and the book leverage ratio of leaders also increases. The effects of a decline in r on these two financial outcomes also exhibit the snowballing effect: the effect of a decline in r is larger when the initial interest rate is lower. When the economy is close to the zero lower bound, a 10 basis point decline in r leads to a 5.2 percent relative increase in debt issued, and a 1 percentage point relative rise in the leverage ratio for industry leaders. Some of the additional debt raised is used to buy back shares, which also contributes to the rise in leverage we observe….”

“…The final set of results concern the relative real effects of lower interest rates on industry leaders versus followers. These results largely follow the stock market and financial effects described above. We find that a decline in r leads to a relative increase in capital expenditures, cash acquisitions, and property, plants, and equipment (PPE) for industry leaders relative to followers. These real effects, like the valuation and financial effects, also snowball at lower levels of the initial interest rate. The results on investment provide empirical support to the prediction in Liu et al. (2021) that industry leaders respond more aggressively to a decline in the interest rate in order to “go for the kill” to ensure their claim on the more valuable future cash flows. Interestingly, one of the strongest results is on cash acquisitions, which suggests that industry leaders may be more likely to target their rivals when interest rates fall from very low levels. Overall, our results have important implications for the broad literature on persistently low interest rates (or r*) and their implications for the macroeconomy, including the literature on “secular stagnation” (e.g., Summers (2014)). Most of the work in this literature has focused on possible causes for the low interest rate, with explanations ranging from demographics, inequality, and low productivity growth. Our work suggests that there is a potentially important feedback effect from low r back to the real economy through market structure and industry competition….”

Table 2 “…The estimates reported in columns (1) and (2) confirm earlier results. A decline in the interest rate is associated with positive returns for the leader portfolio, and this positive return response to a decline in the interest rate is larger in magnitude when the interest rate is lower.Column (3) uses the 10-year real interest. rate level as before. The coefficient on the interaction between the real rate and the change in interest rate is even stronger than in Table 1. Column (4) shows that the results are driven by both positive and negative changes in interest rates. In particular, the excess return results are materially unchanged whether only positive changes in interest rates or only negative changes in interest rate are used. Column (5) shows that the results are not driven by industry leaders that may have higher duration of cash flows for mechanical or spurious reasons. If industry leaders had higher duration for spurious reasons, then they would have a higher price to earnings (PE) ratio and the difference in the PE ratio between the industry leader and follower at the time of interest rate shock would explain our asymmetric valuation response to a decline in r. Column (5) controls for a “PE portfolio” that is long the top 5% of firms by PE in an industry and short the rest. Inclusion of the PE portfolio return does not change the coefficients of interest, and the return of the leader-minus-follower portfolio is itself negatively correlated with the return of the PE portfolio.4 This result shows that lower r impacts relative rm valuations. not only through changing the discount rate but also through the endogenous changes in future cash flows that favor the current leader…”

“…. Figure 2 helps visualize the snowballing effect, and the critical interest rate at which the snowballing effect becomes active. More specifically, the figure reports non-parametric estimates of the effect of a change in interest rates on the leader portfolio ( ) at various grid points for it1, which is the lagged 10-year nominal Treasury rate. Moving from right to left in the figure, the coefficient estimate of becomes statistically significantly negative when the level of the nominal interest rate it1 falls into the 5 and 6% region. The 10 year nominal U.S. Treasury rate has been below 5% since 2000, with only a few exceptions….”

Using evidence from Quebec in the late 1990s when, "...extension of full-time kindergarten to all 5 year olds and the provision of childcare at an out-of pocket price of $5 per day to all 4 year olds. This $5 per day policy was extended to all 3 year olds in 1998, all 2 year olds in 1999, and finally all children aged less than 2 in 2000..."
The paper finds, "....Our results are striking. The introduction of universal childcare in Quebec led to a very large increase in the use of care. The proportion of 0-4 year olds in care rose by 14 percentage points in Quebec relative to the rest of the country, or roughly one-third of the baseline childcare utilization rate. This rise in childcare was associated with a sizeable increase in the labor force participation of married women. Participation rose by 7.7 percentage points in Quebec, or about 14.5% of the baseline. The difference between the rise in participation and the rise in childcare utilization primarily reflectsreduced use of informal childcare arrangements, or the “crowd out” of informal childcare by this new subsidized childcare. Partly as a result of this large “crowd out”, the taxes generated by the new maternal labor supply fallfarshort of paying for the costs of the increased childcare subsidies. We also find consistent and robust evidence of negative effects of the policy change on child outcomes, parenting, and parent outcomes.Child outcomes are worse for a variety of parent-reported measures, such as hyperactivity, inattention, aggressiveness, motor/social skills, child health status, and illness. Parental interactions with children are worse along all measured dimensions, and there is some evidence of deterioration in parental health and a reduction in parental relationship quality. These are subjective measures, but the consistency of the results suggests that more access to childcare is bad for these children (and, at least along some dimensions, for these parents). There are, however, interpretations of these findings which are more benign. While some of these explanations appear inconsistent with the data, we cannot rule out the possibility that our findings represent a short term adjustment to childcare, and not a long-run negative impact...."

Michael Baker, Jonathan Gruber and Kevin Milligan, "Universal Childcare, Maternal Labor Supply and Family Well-Being," National Bureau Of Economic Research, December 2005, https://www.nber.org/system/files/working_papers/w11832/w11832.pdf

Supporting Evidence on child outcomes, “….Table 4 presents the results for child outcomes for the difference-in-differences specification. We consider several summary scores of the individual behavioural measures: hyperactivity-inattention; general anxiety; separation anxiety; physical aggressiveness / opposition; motor and social development; and theppVT score. Appendix C shows how the indices are constructed from the underlying survey questions, and also provides estimates on the individual components using the same specification as we use for the scores. We also consider five measures of health: an indicator for excellent health; indicators for the child never having nose/throat or ear infections; an indicator for having an asthma attack in the last 12 months; and an indicator for injury in the last 12 months….”

Supporting Evidence on labor supply impact, “..Table 3 reports the effects of this policy on the labor supply of married women.27 The first two columns consider mothers of 0-4 year old children. There is a rise in the employment of married women in Quebec, relative to the rest of Canada, of 7.7 percentage points, or 14.5% of baseline participation. In subsidy terms, we find that each 10% subsidy raises maternal work by 2.4%. Given the 1996 Quebec price mean of 0.528 and employment mean of 0.530, this implies an elasticity of maternal work with respect to childcare costs of 0.236, which is towards the low end of estimates in the literature on labor supply effects of childcare costs. As Figure 4 shows, this change in labor force participation once again represents a trend break for Quebec relative to the rest of Canada: whereas female participation in Quebec was 5-6 percentage points below that in the rest of Canada before the policy, it is higher after the policy. There is also a large rise in other measures of labor supply for married women. Average weeks of work rose by 3.5, or more than 13% of baseline, for an elasticity of weeks of work with respect to childcare prices of 0.217. The next three rows show the effect on part-time vs. fulltime work. There is an increase in the odds of women working more than 20 and more than 30 hours per week that is comparable to the entire rise in employment, but no effect on the odds of working more than 40 hours per week, suggesting that women were moving from no work to full-time work as a result of this policy change. Finally, there is no effect on the odds that the mother is enrolled in school…”