This thread continues to shame me with the realization that I retained nothing from my econ classes.
That's ok I've got an MBA, and the econ in that class was shameful, but really honestly the decision process and technical aspect of an MBA is more about are things getting better or worst overall, in my sector, in my area and in my project. If things are looking like they are getting better in all of the above categories, great stay the course, do things look like they are going to get worst, hedge.
My impression of the empirical literature is that it points towards higher taxes putting a (slight) damper on growth rates. Maybe I'm out of date. If you think that's the case, feel free to throw me some citations to whatever the cutting edge articles are. I'm always happy to learn something new.
I don't have time to dig deep into the research (finals in a week) but most of what I have seen has been strong short term effects(mainly due to transferring where consumption/investment take place) and indistinguishable long term effects. Which makes a lot of sense if we think about complicating the model into one of general equilibrium.
Part of the problem can be exemplified in this paper by the Romer's. Wherein they estimate a linear 12 lag model for long term growth. Which basically means they're assuming that government spending is exogeneous and fixed, and that tax increases/decreases aren't related with budget deficits.
Basically that means if we cut taxes and enter a deficit which records a growth increase the tax cut will receive all the credit for the growth increase and the tax increase later to quell the deficit will receive all the credit for the reduction in growth. When it was neither the tax increase or the tax decrease that was the issue, it was the surplus/deficit (and the surplus/deficit were entirely short term)
Shock shock shock they find lower growth with tax increases, the effect decreases over time, and we should not be in any way surprised that their results produced this method. Especially because, surprise surprise, they find a break in the data right around 1980 wherein their estimates change in value significantly. Gee, i can't think of any reason why that would happen (besides tax policy suddenly becoming unhinged from the deficit and therefore prior tax policy and a structural break from a high growth to low growth period which we can't seems to explain)...
I am not even going to mention open economy problems.
Fake Edit: This paper, also supporting the idea that taxes reduce growth explains how the standard Solow model requires that long term growth is fixed.
I think the analysis is slightly better than the Romer paper but suffers similar endogeneity problems which are not fixed (especially in the cross country analysis)
@Goumindong Excellent, we're on the same page then. I'll just sit on my argument that taxes as we use them in reality cause a shift towards more leisure and more consumption now and be happy.
The response to that is "then why can't we see it?"
because total factor productivity is definitely neither constant nor universal, as long-term-growth theory w.r.t. the developing world tells us, for one thing. The other thing is that the neoclassical model doesn't include human capital investment, which is truly annoying to measure. More generally speaking: indices of investment are not, in fact, good measurements of the intuitive notion of deferred consumption (see: durable goods), and indices of employment are not, in fact, good measurements of the intuitive notion of labour input (see: household labour, as the feminist economists like to remind us). The empirics is a wash.
Solow is useful for thinking at the margin for a given society. But to be honest I was never asked to derive the savings rate from a household utility-maximizing framework in deriving Solow; rather taxes were stylized in as diminishing current output "automatically" (and I see that all three of us agree that lower investment -> lower growth anyway; we just disagree over whether taxes increase/decrease leisure/consumption/investment). Even if I was asked, I would have been (and am) leery of trying to build Solow from the ground-up, anyway. The Sraffians won the Cambridge capital debate and K is non-aggregatable from k to begin with, so why bother? It's all stylized relationships between macroscopic variables anyway.
There are parts of the stylized empirics which are easier, in a sense, like whether taxes reduce aftertax afterbenefit income, and there are parts which are more difficult, like whether taxes reduce long-term growth, and we draw from the former to the latter, not the other way around. Otherwise you make the mistake Samuelson did and conclude that the Soviet economy is an unparalleled success, simply because there is a possible world where consumption is long-term depressed in favor of investment.
Look think of a deadweight loss from micro, it reduced quantity right? But it also increased "money not spent". Well as soon as you say that you can get utility from that money not spent and can do other things with it, you will see why deadweight loss does not imply that less things happen. Because "too much money in my pocket" is deadweight loss.
...
Deadweight loss implies we are worse off, not that the particular direction of the worse off implies less investment.
...
DWL is always about utility. Utility is always about optimal allocation of goods, wherein there are tradeoffs between them.
It's precisely the same as the comparative statics when you want to find the effect of a tax on apples on the price/quantity of oranges in a Marshallian demand with three goods. The solution is indeterminate.
Unless you have some intuition regarding the substitution elasticities between apples, oranges, and the third good.
We don't actually disagree here regarding the argument-from-micro. I did say:
No, no. Not a pure exchange general-equilibrium economy with heterogeneous goods - in that, it is true that DWL has an indeterminate effect on recorded trade (=GDP). Second-best solution might easily entail a lot more movement of goods than the first-best, increasing trade whilst decreasing social welfare. And in a pure exchange economy, DWL does, yes, have agents simply stuffing their endowments into second-best uses.
...
Your argument exists insofar as homogeneities fail to hold, but that is true of virtually every macroeconomic statement, since Anything Goes in general eq.
The Solow model is traditionally constructed to have a national income, of which a savings rate (presumably from a separate, unseen intertemporal calculus) separates the national income into consumption and investment. The presumption is thus that the savings rate doesn't change significantly in national income, i.e., that there are precisely none of the unexpected interactions with second-best uses of agent's endowments and people don't save more when their income decreases. You can see this implicitly assumed in having s constant even as the economy grows, for one thing (if you permit s to be indeterminate in Y, there's no good reason to have Solow at all, dang it, never mind any tax-related model augmentations).
If taxes change long term growth rates then the world as we know it cannot exist. Since the world as we know it can and does exist then long term growth rates must not depend on taxes.
Re the Solow Model: In the Solow model S is not indeterminate in Y. It is fixed in Y as a function of the maximization conditions and is fixed independent of taxes. Growth is exogeneous and is captured by writing the production function as (AK^α(1+g)L^β) and it doesn't matter what you do growth tends back to g.
Convergence to the steady-state is faster with a higher savings rate.
We tend to think that there is still more K and L to 'mobilize', even in the developed world, so to speak - albeit in human-specific capital rather than fixed factories. Hence the predominance of Solow-type thinking. But we have poor theories of TFP change w.r.t. tax rates anyway. We stand on the shoulders of giants but the giants do not charge us rent for the privilege, so incentives and such are messy anyway. Innovators do not capture most of the surplus from innovation.
s is fixed independent of Y, yes. Taxes don't exist in the classical model. But it seems reasonable to suggest that agent's endowments in each period are Y+leisure and fixing s suggests that a tax encouraging a movement from Y to leisure reduces both consumption and investment.
The point is that even if we add taxes, g is still exogenous in the model, we always converge to the steady state growth rate regardless of the taxes. They can make us better or worse off, they cannot effect growth.
Once you start talking endogenous growth then you have the indeterminacy problem with the total effect on growth (and you're no longer talking Solow) unless your model formulation pretty much explicitly states that dg/dT is negative, which we don't have evidence for.
They can, if you're not at the steady-state. Convergence can clearly take much longer than the order of multiple decades, since it's been some four centuries since capital-accumulated industrial takeoff was a thing. We're still educating people "more" and such - more and more college, etc.
Well steady state growth does not imply that the level of capital is fixed(in whatever form) when we have population growth, but lets ignore that for a second.
So lets say we don't have steady state growth and are below steady state growth and so are catching up. And we tax which kicks us down the ladder and moves us further away from the steady state capital value. What happens? Convergence speeds up, not slows down. If convergence speeds up and not slows down and we are assuming that we are below the steady state then growth increases.
It would only be if we could cause an increase in investment by taxing that we would decrease growth (because we would be above the steady state capital value and would be converging to a lower growth).
Which is the opposite of your prediction, you said that we necessarily reduce consumption and investment and that growth decreases.
If convergence slowed down the further we got away from steady state and we could keep moving ourselves away from the steady state by taxing, then we would not have a fixed steady state growth rate because taxes would change the steady state growth rate.
But the model does have convergent steady state growth rate and that convergent steady state growth rate is not dependent on taxation.
0
Options
AtomikaLive fast and get fucked or whateverRegistered Userregular
I don't really get this obsession with executive compensation. I'm more concerned with corporations actually paying their taxes, and not fucking over the democratic process... not with how much money they're throwing at their executives. That's between the executives and the shareholders.
In any event, that money typically* finds its way back into the economy, to my understanding.
*Assuming they don't offshore it, which is probably something a guy making $52 million a month is doing.
I don't really get this obsession with executive compensation. I'm more concerned with corporations actually paying their taxes, and not fucking over the democratic process... not with how much money they're throwing at their executives. That's between the executives and the shareholders.
In any event, that money typically* finds its way back into the economy, to my understanding.
*Assuming they don't offshore it, which is probably something a guy making $52 million a month is doing.
The general idea is that many people don't believe that executives create a 300x multiple of economic value that the typical employee does. That said, on the scale that these huge corporations deal in, the money they spend on executives is just a drop in the bucket.
Right, I certainly understand the visceral appeal of being upset that CEOs are making more money than you think they ought to be making -- though there is a related legitimate issue of growing income inequality within a firm and in the larger economy, which historically has had definite macroeconomic consequences. I just view it as about as worthwhile as being upset about people engaging in a very broad definition of "usury" (read: "making money from money") which I don't feel is in itself an ethical dilemma.
Right, I certainly understand the visceral appeal of being upset that CEOs are making more money than you think they ought to be making -- though there is a related legitimate issue of growing income inequality within a firm and in the larger economy, which historically has had definite macroeconomic consequences. I just view it as about as worthwhile as being upset about people engaging in a very broad definition of "usury" (read: "making money from money") which I don't feel is in itself an ethical dilemma.
CEO compensation is basically the easiest way to illustrate the massive growth in inequality in places like the US.
Where do you think some of that gap is coming from?
The complaint isn't about 1 CEO going home with a lot of money. It's the 99% argument: The fact that the top 1% goes home with about 24% of the pie, the highest percentage since WW II, has to mean the other 99% are going home with less. If the company you work for makes more money, you won't be going home with any of it. The top 1% took 95%+ of the economic growth of 2011 home in the USA. Everyone works harder, in tougher times, for their benefits, and their benefits alone.
Re: RBC - well, yes, of course if you posit that shocks are "really" idiosyncratic that the onus is on you to start digging out firm-specific indices, not solve for unknown firm-specific shocks and shove the entire error there to give an arbitrary fit. That would be bad science, indeed.
No, my point was that there was no way you can not do that. Without an assumption that firm specific shocks sum to zero, there is an indeterminacy between specific and non-specific shocks within the model. If we solve the model assuming that firm specific shocks are zero then the sum that actually isn't zero will show up as non-specific shocks. This in and of itself is not a theoretical problem in the model. If firm specific shocks don't sum to zero they are by definition, not firm specific. We can, given that we have the firm specific shocks simply solve out the non specific part of it very simply. As such, all we have is a matter of definition for when firm specific shocks are firm specific or not.
The problem is that any fit of the final productivity is arbitrary. The fits are solutions to the model and the model is a fit and a time vector. Since the time vector of productivity fits is a function of the time vectors of GDP etc there is a necessary correlation between the two. Any regression you do between the two will fail on theoretical grounds. Any comparison of the structure of the productivity compared to the structure of the fit is a direct result as a solution to the model and similar fails on theoretical grounds. Cyclical structures in GDP will of course create cyclical structures in fitted productivity.
And that is what fitted productivity is. Its why we went to the unit root tests and don't look at productivity anymore [at least, iirc] because detecting a unit root in GPD[or not] actually means something whereas fitted productivity values exhibiting the behavior you want does not. At least, they don't provide information relating to the strength of the theory.
Yes, but it generates less demand than just tossing the money out in bales, and we accept the relatively lower hike in demand for supply-side reasons.
No, we accept the relatively lower hike in immediate demand due to uncertainty in income. At least, i cannot find any supply side reasons why spending tomorrow would be any different than spending today.
This is a general point. When the top 0.1% by income receive 50% of all capital gains - from a randomly Googled WSJ headline - we are reliant on a relatively small number of people to be doing capital investment well. So characterize the result as volatility in the marginal productivity of capital and there you go
The real return on investment is a financial and demand side effect, not real productivity effects. Real productivity can only be realized. The difference between real productivity and expected productivity would have effects that correct the demand side of the equation, not the supply side of the equation.
Well, no. In an economy with no state, no G and T, it's still possible to have distinguishably demand-side and supply-side problems. There could still be instances where potential output is shocked (earthquakes being the classic example) or where effective demand fails to exhaust actual output (rational nominal rigidities, etc.).
O.K. We can have supply and demand side problems, but we can't have demand side solutions unless we consider taxes to be demand side because they're the prices charged to consumers for government services.
That would be kinda dumb though, our model works fine and has desirable properties for the questions we're trying to answer.
Well, you don't fit without theoretical motivation. RBC models flail desperately to have any plausible across-the-board technology shocks at all, the solution in the case of ambiguity seems, quite reasonably, to assume that economy-wide non-specific technology shocks don't exist.
On infrastructure stimulus: so that you actually get a useful highway out of the effort, than a hole dug out and filled in again?
On supply-sides: the difference between real productivity and expected productivity is a demand-side effect? How does the archetypical earthquake shock fit into this classification again? Yes having your house collapse into rubble reduces your disposable income but we do not typically characterize such things as demand-side problems, for the dang good reason that it acts primarily by literally grinding capital to dust, not excess demand. Earthquakes make you poorer even if all markets clear instantaneously.
On taxes: argh. Alright, I emphasized that states do more than place restrictions on private activity, they also provide services that states are uniquely suited to provide, and so talking about a PPF that a state must consciously pursue is sensible. That is no call for now assuming that states only provide such services in a quasi-voluntary mutually-beneficial manner. The core of demand-side problems - that people aren't pursuing their bilateral trades because it really, really isn't in their individual interest to do so - still exists and would still militate against the civil service favouring the aggregate welfare against its own selfish best interest, that's why the notorious politics of the Treasury View exist at all. States really don't want to go into debt at such times, for the same reasons nobody else wants to either! The core of demand-side solutions, where we spend so that other people can spend, still exists quite well thank you very much.
I am asking that if we accept that there are supply and demand side problems in the manner of "potential output < potential potential output" and "effective demand < Y" respectively, that we characterize state policy toward either as supply and demand side solutions for the sake of consistency, if nothing else, which seems like a very desirable property indeed. We are quite willing to view deficit-funded tax cuts as demand-side policy even when it is pushed by Republicans, is it so objectionable to view tax-funded long-term state investments in education and infrastructure as supply-side policy merely because these are the kinds of policies favoured by people who are not Jude Wanniski?
I think the bulk of people's outrage at CEO pay comes from how severed it is from actual success. Its perfectly ok for a CEO to drive a company into the damn ground, fire half its employees and still walk off with a multimillion paycheck
I think the bulk of people's outrage at CEO pay comes from how severed it is from actual success. Its perfectly ok for a CEO to drive a company into the damn ground, fire half its employees and still walk off with a multimillion paycheck
Coming from the perspective of writing these agreements and plans (I usually represent the owners of the companies, not the executives) you don't normally approach compensation from any perspective other than "what will make him agree to work here, vs somewhere else" and "how can we align his interests with ours?" Severance, salary and short term cash bonuses go in the former. Equity and incentive comp goes to the latter. On the company side, we try to push as much comp into the incentive bucket as we can, but if you push too hard, the exec will just take another job. Since this is how the market is set, you can't go well outside the market terms and still get someone with a good record.
Have you done such work for companies with widely distributed ownership?
Yes, and to be honest, it's much more executive driven, since there is no party with real power or direct involvement/awareness to negotiate against the executive. In my experience, public companies with at least one large shareholder tend to be managed better than companies without a lead shareholder, and both are worse than private companies.
The model of widely distributed capital appointing senior management always seemed a little quirky to me, and I do wonder whether there's just some mysterious liability or tax reason that justifies moving away from more widespread bond-driven financing structures.
The model of widely distributed capital appointing senior management always seemed a little quirky to me, and I do wonder whether there's just some mysterious liability or tax reason that justifies moving away from more widespread bond-driven financing structures.
Do you mean shifting companies to be more debt financed rather than equity financed in their structure? If so from what I remember of my corporate finance, I think bankruptcy costs were one of the biggest if not the biggest downsides to having a heavily debt driven structure. A lot of it being in the cost of lawyering up and shin kicking the stakeholders have to go through to pick through the assets and sort through the liabilities.
Interest deductiblity of debt for corporate income taxes creates a big incentive to not be fully equity financed.
There was some theory (I'd have to look up the names of the theorems) that under certain idealized conditions it didn't make any difference to the value of the company switching levels between debt financing and equity financing. Its just those little things like taxes and bankruptcy costs that are assumed away have a big effect in the real world.
I find it hard to believe that sorting out competing interests by stakeholders on the occasion companies become completely insolvent is more costly than sorting out competing interests by stakeholders all the time, which is more or less what publicly traded equity entails.
You're thinking of Modigliani-Miller, and yes, that is why I suspected some legal or tax reason why companies might favour one or the other so apparently starkly.
I find it hard to believe that sorting out competing interests by stakeholders on the occasion companies become completely insolvent is more costly than sorting out competing interests by stakeholders all the time, which is more or less what publicly traded equity entails.
You're thinking of Modigliani-Miller, and yes, that is why I suspected some legal or tax reason why companies might favour one or the other so apparently starkly.
The material I read suggested that those costs where way higher than you would otherwise expect, because you have to go through the courts and everyone has to get a bunch of bankruptcy lawyers and expert witnesses and such.
Looking back through one of my books, the costs of financial distress in general, of which bankruptcy is a particular and nasty case, really put a damper a heavily debt based capital structures. Even the perceived risk of going into bankruptcy and incurring those costs will hurt the value of the company, which makes sense if you think about it.
Also, too heavy of a debt based capital structures allows for circumstances where the managers and equity holders can have some really perverse incentives at the expense of the bondholders and everyone else. One case was like an example I mentioned in another thread talking about Bain capital, where it would make sense to take on risky ventures that will have a good chance of bankrupting the company as long as the bulk or all of the downside will be borne by the bondholders and other employees. The distilled case being the CEO taking the company's treasury to Vegas and putting as much as possible on the roulette table at once. Savvy lenders are aware of these potential problems, so that will up the costs of borrowing for riskier, more debt laden companies.
Savant on
0
Options
spacekungfumanPoor and minority-filledRegistered User, __BANNED USERSregular
I work with heavily debt laden companies sometimes, and they are a nightmare because the terms of the numerous credit agreements wind up meaning that future corporate actions may have to be approved by a few banks. The end result is that the companies make inefficient decisions because they thread the maze of events of default.
I work with heavily debt laden companies sometimes, and they are a nightmare because the terms of the numerous credit agreements wind up meaning that future corporate actions may have to be approved by a few banks. The end result is that the companies make inefficient decisions because they thread the maze of events of default.
Yeah, bond covenants are one way for the banks and lenders in general to try to protect themselves from the aforementioned perverse incentives of management and equity holders. Which sounds like it can make an awful mess of things even when the company isn't close to defaulting.
0
Options
spacekungfumanPoor and minority-filledRegistered User, __BANNED USERSregular
I work with heavily debt laden companies sometimes, and they are a nightmare because the terms of the numerous credit agreements wind up meaning that future corporate actions may have to be approved by a few banks. The end result is that the companies make inefficient decisions because they thread the maze of events of default.
Yeah, bond covenants are one way for the banks and lenders in general to try to protect themselves from the aforementioned perverse incentives of management and equity holders. Which sounds like it can make an awful mess of things even when the company isn't close to defaulting.
Part of the problem is that they negotiate these strict covenants, but they don't really look at the companies they are lending to. So they push for one size fits all rules, but grandfather in existing agreements with executives. What you really need to control executive comp is an interested party that stays involved the whole way, which is generally only the case where there is a lead shareholder.
See, the perverse incentives problem binds just as well when it's a shareholder-elected board of directors appointing management. Only now it is always the case that the principal-agent problem is intrusive rather than when the company is plausibly in default. Obviously, investors can resolve some of these issues somewhat, so why can't bondholders do it?
0
Options
spacekungfumanPoor and minority-filledRegistered User, __BANNED USERSregular
See, the perverse incentives problem binds just as well when it's a shareholder-elected board of directors appointing management. Only now it is always the case that the principal-agent problem is intrusive rather than when the company is plausibly in default. Obviously, investors can resolve some of these issues somewhat, so why can't bondholders do it?
Interested bond holders could do it, no question. The issue is that the banks lend to so many companies that they don't have the capacity to stay involved with each company, let alone to have an idea of how an executive is doing. Contrast this with large equity stake holders like PE funds or large corporate groups that make strategic acquisitions, where the shareholder (even if they are not the holder of a majority interest) will stay involved the entire time. That is why I think equity structures with a lead investor make the most sense in this context, even though, in a perfect world, debt holders could accomplish the same thing. Incidentally, this is also why the widely held company model is the worst. Even in the debt holder example, you can wake the slumbering giant who has an incentive to keep exec comp reasonable, it is just hard. With wide spread ownership, there is no giant to wake and the shareholder appointed board will almost always be captured by the executives.
The "large equity stake holder" is presumably the wholly-owned management itself, merely indebted instead. Only now the identity of the lead investor is also the residual claimant from investment.
ronya on
0
Options
spacekungfumanPoor and minority-filledRegistered User, __BANNED USERSregular
The "large equity stake holder" is presumably the wholly-owned management itself, merely indebted instead. Only now the identity of the lead investor is also the residual claimant from investment.
Institutional investors are a special case, because so many of them subscribe to ISS, which serves as a watchdog on matters like executive comp. I personally disagree with a number of ISS's recommendations, and don't like butting heads with them but they are very influential on companies that have a lot of institutional investors who are known to follow their reccomendations.
I'm not sure that I understand your second point. Could you elaborate?
See, the perverse incentives problem binds just as well when it's a shareholder-elected board of directors appointing management. Only now it is always the case that the principal-agent problem is intrusive rather than when the company is plausibly in default. Obviously, investors can resolve some of these issues somewhat, so why can't bondholders do it?
Well, there are multiple principal-agent problems, and there are different specific ones which arise in financial distress between the equity side and the bondholders. The ones I was talking about in particular have management and equity in line with each other with their incentives, which are perverse from the standpoint of the lenders and the value of the firm as a whole. These arise in addition to other agency loss problems being considered, like between management and shareholders, which still exist under a leveraged firm.
There are ways for the bondholders to try to deal with their agency problems, namely with the bond covenants and credit agreements being talked about. But they are limited in what they can accomplish, and are more aimed at "don't go crazy with our money" rather than some other goal like maximizing profits or increasing share prices. Remember that the corporate bondholders under typical circumstances capture a lot less of the upside if the company does well than the shareholders do. The bondholders are more interested in a reduction of the downside, as less default risk means lowering yield which leads to higher prices on the bonds they currently hold.
Are you talking about something beyond that where the company would be essentially without the equity position, or some sort of hybrid between the debt and equity positions? If so I don't know how that would work necessarily. Otherwise we can just look that the pitfalls that face leveraged firms now.
Savant on
0
Options
spacekungfumanPoor and minority-filledRegistered User, __BANNED USERSregular
See, the perverse incentives problem binds just as well when it's a shareholder-elected board of directors appointing management. Only now it is always the case that the principal-agent problem is intrusive rather than when the company is plausibly in default. Obviously, investors can resolve some of these issues somewhat, so why can't bondholders do it?
Well, there are multiple principal-agent problems, and there are different specific ones which arise in financial distress between the equity side and the bondholders. The ones I was talking about in particular have management and equity in line with each other with their incentives, which are perverse from the standpoint of the lenders and the value of the firm as a whole. These arise in addition to other agency loss problems being considered, like between management and shareholders, which still exist under a leveraged firm.
There are ways to try to deal with it, namely the bondholder covenants and credit agreements being talked about. But they are limited in what they can accomplish, and are more aimed at "don't go crazy with our money" rather than some other goal like maximizing profits. Remember that the corporate bondholders under typical circumstances capture a lot less of the upside if the company does well than the shareholders do. The bondholders are more interested in a reduction of the downside, as less default risk means lowering yield which leads to higher prices on the bonds they currently hold.
Are you talking about something beyond that where the company would be essentially without the equity position, or some sort of hybrid between the bond and equity positions? If so I don't know how that would work necessarily. Otherwise we can just look that the pitfalls that face leveraged firms now.
I take Ronya's point to be that a firm managed from the debtholder's standpoint, which emphasizes mitigating downside risk vs taking risks for the hope at a large profit, coupled with the higher involvement of a majority stakeholder whose interests are not aligned with management's (absent arrangements designed to bring management's incentives in line with the debtholder's) could effectively reign in increasing executive compensation, particularly when companies are not doing well. It's an interesting idea, but given the choice, I think executives would flock to equity backed companies, forcing the debt based companies to either loosen up their compensation practices or accept less desirable executives.
See, the perverse incentives problem binds just as well when it's a shareholder-elected board of directors appointing management. Only now it is always the case that the principal-agent problem is intrusive rather than when the company is plausibly in default. Obviously, investors can resolve some of these issues somewhat, so why can't bondholders do it?
Well, there are multiple principal-agent problems, and there are different specific ones which arise in financial distress between the equity side and the bondholders. The ones I was talking about in particular have management and equity in line with each other with their incentives, which are perverse from the standpoint of the lenders and the value of the firm as a whole. These arise in addition to other agency loss problems being considered, like between management and shareholders, which still exist under a leveraged firm.
There are ways to try to deal with it, namely the bondholder covenants and credit agreements being talked about. But they are limited in what they can accomplish, and are more aimed at "don't go crazy with our money" rather than some other goal like maximizing profits. Remember that the corporate bondholders under typical circumstances capture a lot less of the upside if the company does well than the shareholders do. The bondholders are more interested in a reduction of the downside, as less default risk means lowering yield which leads to higher prices on the bonds they currently hold.
Are you talking about something beyond that where the company would be essentially without the equity position, or some sort of hybrid between the bond and equity positions? If so I don't know how that would work necessarily. Otherwise we can just look that the pitfalls that face leveraged firms now.
I take Ronya's point to be that a firm managed from the debtholder's standpoint, which emphasizes mitigating downside risk vs taking risks for the hope at a large profit, coupled with the higher involvement of a majority stakeholder whose interests are not aligned with management's (absent arrangements designed to bring management's incentives in line with the debtholder's) could effectively reign in increasing executive compensation, particularly when companies are not doing well. It's an interesting idea, but given the choice, I think executives would flock to equity backed companies, forcing the debt based companies to either loosen up their compensation practices or accept less desirable executives.
Oh I see now. Well, the problem with having management running things at the behest of the debtholder is that they don't really care how profitable a firm is as long as it is reliably not unprofitable after lending to them. The vision would be one of companies being hulking dinosaurs that would avoid even positive expectation risks in order to maintain a steady cashflow. I could see that maybe working somewhat serviceably for a limited number of businesses or industries, but I doubt that would be any good for the economy as a whole to be adopted in a widespread manner.
It may clamp down on excessive management compensation, but I think that might be somewhat of a tangential issue to be honest. If the CEO gets in because he has a lot of his buddies at the bank on the board of directors, what would stop them from rewarding him excessively as long as it doesn't appreciably increase the risk of financial distress? Why would they care if it hurt profitablity a bit and cut into the shareholder's dividends, or if they have to fire some folks to make room for his gold plated bathroom? It may help stop the golden parachute for the CEO that left behind a burning husk of a company, but that's not the only case where management can bilk the company for more than they are worth.
That would be a different manner from the debtholders having some limited measure of the control over management or having part of the board of directors, I could see that working fine. Then there are also more socialistic models to consider where the government or workers have control over the direction of management of the company, but I'm not really familiar with the theory behind that.
Posts
I don't have time to dig deep into the research (finals in a week) but most of what I have seen has been strong short term effects(mainly due to transferring where consumption/investment take place) and indistinguishable long term effects. Which makes a lot of sense if we think about complicating the model into one of general equilibrium.
Part of the problem can be exemplified in this paper by the Romer's. Wherein they estimate a linear 12 lag model for long term growth. Which basically means they're assuming that government spending is exogeneous and fixed, and that tax increases/decreases aren't related with budget deficits.
Basically that means if we cut taxes and enter a deficit which records a growth increase the tax cut will receive all the credit for the growth increase and the tax increase later to quell the deficit will receive all the credit for the reduction in growth. When it was neither the tax increase or the tax decrease that was the issue, it was the surplus/deficit (and the surplus/deficit were entirely short term)
http://elsa.berkeley.edu/~cromer/RomerDraft307.pdf
Shock shock shock they find lower growth with tax increases, the effect decreases over time, and we should not be in any way surprised that their results produced this method. Especially because, surprise surprise, they find a break in the data right around 1980 wherein their estimates change in value significantly. Gee, i can't think of any reason why that would happen (besides tax policy suddenly becoming unhinged from the deficit and therefore prior tax policy and a structural break from a high growth to low growth period which we can't seems to explain)...
I am not even going to mention open economy problems.
Fake Edit: This paper, also supporting the idea that taxes reduce growth explains how the standard Solow model requires that long term growth is fixed.
http://www.dartmouth.edu/~jskinner/documents/EngenSkinnerTaxEconGrowth.pdf
I think the analysis is slightly better than the Romer paper but suffers similar endogeneity problems which are not fixed (especially in the cross country analysis)
because total factor productivity is definitely neither constant nor universal, as long-term-growth theory w.r.t. the developing world tells us, for one thing. The other thing is that the neoclassical model doesn't include human capital investment, which is truly annoying to measure. More generally speaking: indices of investment are not, in fact, good measurements of the intuitive notion of deferred consumption (see: durable goods), and indices of employment are not, in fact, good measurements of the intuitive notion of labour input (see: household labour, as the feminist economists like to remind us). The empirics is a wash.
Solow is useful for thinking at the margin for a given society. But to be honest I was never asked to derive the savings rate from a household utility-maximizing framework in deriving Solow; rather taxes were stylized in as diminishing current output "automatically" (and I see that all three of us agree that lower investment -> lower growth anyway; we just disagree over whether taxes increase/decrease leisure/consumption/investment). Even if I was asked, I would have been (and am) leery of trying to build Solow from the ground-up, anyway. The Sraffians won the Cambridge capital debate and K is non-aggregatable from k to begin with, so why bother? It's all stylized relationships between macroscopic variables anyway.
There are parts of the stylized empirics which are easier, in a sense, like whether taxes reduce aftertax afterbenefit income, and there are parts which are more difficult, like whether taxes reduce long-term growth, and we draw from the former to the latter, not the other way around. Otherwise you make the mistake Samuelson did and conclude that the Soviet economy is an unparalleled success, simply because there is a possible world where consumption is long-term depressed in favor of investment.
Unless you have some intuition regarding the substitution elasticities between apples, oranges, and the third good.
We don't actually disagree here regarding the argument-from-micro. I did say:
The Solow model is traditionally constructed to have a national income, of which a savings rate (presumably from a separate, unseen intertemporal calculus) separates the national income into consumption and investment. The presumption is thus that the savings rate doesn't change significantly in national income, i.e., that there are precisely none of the unexpected interactions with second-best uses of agent's endowments and people don't save more when their income decreases. You can see this implicitly assumed in having s constant even as the economy grows, for one thing (if you permit s to be indeterminate in Y, there's no good reason to have Solow at all, dang it, never mind any tax-related model augmentations).
http://noahpinionblog.blogspot.com/2011/07/real-laffer.html
If taxes change long term growth rates then the world as we know it cannot exist. Since the world as we know it can and does exist then long term growth rates must not depend on taxes.
Re the Solow Model: In the Solow model S is not indeterminate in Y. It is fixed in Y as a function of the maximization conditions and is fixed independent of taxes. Growth is exogeneous and is captured by writing the production function as (AK^α(1+g)L^β) and it doesn't matter what you do growth tends back to g.
... which is my own suspicion, anyway.
We tend to think that there is still more K and L to 'mobilize', even in the developed world, so to speak - albeit in human-specific capital rather than fixed factories. Hence the predominance of Solow-type thinking. But we have poor theories of TFP change w.r.t. tax rates anyway. We stand on the shoulders of giants but the giants do not charge us rent for the privilege, so incentives and such are messy anyway. Innovators do not capture most of the surplus from innovation.
s is fixed independent of Y, yes. Taxes don't exist in the classical model. But it seems reasonable to suggest that agent's endowments in each period are Y+leisure and fixing s suggests that a tax encouraging a movement from Y to leisure reduces both consumption and investment.
Once you start talking endogenous growth then you have the indeterminacy problem with the total effect on growth (and you're no longer talking Solow) unless your model formulation pretty much explicitly states that dg/dT is negative, which we don't have evidence for.
So lets say we don't have steady state growth and are below steady state growth and so are catching up. And we tax which kicks us down the ladder and moves us further away from the steady state capital value. What happens? Convergence speeds up, not slows down. If convergence speeds up and not slows down and we are assuming that we are below the steady state then growth increases.
It would only be if we could cause an increase in investment by taxing that we would decrease growth (because we would be above the steady state capital value and would be converging to a lower growth).
Which is the opposite of your prediction, you said that we necessarily reduce consumption and investment and that growth decreases.
If convergence slowed down the further we got away from steady state and we could keep moving ourselves away from the steady state by taxing, then we would not have a fixed steady state growth rate because taxes would change the steady state growth rate.
But the model does have convergent steady state growth rate and that convergent steady state growth rate is not dependent on taxation.
In any event, that money typically* finds its way back into the economy, to my understanding.
*Assuming they don't offshore it, which is probably something a guy making $52 million a month is doing.
The general idea is that many people don't believe that executives create a 300x multiple of economic value that the typical employee does. That said, on the scale that these huge corporations deal in, the money they spend on executives is just a drop in the bucket.
CEO compensation is basically the easiest way to illustrate the massive growth in inequality in places like the US.
Where do you think some of that gap is coming from?
No, my point was that there was no way you can not do that. Without an assumption that firm specific shocks sum to zero, there is an indeterminacy between specific and non-specific shocks within the model. If we solve the model assuming that firm specific shocks are zero then the sum that actually isn't zero will show up as non-specific shocks. This in and of itself is not a theoretical problem in the model. If firm specific shocks don't sum to zero they are by definition, not firm specific. We can, given that we have the firm specific shocks simply solve out the non specific part of it very simply. As such, all we have is a matter of definition for when firm specific shocks are firm specific or not.
The problem is that any fit of the final productivity is arbitrary. The fits are solutions to the model and the model is a fit and a time vector. Since the time vector of productivity fits is a function of the time vectors of GDP etc there is a necessary correlation between the two. Any regression you do between the two will fail on theoretical grounds. Any comparison of the structure of the productivity compared to the structure of the fit is a direct result as a solution to the model and similar fails on theoretical grounds. Cyclical structures in GDP will of course create cyclical structures in fitted productivity.
And that is what fitted productivity is. Its why we went to the unit root tests and don't look at productivity anymore [at least, iirc] because detecting a unit root in GPD[or not] actually means something whereas fitted productivity values exhibiting the behavior you want does not. At least, they don't provide information relating to the strength of the theory.
No, we accept the relatively lower hike in immediate demand due to uncertainty in income. At least, i cannot find any supply side reasons why spending tomorrow would be any different than spending today.
The real return on investment is a financial and demand side effect, not real productivity effects. Real productivity can only be realized. The difference between real productivity and expected productivity would have effects that correct the demand side of the equation, not the supply side of the equation.
O.K. We can have supply and demand side problems, but we can't have demand side solutions unless we consider taxes to be demand side because they're the prices charged to consumers for government services.
That would be kinda dumb though, our model works fine and has desirable properties for the questions we're trying to answer.
On infrastructure stimulus: so that you actually get a useful highway out of the effort, than a hole dug out and filled in again?
On supply-sides: the difference between real productivity and expected productivity is a demand-side effect? How does the archetypical earthquake shock fit into this classification again? Yes having your house collapse into rubble reduces your disposable income but we do not typically characterize such things as demand-side problems, for the dang good reason that it acts primarily by literally grinding capital to dust, not excess demand. Earthquakes make you poorer even if all markets clear instantaneously.
On taxes: argh. Alright, I emphasized that states do more than place restrictions on private activity, they also provide services that states are uniquely suited to provide, and so talking about a PPF that a state must consciously pursue is sensible. That is no call for now assuming that states only provide such services in a quasi-voluntary mutually-beneficial manner. The core of demand-side problems - that people aren't pursuing their bilateral trades because it really, really isn't in their individual interest to do so - still exists and would still militate against the civil service favouring the aggregate welfare against its own selfish best interest, that's why the notorious politics of the Treasury View exist at all. States really don't want to go into debt at such times, for the same reasons nobody else wants to either! The core of demand-side solutions, where we spend so that other people can spend, still exists quite well thank you very much.
I am asking that if we accept that there are supply and demand side problems in the manner of "potential output < potential potential output" and "effective demand < Y" respectively, that we characterize state policy toward either as supply and demand side solutions for the sake of consistency, if nothing else, which seems like a very desirable property indeed. We are quite willing to view deficit-funded tax cuts as demand-side policy even when it is pushed by Republicans, is it so objectionable to view tax-funded long-term state investments in education and infrastructure as supply-side policy merely because these are the kinds of policies favoured by people who are not Jude Wanniski?
Coming from the perspective of writing these agreements and plans (I usually represent the owners of the companies, not the executives) you don't normally approach compensation from any perspective other than "what will make him agree to work here, vs somewhere else" and "how can we align his interests with ours?" Severance, salary and short term cash bonuses go in the former. Equity and incentive comp goes to the latter. On the company side, we try to push as much comp into the incentive bucket as we can, but if you push too hard, the exec will just take another job. Since this is how the market is set, you can't go well outside the market terms and still get someone with a good record.
Yes, and to be honest, it's much more executive driven, since there is no party with real power or direct involvement/awareness to negotiate against the executive. In my experience, public companies with at least one large shareholder tend to be managed better than companies without a lead shareholder, and both are worse than private companies.
Do you mean shifting companies to be more debt financed rather than equity financed in their structure? If so from what I remember of my corporate finance, I think bankruptcy costs were one of the biggest if not the biggest downsides to having a heavily debt driven structure. A lot of it being in the cost of lawyering up and shin kicking the stakeholders have to go through to pick through the assets and sort through the liabilities.
Interest deductiblity of debt for corporate income taxes creates a big incentive to not be fully equity financed.
There was some theory (I'd have to look up the names of the theorems) that under certain idealized conditions it didn't make any difference to the value of the company switching levels between debt financing and equity financing. Its just those little things like taxes and bankruptcy costs that are assumed away have a big effect in the real world.
You're thinking of Modigliani-Miller, and yes, that is why I suspected some legal or tax reason why companies might favour one or the other so apparently starkly.
The material I read suggested that those costs where way higher than you would otherwise expect, because you have to go through the courts and everyone has to get a bunch of bankruptcy lawyers and expert witnesses and such.
Looking back through one of my books, the costs of financial distress in general, of which bankruptcy is a particular and nasty case, really put a damper a heavily debt based capital structures. Even the perceived risk of going into bankruptcy and incurring those costs will hurt the value of the company, which makes sense if you think about it.
Also, too heavy of a debt based capital structures allows for circumstances where the managers and equity holders can have some really perverse incentives at the expense of the bondholders and everyone else. One case was like an example I mentioned in another thread talking about Bain capital, where it would make sense to take on risky ventures that will have a good chance of bankrupting the company as long as the bulk or all of the downside will be borne by the bondholders and other employees. The distilled case being the CEO taking the company's treasury to Vegas and putting as much as possible on the roulette table at once. Savvy lenders are aware of these potential problems, so that will up the costs of borrowing for riskier, more debt laden companies.
Yeah, bond covenants are one way for the banks and lenders in general to try to protect themselves from the aforementioned perverse incentives of management and equity holders. Which sounds like it can make an awful mess of things even when the company isn't close to defaulting.
Part of the problem is that they negotiate these strict covenants, but they don't really look at the companies they are lending to. So they push for one size fits all rules, but grandfather in existing agreements with executives. What you really need to control executive comp is an interested party that stays involved the whole way, which is generally only the case where there is a lead shareholder.
Interested bond holders could do it, no question. The issue is that the banks lend to so many companies that they don't have the capacity to stay involved with each company, let alone to have an idea of how an executive is doing. Contrast this with large equity stake holders like PE funds or large corporate groups that make strategic acquisitions, where the shareholder (even if they are not the holder of a majority interest) will stay involved the entire time. That is why I think equity structures with a lead investor make the most sense in this context, even though, in a perfect world, debt holders could accomplish the same thing. Incidentally, this is also why the widely held company model is the worst. Even in the debt holder example, you can wake the slumbering giant who has an incentive to keep exec comp reasonable, it is just hard. With wide spread ownership, there is no giant to wake and the shareholder appointed board will almost always be captured by the executives.
The "large equity stake holder" is presumably the wholly-owned management itself, merely indebted instead. Only now the identity of the lead investor is also the residual claimant from investment.
Institutional investors are a special case, because so many of them subscribe to ISS, which serves as a watchdog on matters like executive comp. I personally disagree with a number of ISS's recommendations, and don't like butting heads with them but they are very influential on companies that have a lot of institutional investors who are known to follow their reccomendations.
I'm not sure that I understand your second point. Could you elaborate?
Well, there are multiple principal-agent problems, and there are different specific ones which arise in financial distress between the equity side and the bondholders. The ones I was talking about in particular have management and equity in line with each other with their incentives, which are perverse from the standpoint of the lenders and the value of the firm as a whole. These arise in addition to other agency loss problems being considered, like between management and shareholders, which still exist under a leveraged firm.
There are ways for the bondholders to try to deal with their agency problems, namely with the bond covenants and credit agreements being talked about. But they are limited in what they can accomplish, and are more aimed at "don't go crazy with our money" rather than some other goal like maximizing profits or increasing share prices. Remember that the corporate bondholders under typical circumstances capture a lot less of the upside if the company does well than the shareholders do. The bondholders are more interested in a reduction of the downside, as less default risk means lowering yield which leads to higher prices on the bonds they currently hold.
Are you talking about something beyond that where the company would be essentially without the equity position, or some sort of hybrid between the debt and equity positions? If so I don't know how that would work necessarily. Otherwise we can just look that the pitfalls that face leveraged firms now.
I take Ronya's point to be that a firm managed from the debtholder's standpoint, which emphasizes mitigating downside risk vs taking risks for the hope at a large profit, coupled with the higher involvement of a majority stakeholder whose interests are not aligned with management's (absent arrangements designed to bring management's incentives in line with the debtholder's) could effectively reign in increasing executive compensation, particularly when companies are not doing well. It's an interesting idea, but given the choice, I think executives would flock to equity backed companies, forcing the debt based companies to either loosen up their compensation practices or accept less desirable executives.
Oh I see now. Well, the problem with having management running things at the behest of the debtholder is that they don't really care how profitable a firm is as long as it is reliably not unprofitable after lending to them. The vision would be one of companies being hulking dinosaurs that would avoid even positive expectation risks in order to maintain a steady cashflow. I could see that maybe working somewhat serviceably for a limited number of businesses or industries, but I doubt that would be any good for the economy as a whole to be adopted in a widespread manner.
It may clamp down on excessive management compensation, but I think that might be somewhat of a tangential issue to be honest. If the CEO gets in because he has a lot of his buddies at the bank on the board of directors, what would stop them from rewarding him excessively as long as it doesn't appreciably increase the risk of financial distress? Why would they care if it hurt profitablity a bit and cut into the shareholder's dividends, or if they have to fire some folks to make room for his gold plated bathroom? It may help stop the golden parachute for the CEO that left behind a burning husk of a company, but that's not the only case where management can bilk the company for more than they are worth.
That would be a different manner from the debtholders having some limited measure of the control over management or having part of the board of directors, I could see that working fine. Then there are also more socialistic models to consider where the government or workers have control over the direction of management of the company, but I'm not really familiar with the theory behind that.