Krugman on Shaping Expectations
I’m not going to say this even qualifies as a “Kontradiction,” but it’s close. By now, every serious geeconosphere wonk should know that Paul Krugman’s recommendation to Japan back in the day, was that its central bank should credibly commit to future inflation. That would cause the public to expect a lower purchasing power of money in the future, meaning that even if nominal interest rates were stuck at zero, the real interest rate would be negative, inducing the public to go spend. Here’s Krugman summarizing his case in November 2009:
We’re in a liquidity trap, with interest rates up against the zero bound. This means that conventional monetary policy isn’t sufficient. What should we do?
The first-best answer — that is, the answer that economic models, like my old Japan’s trap analysis, suggest would be optimal — would be to credibly commit to higher inflation, so as to reduce real interest rates.
But the key thing to recognize about this answer is that it’s all about expectations — the central bank only has traction over expected inflation to the extent that it can convince people that it will deliver that inflation after the liquidity trap is over. So to make this policy work you have to (i) convince current policymakers that it’s the right answer (ii) Make that argument persuasive enough that it will guide the actions of future policymakers (iii) Convince investors, consumers, and firms that you have in fact achieved (i) and (ii).
In reality, we haven’t even gotten anywhere near (i): the conventional wisdom is still that any rise in expected inflation above 2 percent is a bad thing, when it’s actually good.
So some readers have asked why I’m not making the same arguments for America now that I was making for Japan a decade ago. The answer is that I don’t think I’ll get anywhere, at least not until or unless the slump goes on for a long time.
OK so back then, Krugman was saying the “first best” answer, based on a rigorous economic model, and what he himself told the Japanese, was to manipulate the public’s expectations about future policy. But, since that had some practical roadblocks, a fall-back was to rely on deficit spending today.
Contrast that treatment with Krugman’s analysis of a recent paper promoting supply-side solutions:
There’s a new Vox paper suggesting that supply-side policies — structural reform that makes workers more productive, prices more flexible, or whatever — may be the answer to a liquidity trap. Some people may seize on this claim. So it’s worth pointing out that a casual interpretation here gets it all wrong.
Here’s what we already know: in the looking-glass world of the liquidity trap, supply-side improvements are generally counter-productive…
So how does the Vox paper get a different story? By assuming that the payoff to supply-side improvements doesn’t come until the economy has already spontaneously emerged from the liquidity trap, and so these perversities are gone. The idea then is that consumers, rationally perceiving this future improvement, feel richer and spend more now.
Oh, Kay. It might be true, although it depends on consumers being very well informed. I don’t know anyone in the Princeton economics department, let alone Main Street USA, who does family budgeting based on perceptions about how reforms in tax law or labor regulation will affect incomes five years from now. But whatever.
Clearly not an actual contradiction, and perhaps not even a Krugman Kontradiction. But it seems Krugman is much more sympathetic to the idea of central bankers promising higher future inflation, than to the idea of Treasury officials promising lower future taxes. It seems that Main Street USA, when doing the family budgeting, must read more Scott Sumner than Art Laffer.
In Scott Sumner’s mind, the American family, while doing its budget, also watches NGDP like a hawk in order to decide whether to spend more money. “Oh, look dear, NGDP is coming back up. Time to spend again, like the robot econotechnocrats want us to!”
Silas that’s a bald-faced lie. Scott doesn’t say the American family looks at current NGDP. Rather, Scott says that husbands and wives argue over the path of future NGDP.
If we can’t have integrity on the internet, we are mere apes.
This is just one reason why this blog is so awesome. Hilarious.
So expectations play no role in people’s economic planning and when we see deviations in NGDP-trends reflected in such indicators as stock-prices , interest rates, unemployment rates etc etc people just don’t notice ?
People don’t form expectations about NGDP, no. Most people never use the concept and have never heard the abbreviation. Even intelligent business owners don’t have a reason to care about it: they care about the potential demand fro products *they sell* or could be selling, not “total purchases across the entire economy”.
They may not explicitly be thinking about NGDP – so that extent you are right. But surely what really matters is not that literalistic interpretations but rather whether expectations that are ultimately based on NGDP-trend will affect people’s behavior.
Krugman obviously wants to use expectations when it suits him and mock it as absurd when it suits him.
What I found really incorrect in his post was the statement that “supply-side improvements are generally counter-productive…” in a liquidity trap, when surely it is the need for supply-side factors that led to the liquidity trap in the first place, and the lack of them that keeps it in place.
But surely what really matters is not that literalistic interpretations but rather whether expectations that are ultimately based on NGDP-trend will affect people’s behavior.
You mean like an individual’s expectations of future prices for the particular markets they intend to sell into, thus enabling them to calculate current prices they are willing to pay?
You don’t need NGDP targeting for that.
What Major_Freedom said. Also, do not equate “expectation of the sum” (NGDP) with “sum of the expectations”. The latter does not somehow add up to the former in any useful way.
If the stock market was tanking and unemployment was rising because businesses were cutting back on investment as a result of a future anticipated fall in aggregate demand wouldn’t that have a direct effect on “an individual’s expectations of future prices for the particular markets they intend to sell into” ?
By extension: Stabling AD makes economic calculation more reliable since it eliminates fluctuations that reflect changes in the demand for money rather than changes in underlying consumer preference.
@Rob: because businesses were cutting back on investment as a result of a future anticipated fall in aggregate demand …
*bzzt* Again, businesses don’t make decisions this way.
If the stock market was tanking and unemployment was rising because businesses were cutting back on investment as a result of a future anticipated fall in aggregate demand
Rob, you just used the premise “individuals plan using NGDP” in your argument.
That premise IS the claim being argued over!
Businesses don’t cut back on investment based on falling aggregate demand. They cut back on investment when there is an expected decline in future profitability, which is a function of both output prices AND input prices, and not only that, but for specific markets as well.
If the stock market tanks, it won’t be because of falling aggregate demand. The stock market falls along with aggregate demand because of a third reason that is neither stock market prices nor aggregate demand.
By extension: Stabling AD makes economic calculation more reliable since it eliminates fluctuations that reflect changes in the demand for money rather than changes in underlying consumer preference.
That again presumes the premise “individuals plan according to NGDP” to be true, when that is the very thing being argued over.
…surely it is the need for supply-side factors that led to the liquidity trap in the first place, and the lack of them that keeps it in place.
Bingo.
Austrian Economics is based on the strong assumption that agents base their decisions on expectations about the future.
As you state – investment decisions are based on expected future profits (themselves based on expected future prices). If an expected supply-shock (say a war that may disrupt production of a raw material) causes expectation of future profits to diminish then some business may well decide it is safer to hold cash rather than to invest. Demand for money has increased. Aggregate demand has fallen. NGDP-forecasts will reflect this.
The economy will need to find a new equilibrium based on the supply-shock , and the demand-side changes it triggered. In the view of people who favor NGDP-targeting it will be much easier to find that new equilibrium if the AD-shock can be neutralized via an expansion in the money supply to meet the increased demand I have yet to see an argument that holds up as to why this kind of expansion is in any way related to the kind of expansions that underlie the ABCT where the expansion is not in response to an increased demand.
Rob, unless that injection of money removes the *reason* people have to be uncertain about future profitability, then it’s just papering over (pardon the pun) the problem. If I’m not sure I’ll be able to buy my inputs next week, then for the government to make people buy lots of crap they wouldn’t have otherwise bought … doesn’t change this at all.
There are reasons, good reasons, for holding on to cash, and any attempt to make people spend more just compounds the problem while goosing a misleading metric.
Do you understand why it wouldn’t be a good idea to make each person hit a spending target each hour? Then you understand the problems in artificially moving up purchases.
I think this touches on a lesson that many economists fail to grasp.
Only that which action can be directed towards, is relevant in economic science. NGDP is an (estimated) aggregate statistic that no actor economizes or acts upon.
It makes no difference to the entrepreneur whether total spending is higher or lower. He is only (rightfully) concerned about output prices and input prices in a particular market.
The only parties who would have any interest in such aggregate statistics are those attempting to centrally plan the whole economy, or help those who are attempting to centrally plan the whole economy.
NGDP is an EFFECT of individual choices in output prices and input prices. It’s not a CAUSAL factor in determining whether employment or production goes up or down.
To believe that forcefully raising or lowering NGDP through inflation can somehow benefit everyone, is like believing effects can be controlled in order to determine causes.
Suppose I had the taxation power to steal everyone’s money. Suppose it totalled $1 trillion, for an annual NGDP of $15 trillion.
Suppose I then spent $1 trillion, then taxed the receivers $1 trillion, then spent $1 trillion again, over and over, 15 times in the following year, on puppy chow. Suppose I also printed some new money, and spent and respent an additional 0.05 x $15 trillion = $750 billion on puppy chow in the current year.
NGDP grew by 5% this year, but only 5% of the world’s population seems to be upright.
What went wrong? I succeeded in raising NGDP by 5%, so the economy should have grown.
Oh that’s right, I almost forgot! NGDP is an effect concept, not a cause concept.
Don’t let the NGDP guys into a hospital. They might start pulling patient’s hair out thinking it will cure their cancer.
Good points. Also, the models behind NGDP targeting can’t explain why it would be a bad idea to hit the target by rounding up random citizens and forcing them to spend $X within the next hour.
It’s because they can’t grasp why it might be *good* not to blow all your money according to the CB-dictated schedule.
There’s really no limit to the absurdity you have to endorse once you deem NGDP to be the key to economic goodness: how about ordering people to buy services they normally provide themselves (cooking, cleaning, “stimulation”, etc)?
There’s really no limit to the absurdity you have to endorse once you deem NGDP to be the key to economic goodness: how about ordering people to buy services they normally provide themselves (cooking, cleaning, “stimulation”, etc)?
Imagine the jump in reported spending that could be added to NGDP by legalizing all purchases of drugs, prostitutes, and politicians (redundant?).
“Don’t let the NGDP guys into a hospital. They might start pulling patient’s hair out thinking it will cure their cancer.”
That’s the best analogy for NGDP targeting that I’ve heard so far
Maybe the government should issue Krugmanite “scrip” to the unemployed for dirty sloppy lap dances at Detroit “adult entertainment” clubs to jack up NGDP AND to help the single moms get and have more “aggregate demand”.
Bob_Roddis, this is a family site.
And the USA is a family country and would never do such a thing. But Canada might.
http://www.washingtonpost.com/wp-dyn/articles/A35851-2004Dec4.html
Silas wrote:
Even intelligent business owners don’t have a reason to care about it: they care about the potential demand fro products *they sell* or could be selling, not “total purchases across the entire economy”.
And let us not forget, everyone, that–contrary to repeated claims by everyone from Scott Sumner down to Matt Yglesias–NGDP is NOT the same thing as “total spending.”
You make some goods points.
The aim of the expansion of the money supply is not to get people to buy stuff they don’t really want but to get back in the most optimal fashion to an equilibrium where markets are clearing and the structure of production reflects consumer preferences.
Its fairly easy to prove that this will take longer via the route of all prices adjusting to the change in the demand for money than if the supply of money increases to meet the new demand.
Having a system that tends to stabilize AD (and remember I favor a free-banking solution and distrusts a CB-based solution probably as much as you) also tends to “removes the reason people have to be uncertain about future profitability” when that uncertainty is only based on concerns about the future value of money (which is just a veil) , and gets focus instead on the real demand and supply factors for real goods that will require genuine market adjustments.
Rob the uncertainty is about future ability to satisfy demand. Uncertainty about the future value of money is one way this is *reflected*, but it is not the *driver* of the process, and so is not something that can be smoothed away via monetary policy.
The aim of the expansion of the money supply is not to get people to buy stuff they don’t really want but to get back in the most optimal fashion to an equilibrium where markets are clearing and the structure of production reflects consumer preferences.
Okay, but part of those preferences are for things that have disrupted production structures, and thus require extensive re-organization involving entrepreneurial insight. To the extent entrepreneurs fail at realigning production structures, money (as something more certain to be saleable) becomes more valuable, and this is an important signal.
What proof are you referring to that it’s faster for new money to adjust prices up than for them to fall? I suspect that the proof has a hidden assumption about the goodness of paying off debts and not defaulting, but in my opinion, defaults are a signal of too much reliance on debt funding, a signal that is obscured by avoiding defaults via new money.
Nominal GDP should be a reasonable proxy for MV and this fell in 2008.
The object is to adjust the money supply to match the current price level – as an alternative to allowing all prices to fall to match the money level.
I agree that many of the current economic problems may be structural in nature (and one of the main structural problems may in fact be the high-degree of price-stickiness we face) so money elasticity is not a “universal cure”.
However I feel that it is important to recognize that CB actions not only in over expanding the money supply in the boom , but then allowing it to over-contract during the bust are responsible for creating and maintaining the current -recession.
Rob,
Are you saying the Fed allowed the money supply to over-contract in response to this recession?
Yes, I would argue that back in 2008 had we had a free banking system then they would have expanded the money supply in response to the increased demand for money.
The fed failed to do this and made the recession much worse than it needed to be.
Yes, I would argue that back in 2008 had we had a free banking system then they would have expanded the money supply in response to the increased demand for money.
The fed failed to do this to a sufficient degree and made the recession much worse than it needed to be.
Well, Rob, the Fed still hasn’t satisfied *my* demand for money, so it needs to print up some more money and loan it to me at 0% interest.
Rob,
Ok, I can see arguing the Fed didn’t do enough to expand the money supply – to a suffient degree – (though I would disagree). But, I don’t see how you could argue that they allowed it to contract.
http://research.stlouisfed.org/fredgraph.png?g=3qw
And, I don’t know how one can tell what would have happened to the money supply under ‘free banking’ (why it wouldn’t have been just as likely, or more likely, that it would have fallen…)
Richard,
Tthe fed allowed MV to fall. They increased M f, but not enough to offset the fall in V.
And in regards to free banking – their is a whole theory behind this (See Selgin’s Theory of Free Banking http://oll.libertyfund.org/?option=com_staticxt&staticfile=show.php%3Ftitle=2307)
The short version is that when Free Banks face an increased demand for money they can safely expand the supply to meet it – and this would have stabilized MV in a way that the fed failed to do.
Rob,
I read “money supply” to mean the stock of money (M).
I confess I have not read Selgin’s book (yet), but the idea that free banks, w/o gov’t intervention, can expand to maintain “MV” *in all cases* strikes me as a bit much.
Selgin’s book is worth reading.
Free banks are not a panacea – they are just part of an overall free market solution – but to me the theory makes sense.
I do plan to read Selgin’s book soon, if only because I am still left unpersuaded by the free banking theory as presented in articles he’s written in response to 100%ers. (I am sure there is much to it he could not include in shorter articles).
BTW, forgive my igonorance, but how did you determine that MV fell?
Nominal GDP should be a reasonable proxy for MV and this fell in 2008.
(Re posted as the comment appeared in wrong place first time)
Oh, we’re merely neural network-based nature-biased high-accuracy Bayesian estimators?
Sorry, channeling Daniel_Kuehn’s obliviousness for a second there…
Maybe not. But I wonder if you could get a bunch of interacting neural nets to go into something like a recession!
Woah, did Krugman just say that inflation can work because most families are stupid, and that supply-side-ism could only work if people were not stupid? In other words, the more stupid that people are, in Krugman’s mind, the better Keynesianism can work?
Putting aside that incredible admission of the depravity of Keynesianism in that it preys on people’s ignorance and manipulates them like cattle, even by his own logical constraints, there is no reason to accept his position.
His position is that people are smart enough to anticipate rising prices in the future, given the central bank committing to a future inflation target, thus become willing to pay higher prices for things now, which by virtue of exchanges taking place, will allegedly “get the economy moving again.”
But if people are smart enough to anticipate rising prices in the future, then surely they would also be smart enough to know what falling prices means, and therefore given the central bank committing to a future stoppage of inflation, people will become willing to pay lower prices for things now, which would also “get the economy moving again.”
Seriously, what’s so bad about having a smaller than arbitrary amount of $X money in circulation, rather than a higher than arbitrary amount of $X money in circulation? Who would actually believe that people living in the year 1890 could have been as prosperous as we are in 2011 if they only had the same quantity of money in circulation then as we do now? Nobody who thinks for more than 5 seconds would believe that. Well then, why should we have a quantity of money in circulation now that would resemble what would exist in the future relative to now? It would be like us living in 1890 and we’re all believing that we can have airplanes and cars from the year 2011 if only we convinced the central bank to create more money.
If it’s absurd to think this, then it’s also absurd to think that we should call for more money now that would have otherwise resembled the quantity in 2200 AD, or 2100 AD, or 2050 AD, or 2020 AD, or any other quantity that would have otherwise existed in the future, but increased sooner today.
If there are idle resources or unemployment, on the basis that the Fed is promising inflation it can’t deliver, which is periodic in every deflationary bust, which then makes owners of such resources “hold out” for anticipated higher prices, then surely with a policy of price deflation, the owners of resources and labor will sell sooner rather than later, given that prices in the downward direction, like the upward direction, aren’t perfectly flexible either.
I always thought Herr Krugman thought everyone was stupid, except him of course.
That’s one of the most clear Krugtradictions, at least of the recent ones you’ve pointed out.
Wenzel just posted an actual contradiction from Krugman on EPJ.
Anybody can do that.
That’s funny, and tragic.
You crack me up
definately a Kontradiction – the only exception I can think of is that proposed future tax reform is always subject to the whims of future politicians, whereas it’s a pretty safe bet that Bernanke won’t be slowing down the printing press anytime soon…
Cold fusion will have to wait.
As highly as DK thinks of Murphy, it’s not Murphy that is holding up the viability of cold fusion.
Mammoth wrote: “Cold fusion will have to wait.”
But Mammoth, that’s optimal. If I delivered a ridiculously cheap energy source to the world right now, it might get worse than 1933. Didn’t you read Krugman’s quotation above? I promise–scout’s honor!–that I will deliver the blueprints for the cold fusion reactor to Joe Romm in November 2016.
Now get out there and spend, guys. We’ve got an economy to save.
Cold fusion might lead to people not having to pay for energy.
That would hurt NGDP and cost jobs in the energy sector. Ban it.
But the technocrats in Washington can easily predict what everyone will choose in the future, so they can plan to print and spend less on totally unrelated things like more bridges to nowhere. It’s why the regulators are doing this instead of using their gift of foresight to make millions in the market. They’re doing it for us you see.
Problem solved. No ban needed.
Silas wrote:
Do you understand why it wouldn’t be a good idea to make each person hit a spending target each hour? Then you understand the problems in artificially moving up purchases.
Silas, don’t let me forget that one before my debate with Sumner. I have to think about it more, but that might be a real zinger.
Wow, you’re going to owe me a lot of credits soon! Don’t forget the one about artificially fixing NGDP.
I think those two points are really about the same point: NGDP has no fundamental connection to (“entanglement with”) the things we care about, so any prescription based on it will have tons of arbitrary elements:
– Why 5%? What about 5%/year is special?
– Why does the transfer of money for a “purchase” count for this metric? Couldn’t people produce a bunch of these without accomplishing anything productive?
– Why does the yearly NGDP have to be a certain value? Why not a month? A week? A second? If it’s acceptable for NGDP to “slump” from midnight to 5 am, why not March through June? 2008 to 2012? Now until we’ve constructed the robots that let us all retire?
– What if we start paying each other in gold, bonds, or a foreign currency?
– Why not goose expenditures by just forcing non-market production onto the cash nexus? Sorry, Joe, your wife can’t give you free marital services, you have to go downtown for “that”.
It’s like shooting ducks in a row … or something.
I hereby nominate Silas to write articles for Mises.org.
Thanks, Major_Freedom. I trust you follow my blog so you can get your Silas fix in the meantime?
(Btw, I was kicked off the Mises blog because I’m too pro-IP and I pissed off Stephan_Kinsella too many times…)
I became too intimidated about your blog when you wrote about the computations and lingo behind Bitcoin. Went over my head faster than the Bernank’s helicopter.
As for IP, I’ll only say this: If you are pro-IP, then you are pro-owning other people’s brains, because ideas are really just brain states. If you claim to own an idea, then you claim to own a particular brain state pattern.
It would be like claiming to own the thought “happy”, and then demanding that everyone who has the brain state associated with “happy” must refrain from manifesting that thinking into action.
What I find pro-IP folks always do when repeatedly and continually pressed, is to resort to committing appeal to consequences fallacies. “It would reduce investment”, “It would stifle innovation”, “It would make people reduce their effort”, “It would shrink the production of wealth”, “It would be bad for the economy”, etc.
Now, I don’t want to hijack this thread into a discussion of IP, I just wanted to do what you did, which is let you know of my general position towards it. If ever Bob does a post on IP, then I would be more than willing to beat that dead horse again for kicks.
I did some quick reading of your IP posts on your blog, and some of them date back to 2008. Wow dude, it looks like back then you and Murphy hated each other’s guts. The whole carbon cap trade thing really got heated.
Well, it wouldn’t be any fun if we just parroted stuff the other already agreed with, would it? It would probably make you and me too complacent in our beliefs.
Anyway, if you want to learn my thoughts on IP, I recommend this comment (and the post that started it, and the intellectual property tab if you have time).
I won’t say anymore about the topic in this thread.
Sorry, I meant to link this comment.
The data, though, clearly shows expectations do matter. Deal with the facts.
See Beckworth: http://macromarketmusings.blogspot.com/2011/11/some-evidence-on-importance-of.html
Oh my gosh Anon1, did you just prove to Silas and me that households actually care about nominal GDP when making their plans, by pointing to a post in which Beckworth showed that professional forecasters’ predictions of future NGDP tend to track actual future NGDP?
I swear if NGDP does not rise by 5% next year, I’m totally going to hold onto more of my money and not buy that new computer I have been holding out on.
I mean, I really care what EVERYONE is spending their money on, because…I sell goods and services to everyone? Yeah, that’s the ticket.
Wouldn’t a prudent business person , seeing that NGDP forecast are down and knowing this will probably affect his ability to sell in the future , care about that and want to factor it into his plans?
Why is this an issue that draws ridicule – what am I missing ?
Rob, how many business owners know what “NGDP” even is? I am not ridiculing the theory that NGDP forecasting is a good rule of thumb for the Fed to follow, I am ridiculing Scott Sumner’s frequent assertions that businesses in the real world act because they became worried that the Fed wasn’t going to keep NGDP growing at the historical trend rate. They did no such thing. It would be like me saying businesses thought the economy could support a longer Hayekian triangle.
Wouldn’t a prudent business person , seeing that NGDP forecast are down and knowing this will probably affect his ability to sell in the future , care about that and want to factor it into his plans?
His expectations of his own sales are what he sells into, not NGDP.
Falling forecasted NGDP is just the sum of expected sales for each economic actor (who sells something that is counted by NGDP).
If forecasted NGDP rises, then that means individual economic actors are expecting higher future revenues for their own lines of work and bid up the prices of scarce factors of production.
If forecasted NGDP falls, then that means individual economic actors are expecting lower future revenues for their own lines of work and bid down the prices of scarce factors of production.
The sum statistic “NGDP” is itself not an object of economic action. It’s why almost all businessmen don’t even know what it even means. How can a statistic that almost all businessmen don’t even know about, possible be a factor that affects their decisions?
if we assume that whatever expectations consumers have about the future that none of them are based either explicitly on NGDP forecasts or on any of its derivatives then I’m not sure how this damages the NGDP targetters arguments very much.
Under an NGDPT Regime : No-one need to worry about variations in AD since the CB is committed to keeping it stable
Under a fixed money regime: Individuals do need to factor variations in AD into their planning. Failure to do to so may lead to bankruptcy.
You can argue that the second option is better, since it keep the economy nimble and better able to respond to changed demand including changed demand for money. But to do so I believe is inconsistent with saying that individuals don’t factor in things that are proxies for NGDP into their planning.
Under an NGDPT Regime : No-one need to worry about variations in AD since the CB is committed to keeping it stable
No individual businessmen worries about variations in aggregate demand anyways. They only worry about the demand in the particular market they sell into.
Under a fixed money regime: Individuals do need to factor variations in AD into their planning. Failure to do to so may lead to bankruptcy.
Why should the choice be either target NGDP or fixed money? Aren’t you ignoring a third alternative?
Well, other things being equal then if AD falls the value of money rises and price adjustments will have to be made in every single market in response.
So if they are not concerned about AD – then they probably should b because in a fixed money regime where AD does vary those businesses that are able to plan and adjust to these changes will do better than those that do not. These AD changes will ,as you say, manifest themselves differently in different markets but I think the general point sill holds.
I honestly don’t know to what extent businesses use NGDP forecasts for planning purposes – my main points are really that efficient businesses will use all available information in in decision making (and I can’t see why they would particularly ignore NGDP data) , and that economic calculation will be more accurate under a regime of stable AD.
% of successful businesses that never paid attention to NGDP: 100 minus epsilon.
Sure Silas, but:
Having the Fed pledge to buy an unlimited quantity of assets, until it hits a NGDP target: priceless.