Paul Krugman — nothing but a die-hard neoclassical economist

3 July, 2016 at 11:01 | Posted in Economics | 21 Comments

In his review of Mervyn King’s The End of Alchemy: Money, Banking, and the Future of the Global Economy — on which I had a post up yesterday — Krugman writes:

Economist KrugmanIs this argument right, analytically? I’d like to see King lay out a specific model for his claims, because I suspect that this is exactly the kind of situation in which words alone can create an illusion of logical coherence that dissipates when you try to do the math. Also, it’s unclear what this has to do with radical uncertainty. But this is a topic that really should be hashed out in technical working papers.

This passage really says it all.

Despite all his radical rhetoric, Krugman is — where it really counts — nothing but a die-hard neoclassical economist. Just as people like Milton Friedman, Robert Lucas or Greg Mankiw.

The only economic analysis that Krugman and other mainstream economists accept is the one that takes place within the analytic-formalistic modeling strategy that makes up the core of mainstream economics. All models and theories that do not live up to the precepts of the mainstream methodological canon are pruned. You’re free to take your models — not using (mathematical) models at all is, as made clear by Krugman’s comment on King, totally unthinkable —  and apply them to whatever you want – as long as you do it within the mainstream approach and its modeling strategy. If you do not follow this particular mathematical-deductive analytical formalism you’re not even considered doing economics. ‘If it isn’t modeled, it isn’t economics.’

That isn’t pluralism.

That’s a methodological reductionist straightjacket.

So, even though we have seen a proliferation of models, it has almost exclusively taken place as a kind of axiomatic variation within the standard ‘urmodel,’ which is always used as a self-evident bench-mark.

(h/t Nanikore)

Applying closed analytical-formalist-mathematical-deductivist-axiomatic models, built on atomistic-reductionist assumptions to a world assumed to consist of atomistic-isolated entities, is a sure recipe for failure when the real world is known to be an open system where complex and relational structures and agents interact. Validly deducing things in models of that kind doesn’t much help us understanding or explaining what is taking place in the real world we happen to live in. Validly deducing things from patently unreal assumptions — that we all know are purely fictional — makes most of the modeling exercises pursued by mainstream economists rather pointless. It’s simply not the stuff that real understanding and explanation in science is made of. Just telling us that the plethora of mathematical models that make up modern economics  ‘expand the range of the discipline’s insights’ is nothing short of hand waving.

No matter how many thousands of ‘technical working papers’ or models mainstream economists come up with, as long as they are just ‘wildly inconsistent’ axiomatic variations of the same old mathematical-deductive ilk, they will not take us one single inch closer to giving us relevant and usable means to further our understanding and explanation of real economies.

King knows that. Krugman obviously not.



  1. The tell is that Mervyn King was Governor of the Bank of England and Chairman of its Monetary Policy Committee for ten years while Paul Krugman is a textbook writer. So was Paul Samuelson, who started this whole thing.

    Both Samuelson and Krugman won the Sveriges Riksbanks pseudo-Nobel for theoretical contributions but the closest either got to practical employment was as advisors and public intellectuals.

    While it is true that Krugman sidekicks “Keynesians” Larry Summers and Brad DeLong, who like Krugman also self-identify as neoclassical economists, did serve as officials of the US Treasury in the Clinton Administration, the Clinton surpluses resulted in the recession that Clinton left G. W. Bush.

    While Democrats in general view the surpluses as an achievement to be replicated, they arguably contributed to the 2007 crisis through an ensuing increase in private indebtedness, based on analysis using SFC macro-modeling and sectoral balances.

    No wonder economics is called “the dismal science” with “experts” like this in the Samuelson mold (math fetishists).

  2. Yeah, so! He never pretended to be something else!

    • Absolutely right. Krugman has never made a secret of himself being a ‘proud’ neoclassical mainstream economist.
      The problem is that a lot of people seem to think that — since he sometimes chooses to criticize the fresh water part of the family — he doesn’t really belong in the family. But he does — and therefor he never questions the deductivist foundations of mainstream economics. And that’s also the reason he’s no alternative for those he really want to make economics a realist and relevant science! If you look for real alternatives, I would rather suggest you read Keynes, Minsky, Polanyi, Galbraith, Myrdal, Hodgson, Keen, Mirowski, Lawson, Davidson, Kalecki, Chick, Dow, ……

      • Yeah, I read most of them. I like much of what some of then have written but I’m not convinced that they have offered an alternative!

      • In social science there is recognition that different approaches apply to different aspects of the subject matter depending on the levels of generality and granularity. These approaches are formal or conceptual, thin or thick, nomothetic or idiographic, causal or hermeneutic. In contrast, conventional economists have declared that the methodological debate is over and thin, nomothetic, and causal have won. Formal, thin, idiographic and hermeneutic are “heterodox.”

        Both more general and more granular approaches are required in social science. Preferring the formal, thin, nomothetic and causal to be more “scientific” is to take natural science as the norm rather than recognizing that the natural, life and social sciences have different subject matters than require different approaches to study comprehensively.

        Even in the natural sciences, this is the case. Kepler and Newton approaches to planetary motion, as great advances they were, revealed nothing about the difference in the composition of the planets, since this was irrelevant that the level of generality being aimed at.

        So there is nothing wrong with taking different approaches depending on the level of generality or granularity. However, whichever approach is selected has be to correctly conceived and applied. There is nothing wrong with using a simple IS-LM approach in some cases as a gadget. But one has to get the assumptions right for the system being addressed, which Krugman does not, since he doesn’t correctly understand the present monetary system or the banking system. But similar things can be said of DSGE modeling used by central banks.

        Central bankers know this; there is no Taylor rule simply to follow in order to achieve success. So in their deliberations they combine the formal and conceptual, thin and thick, nomothetic and idiographic, and causal and hermeneutic in assessing the state of the econony and how to address it with monetary policy. Having served as a central bank chief, King knows this. Krugman is still stuck in the textbook. Worse, his book is wrong on major points about government finance, money & banking, and finance, as Post Keynesian and MMT economists have pointed out.

        Moreover, no way has been devised to incorporate Minsky’s conceptual analysis into a formal model yet, although Keen is working toward it. Nor is there a way to incorporate Bill Black’s granular analysis of the legal aspects of the banking system that were heavily involved in the lead up to the 2008 crisis. As result, conventional economists were blindsided and write the crisis off to a shock that could not be foreseen (in the models).

        But the FBI warned of rampant fraud in residential housing and the mortgage industry in December 2004, while Alan Greenspan blew it off as “froth” in some markets. The model showed everything OK in the banking and financial sector, while termites were eating away the foundation. Subsequent investigation and multi-billion dollar settlements reveal how deep and pervasive parasitism was and how it influenced the crisis as a major factor if not a principle cause. It was a necessary condition for the level of Ponzi finance and a sufficient condition for a devastating financial bubble owing to the importance of the asset class and the systemic risk created.

        Krugman’s “where’s your model?” and “what’s your alternative?” are off-base. It’s a facile defense of a failed approach.

        This could be written off as bad science if it were merely theoretical but macro is a policy science and bad policy based on bad science is bad socially, politically and economically.

  3. From the outside, it seems insane that a profession would simply ignore reality, because it lacks the ability to model certain aspects of it analytically. Krugman’s own fame within economics is due to his role in inventing analytic methods for treating increasing returns and their effects on the pattern of trade, and therefore opening their consideration by neoclassical economics. Increasing returns were determining the patterns of trade long before neoclassical economics could say anything analytical about them, and when economists could observe and write discursively, notice was taken. Bertil Ohlin said fairly insightful things about the importance of increasing returns to trade back in the day, just to take an example.
    For Krugman, “it’s unclear what this has to do with radical uncertainty” though it is clear enough to me: an axiom of “radical uncertainty” is incompatible with the tractability of analytic models. You cannot assume a pervasive uncertainty and still motivate a definite analytic result. “Profit-maximizing” agents cannot profit-maximize, if “maximize” has no working definition, which it cannot under conditions of radical uncertainty.
    This doesn’t prevent financiers, businessmen and politicians from getting on with it, but it does mean that they get on with it, while relying on cues and hedges and structures of coordination and insurance that can be barely glimpsed from the corner of an analytic eye. It means, as I noted in another comment, that rent-seeking rather than profit-maximizing guides decision-making — and that’s a rather important distinction practically, even if neoclassical economics strives to ignore or misunderstand it.

  4. The present post reminds me of the Keen v Krugman debate from 2012, which began with this discursion from Krugman:

    I always try to find the simplest representation I can of whatever story I’m trying to tell about the economy. The goal, in particular, is to identify which assumptions are really crucial — and in so doing to catch yourself when you’re making implicit assumptions that can’t stand clear scrutiny.

    Keen doesn’t seem to be doing that. His paper contains a number of assertions about what is crucial, without much explanation of why these things are crucial. And I guess I just don’t see it.

    In particular, he asserts that putting banks in the story is essential. Now, I’m all for including the banking sector in stories where it’s relevant; but why is it so crucial to a story about debt and leverage?

    Keen says that it’s because once you include banks, lending increases the money supply. OK, but why does that matter? He seems to assume that aggregate demand can’t increase unless the money supply rises, but that’s only true if the velocity of money is fixed; so have we suddenly become strict monetarists while I wasn’t looking? In the kind of model Gauti and I use, lending very much can and does increase aggregate demand, so what is the problem?

    Keen then goes on to assert that lending is, by definition (at least as I understand it), an addition to aggregate demand. I guess I don’t get that at all. If I decide to cut back on my spending and stash the funds in a bank, which lends them out to someone else, this doesn’t have to represent a net increase in demand. Yes, in some (many) cases lending is associated with higher demand, because resources are being transferred to people with a higher propensity to spend; but Keen seems to be saying something else, and I’m not sure what. I think it has something to do with the notion that creating money = creating demand, but again that isn’t right in any model I understand.

    My point is that there seems to be a lot of implicit theorizing going on here — and at least at first glance, the implicit theorizing doesn’t make a lot of sense.

    I thought this was revealing, but I also found it completely exasperating. He’s all for including banking where it’s relevant, but he cannot see why it is relevant to a story of debt and leverage. Other thantry to take away his supply of hallucinogens, what can any of us do with that?

    • Unless he has changed his mind recently, Krugman still thinks that banks are only intermediaries between savers and borrowers, that there is a fixed amount of loanable funds, and that money is neutral in the long run. Lars has posted on all these mistakes.

    • Banks are only intermediary in a certain sense: Any loan that is granted implies that someone’s spending exceeds his or her income. That does necessarily imply that someone else’s spending must fall short of his or her income. Otherwise the economy as a whole would spend more than total income, which is an impossibility (assuming that there is no international spending, or that we look at the world as a whole).
      Moreover, Krugman’s view — if I interpret it correctly — is that you can analyze debt and leverage in a model that sidesteps the workings of the banking sector, precisely for the above reasons. One example of this is Eggertson and Krugman (2013).

      • Eggertson and Krugman (2013) as far as I can tell presents a tautology. It is argument by assertion dressed up, but with no where to go.
        As for the rest — Any loan that is granted implies that someone’s spending exceeds his or her income. — it sounds like loanable funds, and what Tom Hickey says, applies.
        Rather than think about money as a viscous liquid or banks making discrete loans, think about money as a score-keeping system and banks as player-referees managing a payments and credit system. Banks as intermediaries do not transmit a liquid flow so much as they mirror and authenticate the working capital of households and firms. Leverage is an aspect of that mirroring.

      • No, there is no fixed amount of loanable funds.
        The point I’m making is simple. In order for national accounting to hold, it is by necessity so that one individual’s borrowing — and therefore spend in excess of his or her income — implies that some other individuals must save the precise same amount — and therefore spend less than their income. If this accounting does not hold, aggregate spending would exceed or fall short of aggregate income, which it cannot (again ignoring international lending).
        In this sense, banks are mere intermediaries in the aggregate. If that “sounds” like a loanable funds theory, I do not care. It is true by construction.
        Please let me know which part of the above you are in disagreement with. If not, it won’t be possible to have a conversation. Mr. Hickey’s statement on loanable funds was not very informative.

      • “Banks are only intermediary in a certain sense:”

        No one denies that and it is not the issue, Banks are intermediaties in the sense that in providing leverage they liquefy collateral or expected future income. But the loanable funds theory assumes in addition that banks loan out deposits that they take.

        “Second, at the end of his most recent blog it is pretty clear that Krugman leaves banks out of his model because he doesn’t understand “what banks do”. He starts by saying “If I decide to cut back on my spending and stash the funds in a bank, which lends them out to someone else…” Well, if he had actually read Minsky, he would understand that this is the description of a loan shark, not a bank.” — Randy Wray, WHY MINSKY MATTERS: Part One

        Banks don’t loan out deposits. Loans create deposits. Nor do the loan out bank reserves held on balance in their accounts at the central bank. These function as settlement balances and exist only on the central banks spreadsheet. The only way reserve balances enter non-government is when banks exchange reserve balances for vault cash and then pass it through the window into circulation.

        Krugman and Mankiw on loanable funds — so wrong, so wrong

        Philip Pilkington: How Krugman’s Addiction to the ISLM Model Has Led to Repeated Bad Forecasts

      • pontus: in order for national accounting to hold, it is by necessity so that one individual’s borrowing — and therefore spend in excess of his or her income — implies that some other individuals must save the precise same amount — and therefore spend less than their income.
        I think the conventions of national accounting are confusing you. In any case, this proposition as you state it is not true, logically or factually.
        National accounting is a double-entry bookkeeping system: each observed transaction is presumed to create two entries, which exactly balance each other, so that spending and income mirror one another, are two aspects of the same stream of transactional business activity. If I buy something, someone must have sold it: every purchase must be a sale. By the same bookkeeping logic, every transaction in which one entity borrows entails another entity lending.
        A conceptual difficulty arises regarding the definition of “savings” as refraining from consumption, because “saving” as not-spending does not logically entail a transaction. If I save $1 by putting it in my sock drawer, I have not spent it or lent it.
        Those sock drawer hoards are a problem for your idea that people must borrow to spend more than their income or that others must lend them precisely the excess of their spending over their incomes.
        National income accounting identifies a category it calls “saving” that it opposes in its double-entry scheme to “investment”. Honestly, I have to say it does not make a lot of sense: while investment is spending and transactional, saving is not. In practice, a large part of “saving” in the national accounts is a residual, which is to say it is not estimated from observed transactions because there are none: the statisticians fudge. They make up a number to make things balance. That’s right. They pull it out of the air. I really wish they would omit “saving” from the national accounts altogether — it would be cleaner conceptually.
        In terms of economic theory, we might like to distinguish saving from consumption, with a conservation law that required saving and consumption to total production. All production would be allocated to either consumption now or investment for the future. In this case, saving and investment are equivalent, indeed identical, and I suppose the national income accountants were thinking along such lines. Money makes such thinking a hopeless muddle — money and time.
        People save money, not production. Our opportunities to produce cannot be saved; we act now, only now. If we lose our chance, it is lost. Money creates a fiction, a virtual world where we can save and invest. It is marvellous really. But, money is not so tightly fitted onto the “real” world that one cannot create money to spend without anyone saving their foregone consumption. And, the national accounts do nothing to tighten the fit.

      • Hi Bruce
        No the sock hoarding analogy does not present a problem. What I am saying is simply that my income is your spending, and your income is my spending. So if I spend more than my income, you must spend than your income by the same amount.
        To hammer home that point, suppose that my income is y1, my spending is s1, and whatever else I do with my income that is not spending is called x1 (this can be cash hoarding in socks or whatever). Let’s denote the same variables for you by the indexation 2.
        Our budget constraints are
        If your income is my spending, and my income is your spending then y1=s2 and y2=s1, so
        Simplifying we find that
        Thus, no matter how we do things — no matter if banks create money “out of thin air”, no matter if you save by stuffing your mattress — as long as our income depends on others’ spending, it is still true that if you spend more than your income then someone must spend less (of course, income may very well change because of such decisions, but that’s not what this is about).
        Is this loanable funds? I don’t like these labels a lot, so I don’t know. But if it is, loanable funds is correct.

      • pontus,
        Thank you for clarifying your thinking with mathematics. I think we might have to concede to Krugman on the clarifying power of mathematics. 😉
        What you describe is commonly called “the circular flow”. The economy is conceived of as two flows — one of money and the other of goods and services; money moves from hand to hand in one direction and goods move in the other. In the circular flow of two, what I sell to you is my income, so that what you spend is my income.
        In the circular flow, income is a rate. A larger income is a faster rate of spending. If you spend more — that is, at a faster rate — my income increases.
        Of course, when my income increases, I may feel that I can spend more (aka at a faster rate). If I do spend more, your income increases.
        One can imagine circumstances in which I go to the hoard in my sock drawer to finance an increment in spending beyond what I can finance from my current rate of income. My increased spending increases your income. You feel you can increase your spending, given your additional income, and you do so, increasing my income, which helps me sustain my higher level of spending without dipping into the sock drawer. This virtuous spiral is the basis for the concept of the multiplier used in Keynesian analysis of fiscal policy.
        So no, there need be no conservation of funds requiring that, should you spend more than your income, I must spend less. On the contrary.
        There may be a constraint in production capacity. There are two flows, after all, and if I am not able to increase my production of goods and services to match the effective demand of your increased rate of spending, then I will not be able to earn an increased income. The multiplier only works to bring up the pace of activity in circumstances of slack.
        The circular flow is unrelated to the double-entry bookkeeping of the national accounts and the bookkeeping conventions imply no constraint. The bookkeeping simply records the transactions as they take place and will faithfully measure in periodic sums the rates of transactional activity.
        Without going into detail, Loanable Funds is an artifact of an imaginary world in which money has no existence of its own, but simply maps a fully utilized production capacity. Then, to “save” requires foregoing consumption so that production capacity can be freed; lending that savings entails transferring that reserved production capacity to the use of the borrower.

  5. Prof. Syll’s analysis is onomasiologically incomplete.
    No example is given of “relevant and usable means to further our understanding and explanation of real economies”.
    There is no example, not even even a vague outline, of what Prof. Syll might regard as a satisfactory economic analysis of “an open system where complex and relational structures and agents interact”. Because it is devoid of any such substance, Prof.Syll’s argument provides no understanding of how Krugman’s many errors could have been avoided.

    Instead we are left with the absurd impression that everything Krugman and all other economists have ever written in the past, or can ever write in the future, is doomed to be epistemologically and ontologically unsound.

    P.S. Krugman’s review of King is excellent, IMHO.

    • Krugman’s punchline was good – that despite calling for the wholesale rejectionf neo-classical economics – King’s proposals for policy were a bizarre anti-climax in that they were yawningly mundane in their conventionality. As Krugman said, they could have been written out of the IMF’s handbook. Where I disagree with Krugman is that if we are going to get real solutions, we are going to have understand the real problems, and that requires fresh thinking and the removal of a lot of unnecessary clutter to allow a subject that looks more like the humanities and other social sciences – that is where I agree with King. Like an historian they have to start with the primary (including qualitative) evidence and build up the picture from the ground up. Samuelson’s project has ultimately reached its inevitable end – but the subject which has invested so much in it is finding reality hard to swallow. I don’t know what it is going to take to finally kill the beast.

  6. I have my reservations about Krugman, too, but I think you’re being somewhat unfair to him on this and your previous posts about his review of King’s book. You have to read his remark in context and the context, here, is King’s assertion that “Monetary stimulus via low interest rates works largely by giving incentives to bring forward spending from the future to the present. But this is a short-term effect.”.

    Remember, it’s King, not Krugman who claims that stimulus works by bringing forward spending from the future (i.e. bringing forward future demand, as opposed to creating demand where none had exchanged before). King’s assertion smacks of the worst neoclassical nonsense to me (“intertemporal maximisation” etc.). In such a context, I don’t think it’s unreasonable for Krugman to say “set out your model in detail for me”.

    True, setting out such a model would imply accepting the core assumptions of what you call the ur-model, but has not King already done that by his implicit acceptance of intertemporal maximisation or whatever it is he is basing his curious assertion about “bringing forward” demand on? If it turned out that his model does not work they
    way he thinks it does, even after accepting those assumptions, then there’s a problem with his argument, isn’t there? Of course, the fact that his model (or Krugman’s models) do work if you accept their (unrealistic) assumptions, doesn’t prove very much either, I’ll gladly grant you that.

    And, on the youtube video, take a look at this: 🙂

    • No, it is not loanable funds, and that is Krugman’s mistake.

      The income-expenditure model used by national accounting is based on accounting identities, to which all sane people agree since it is just a matter of balancing books.

      There is no causality involved in accounting identities. None. Nada. Zip. It’s simply double entry. As far as I know, everyone agrees with this.

      This implies that finance and economics must be stock-flow consistent. But consistency is tautologous and makes no claim about the real world. That requires theoretical interpretation and hypothesis testing.

      However, accounting identities are used in economics. A lot of macro is based on them. But imputing causality implies theory and theories are either correct or incorrect based on facts in addition to accounting.

      The evidence for theoretical imputations of causality includes operations. Operations are often incorrectly understood, with the result that causality is sometimes reversed.

      Loanable funds is based on assuming that saving causes investment. As Keynes pointed out the causality is reversed. Investment causes saving.

      Firm spending results in household (workers’ and owner’s) incomes. The income that is not spent on consumption is a residual called “saving.” It must be equal to the residual of firm income from consumption goods, which is called investment. C = C so S = I.

      This is true in a closed system of consolidated private domestic sector, and it remains true of the entire system in a multi-sector model such as national accounts that include DP, G, and NX, since all books balance. The sum of the balances of all sectors must be zero by identity. That just says all books balance.

      But it is firm spending that results in income to the household sector, so firm investment feeds household consumption (C) and its residual that is not spent in the period, which is labeled saving (S), in the income-expenditure model.

      But hat that is not loanable funds, which uses saving in a somewhat different sense as bank saving (deposits). The mistaken idea is that banks take deposits and then loan them out, so the stock of deposits funds the total of borrowings from banks by economic entities. This definition of “saving” is different from S in the income-expenditure model, which is inclusive of more than bank saving.

      That model of banking is wrong operationally. Banks don’t work like that. Loans create deposits. A loan is a bank asset/customer liability and a bank deposit is a bank liability/customer asset. The accounting shows that prior saving is NOT required for a bank to make a loan, only a qualified customer. Banks do not loan out deposit they take. They create deposits in the process of extending loans.

      Loan officers don’t check with ALM prior to making loans to determine whether the bank has enough deposits to cover the loans being negotiated. They inform ALM of loans in the pipeline so that ALM can insure that the bank’s settlement account at the Fed balances in the period in order to avoid incurring the penalty rate.

      Consider the banking system as a single bank. As loans are made, deposit accounts are credited. Those deposit accounts are used in shifting the funds created by loans around intra-bank. No deposits are needed to create new loans. This is true of the banking system, with the central bank ensuring as the lender of last resort that liquidity is always available to clear in the payments system as deposits get shifted around and cash is withdrawn at the window.

      In the IS-LM model based on loanable funds, the stock of bank savings (M2) gets increased or decreased by changes in the interest rate, which, being the price of money, influences the propensity to spend or save (liquidity preference). If consumption increases, then there is less available for investment, which affects growth and leads to inflation.

      The IS-LM model is a causal theory that assumes loanable funds.The interest rate is the lever the central bank uses for control money demand and discipline consumption desire. The idea is that as the stock of savings increases there is more available for investment, dampening inflationary pressure and making more funds available for investment.

      In this model, central bank technocrats acts on behalf of markets to maintain equilibrium between saving and investment, optimizing growth, also with an eye on price level and wage pressure to manage inflation. Because natural rates and all that Wicksellian stuff.

      But reality refuses to comply.

      Krugman still doesn’t get endogenous money, and the textbooks, his included, don’t reflect it yet, even though central bank papers are now affirming what circuitists and chartalists have been saying.

      • So that’s where you and Mr. Wilder disagree. Mr. Wilder states that
        “I think the conventions of national accounting are confusing you. In any case, this proposition as you state it, is not true — logically or factually.”
        While you claim that
        “The income-expenditure model used by national accounting is based on accounting identities, to which all sane people agree since it is just a matter of balancing books. […] As far as I know, everyone agrees with this.”
        At least now you know that not everyone agrees with this!
        As for the other part of your theorizing, I don’t have an opinion.

      • “At least now you know that not everyone agrees with this!”

        I doubt that Bruce Wilder doesn’t agree that accounting identities are tautologies and that they don’t have causal import unless theoretically interpreted. But I’ll let him speak to that.

        BTW, some of the confusion is perhaps created by incorrect symbolism. The identity sign is three horizontal lines and the equality sign by two lines. The identity sign is the symbol of tautology whereas the equality sign is used in writing functions imputing causality.

        Identities as identities are not functions symbolizing input-output and its direction causally in terms of independent and dependent variables. However, identities are often expressed using the equals sign when they are meant as tautologies.

Sorry, the comment form is closed at this time.

Create a free website or blog at
Entries and comments feeds.