Archive

Author Archive

The imputed rent is to darn high! More on Piketty

April 22, 2014 2 comments

Quite some people who are reading Piketty are somewhat puzzled by ‘r’, the return on capital. How can r keep increasing when the amount of capital, as a % of GDP, doubles?! What is ‘r’ anyway? Piketty is actually quite clear on ‘r’: he uses the definitions and data from the national accounts. One of the constituent parts of ‘r’ are house rents: actual rents as well as imputed rents. Imputed rents are an imputed stream of income of owner occupants of houses and are supposed to be as high for these owners as actual rents for comparable houses (becomes a little difficult for houses of the top 0,1%, but for most owner occupied houses this procedure is reasonable). And when we look at imputed rent as a % of total household expenditure we see that this did increase quite a lot (should have calculated this as a part of disposable income or GDP but that required a whole lot of additional calculations while these data were directly available on the Eurostat site). Even for neoclassical economists, who do not believe that house prices are house prices and calculate the value of housing capital as the discounted value of the stream of rents (look here, in french, via Marginal Revolution) this will mean that the amount of capital increases, compared with household expenditure (and a fortiori when the ECB lowers interest rates, i.e. the discount factor). I do admit that it is troublesome to mark houses to market when they are not on the market. But discounting a stream of income in an uncertain world is troublesome, too, and it does not make any sense to calculate the extra ordinary rise of house prices in many countries away! Anyway – the imputed rents are too darn high! Which explains part of the conundrum.

Update: I forgot to state that quite a part of the imputed rent gain of owner occupiers is of course siphoned off by the banks which provided mortgages with newly minted money

rents

.

 

Categories: Uncategorized

On Piketty and definitions. ‘Financial capital’ is not the same thing as ‘physical capital’ (two graphs)

April 21, 2014 4 comments

The discussion about Piketty is getting messed up. The vagueness of economic parlance allows people to accuse him of mistakes he doesn’t make. The meaning of the word ‘capital’ is a case in point. Financial capital (like bonds, cash, receivables (which are included in the estimates of financial wealth in the national accounts and which have the same magnitude as cash, deposit money and money in savings account combined)) are not the same thing as physical capital (houses, cars, buildings, machinery and the like). Financial capital is to quite some extent owned by old, retired women. Physical capital is used by people at work. The Piketty book, ‘Capital in the 21st century’, is about the inexorable rise of financial capital, not about the increase of physical capital, as Clive Crook rightly notes (below).

Over the course of history, capital accumulation [physical capital, M.K.] has yielded growth in living standards that people in earlier centuries could not have imagined, let alone predicted — and it wasn’t just the owners of capital [financial capital, M.K.] who benefited. Future capital accumulation may or may not increase the capital share of output; it may or may not widen inequality. If it does, that’s a bad thing, and governments should act. But even if it does, it won’t matter as much as whether and how quickly wages and living standards rise.

That is, or ought to be, the defining issue of our era, and it’s one on which “Capital in the 21st Century” has almost nothing to say.

That’s right. But the english word ‘capital’ can mean financial capital as well as physical capital. It’s a bad thing that economics as a science did not come up with more precise phrases and definitions – but that’s the way it is. Piketty does talk about physical capital – as financial capital is always some kind of ownership of physical or other financial capital and increases in physical capital can give rise to an increase in financial capital, like bonds (when the government uses bonds to finance bridges) or stock – but in his book he does not try to explain economic growth, as Crook rightly mentions. He points out and explains the inexorable, endogenous rise of financial capital (graph 1). And it’s sad to see that Crook mixes up both meanings of the phrase ‘capital’ in one sentence. And to see Tyler Cowen approvingly albeit somewhat cunningly quoting this statement.

Financial capital is of course another kind of technology – a market economy without ‘payables’ and ‘receivables’ is inconceivable. Piketty however shows that the increase of financial capital does not come without risks – he does not give enough attention to pension funds, but these are some of the largest owners of financial capital (graph 2). And the fickle character of financial markets does mean that the increase of financial markets and the increasing importance of returns and the value of financial market is, in fact and alas, a defining issue of my financial future. This will still take a while and today the sun is shining. But in my country at least part of the retired experience ridiculous swings in pensions – while the stock of physical capital is steadily increasing!

For some idea about the different growth rates of financial and physical capital see below, source: Centraal Bureau voor de Statistiek. Sadly, and tellingly, the physical series has been discontinued – we’re decreasingly interested in baking pies and increasingly in slicing them.

real capital

capital4

Update: The sun was shining. Now it rains.

Categories: Uncategorized

Pensions: an Iznop game? (three graphs)

April 20, 2014 Leave a comment

Does Europe have to panic about pensions? Or about unemployment? Paul Samuelson famously wrote, back in 1967,

Social security is a Ponzi scheme that works. The beauty of social security is that it is actuarially unsound. Everyone who reaches retirement age is given benefit privileges that far exceed anything he paid in – exceed his payments by more than ten times (or five times counting employer payments). How is it possible? It stems from the fact that National Product is growing at a compound interest rate and can be expected to do so as far as the eye cannot see. Always there are more youths than old folks in a growing population. More important, with real income going up with 3% a year, the taxable base on which benefits rest is always greater than the taxes paid historically by the generation now retired…. A growing nation is the greatest Ponzi scheme ever contrived

Samuelson was of course right. Here you find (in french) an INSEE study which calculates that (due to the fact that about ten years ago french pensions were not indexed to wages anymore but to inflation) the burden of french pensions will not increase, thanks to growing labour productivity. France however has a relatively high fertility rate (about 2,0, just below the replacement level of 2,1). Countries like Germany and Italy and Spain however have fertility levels of only about 1,4. And, to paraphrase Samuelson, ‘Always there are more old folks than youth in a shrinking population’. Is this a problem? Not entirely. Dear comrades and (what’s the female equivalent?), unemployment in the EU is at a historical high of 12%, putting these people to work will mean less unemployment benefits and much more production, which will pay for the pensions problems and the more so when we add productivity increases. Also, fortunately and wisely, during the Ponzi-days the de facto retirement age was lowered all over Europe, which means that at this moment there is, surely in Southern and Eastern Europe, a large labour reserve (graph 1). True, after about 2000 labour participation rates of this age group increased everywhere (graph 2, though there seem to be recent setbacks because of ultra high unemployment). Read more…

Categories: Uncategorized

The solitude of Latin America. Nobel lecture, 8 December 1982, Gabriel Garcia Marquez

April 18, 2014 1 comment

The Solitude of Latin America

Antonio Pigafetta, a Florentine navigator who went with Magellan on the first voyage around the world, wrote, upon his passage through our southern lands of America, a strictly accurate account that nonetheless resembles a venture into fantasy. In it he recorded that he had seen hogs with navels on their haunches, clawless birds whose hens laid eggs on the backs of their mates, and others still, resembling tongueless pelicans, with beaks like spoons. He wrote of having seen a misbegotten creature with the head and ears of a mule, a camel’s body, the legs of a deer and the whinny of a horse. He described how the first native encountered in Patagonia was confronted with a mirror, whereupon that impassioned giant lost his senses to the terror of his own image.

This short and fascinating book, which even then contained the seeds of our present-day novels, is by no means the most staggering account of our reality in that age. The Chronicles of the Indies left us countless others. Eldorado, our so avidly sought and illusory land, appeared on numerous maps for many a long year, shifting its place and form to suit the fantasy of cartographers. In his search for the fountain of eternal youth, the mythical Alvar Núñez Cabeza de Vaca explored the north of Mexico for eight years, in a deluded expedition whose members devoured each other and only five of whom returned, of the six hundred who had undertaken it. One of the many unfathomed mysteries of that age is that of the eleven thousand mules, each loaded with one hundred pounds of gold, that left Cuzco one day to pay the ransom of Atahualpa and never reached their destination. Subsequently, in colonial times, hens were sold in Cartagena de Indias, that had been raised on alluvial land and whose gizzards contained tiny lumps of gold. One founder’s lust for gold beset us until recently. As late as the last century, a German mission appointed to study the construction of an interoceanic railroad across the Isthmus of Panama concluded that the project was feasible on one condition: that the rails not be made of iron, which was scarce in the region, but of gold.

Our independence from Spanish domination did not put us beyond the reach of madness. Read more…

Categories: Uncategorized

Solar in the EU: triple E.

April 17, 2014 1 comment

The future is solar, as we all know. And the future is here (graphs). Solar panel area tripled in ten years which, as panels are becoming more effective, means that solar power generation must have increased fourfold or something like that. Which is only the beginning. A quick internet search yielded, however, that China and the USA seem to be outgrowing the EU.

An interesting aspect: solar electricity is not only durable – it’s also ‘local’. The modern state is to quite an extent a network state: sewer systems, highway systems, railway systems, the electricity grid, television and telephone networks etcetera. Solar electricity is, though compatible with networks, not necessarily dependent on large, national grids and allows for a more dispersed use of capital (financially as well as physically).

The comparative advantage of the southern EU states is not yet visible in the statistics.

Source: Eurostat, for a number of states no data available.

solar1 Read more…

Categories: Uncategorized

Unemployment in Europe: the totally stunning regional differences

April 15, 2014 Leave a comment

‘Totally stunning’: two hyperboles. But I had to use them. Could anybody only six years ago have imagined a Eurozone core with 4% unemployment or less (here the new regional unemployment data) and a southern periphery with large areas with unemployment of over 30%. Broad unemployment in these regions must be somewhere between 35 and 40%. Hey, Andalucia has 36% normal unemployment… Mind that inflation is going down in the core, too. Unemployment percentages of 3 to 4% are i.e. totally feasable.

Unemployment2

Mind also that borders between countries do explain part of the differences. But only a part. Mind also that unemployment in Eastern Germany is finally becoming less high – but it took, despite lavish transfers, almost 25 years and mass migration before this happened. The ‘Wirtschaftswunder’, based upon debt redemption and equality, worked better than neoliberal internal devaluation (in the fifties unemployment went down from about 12% to about 2% in 9 years).

Categories: Uncategorized

Piketty’s dataset: part of a trend which is changing economics.

April 14, 2014 Leave a comment

Update: via Business Insider: this 2012 Cato Institute report by Steve Hanke and Nicholas Krus which, starting in France in 1796, carefully lists all 56 known episodes of hyperinflation (21 of which were connected with demise of Soviet Union and Yugoslavia).

I’m reading Thomas Piketty’s book about wealth, capital and inequality. At this moment one remark:

His book is based upon a very extensive ‘open source’ dataset which spans the centuries and the globe (wealth, return on capital, labour share, share of capital etc.). This seems to be part of a trend as Piketty is not the only economist who does this. Other examples are:

Carmen Reinhart and Kenneth Rogoff with their ‘This time it’s different. Eight centuries of financially folly‘ dataset, which spans the centuries and the globe (debt).

The late Angus Maddison data on GDP  (dataset continued by ‘a group of close colleagues’) which span the millenia and the globe

The Bank for International Settlements with their recent dataset on house prices which span decades (for Norway: centuries) and the globe.

The (real) wages datasets of the International Institute of Social History (moderator: Jan Luiten van Zanden) which span the centuries and the globe.

These datasets are changing or did already change the science of economics. A common theme: there is no such thing as a stable monetary capitalist economy.

I do think that, as long as we have the Sveriges Riksbank prize in economics science in memory of Alfred Nobel, the founding and maintenance of such datahubs should be one of the arguments to award the prize.

Categories: Uncategorized

The ECB is failing its own, flawed, goals

April 13, 2014 Leave a comment

One of the functions of the 2% Eurozone inflation target of the ECB is to make processes of internal devaluation easier. This should, according to the ECB, be possible without outright deflation of the price level.  According to the 2011 ECB manual ‘The monetary policy of the ECB‘ (161 pages):

Taking the existence of unavoidable inflation differences into account, it has been argued that the ECB’s monetary policy should aim to achieve – over the medium term – an inflation rate for the area as a whole that is high enough to prevent regions with structurally lower inflation rates from having to meet the costs of possible downward nominal rigidities or entering periods of protracted deflation. According to all available studies, a rate of inflation below, but close to, 2% for the euro area provides a sufficient margin also in this respect.

In other words: 4% inflation in the Netherlands and Germany is necessary to enable Spain and Portugal and Greece and Italy to lower their price level relative to the Dutch and German level without having to lower nominal wages. That’s what this is all about. One can wonder if such a policy is effective anyway. As a recent ECB working paper states: Read more…

Categories: Uncategorized

The unpleasant political arithmetic of Greek bonds (2 graphs)

April 11, 2014 Leave a comment

Greek bonds fly of the shelves‘. Yesterday, Greece successfully issued 3 billion 4,95% bonds – and I’m still not seeing that one coming. Greece does not need new debt to be able to pay its interest bill. It needs frontloaded debt reduction to get debts down to a sustainable level, i.e. 60% of 2015 GDP (conservatively estimated, i.e. taking 0% growth and 3% deflation into account).Credible measures and structural reforms to make the remainder of the debt more flexible (i.e. ‘deflation protected’ bonds) have to be put in place. Tax measures – like a land tax – which increase the velocity of the stock of wealth and induce market oriented and demand boosting use of this wealth have to be introduced. Read more…

Categories: Uncategorized

Growth in college food banks

April 10, 2014 Leave a comment

From: David Ruccio

296Hunger0405

As the Washington Post explains, Read more…

Categories: Uncategorized

Some links, 10/4/2014. Cooperations, energy (graph), Peter Praet (ECB) on the impossibilities of monetary policy in the EZ

April 10, 2014 Leave a comment

The ILO is officially charged with promoting and estimating cooperatives:

Here a little about the new statistics on cooperations. Here a report about cooperations (mind that Goldman Sachs was a partnership until 1999):

As business organization, cooperatives contribute to economic development, generating more than 100 million jobs and securing the livelihoods of nearly a quarter the world’s population. Cooperatives provide an important channel for bridging market values and human values … The financial and ensuing economic crisis has had negative impacts on the majority of enterprises; however, cooperative enterprises around the world are showing resilience to the crisis. Financial cooperatives remain financially sound; consumer cooperatives are reporting increased turnover; worker cooperatives are seeing growth as people choose the cooperative form of enterprise to respond to new economic realities. This report provides historical evidence and current empirical evidence that proves that the cooperative model of enterprise survives crisis, but more importantly that it is a sustainable form of enterprise able to withstand crisis, maintaining the livelihoods of the communities in which they operate

From the Worldwatch institute: energy priorities: Read more…

Categories: Uncategorized

Are New-Keynesians accidentally discovering Keynes?

April 9, 2014 5 comments

Are ‘New-Keynesians’ discovering Keynes? Paul Krugman links on his blog to an Eggertsson/Merohtra paper which allows the ‘natural rate of interest’ to fluctuate. Which actually sounds somewhat Keynesian. In new/neo/old classical thinking the natural rate of interest equilibrates, in the unspecified run, supply and demand in all markets, including the labour market. An idea which, according to Keynes, was not so much wrong but useless. David Glasner, on his ‘Uneasy Money’ blog, states about Keynes this (emphasis added):

Keynes did not conclude, as had Sraffa, that there is no natural rate of interest. Rather, he made a very different argument: that the natural rate of interest is a useless concept, because there are many natural rates each corresponding to a different the level of income and employment, a consideration that Hayek, and presumably Fisher, had avoided by assuming full intertemporal equilibrium.

This last assumption is, according to Eggersson and Merohtra, still crucial for ‘microfounded’ models (which are not founded upon micro-relations at all, but that’s another discussion, see the end). But adding even a little realism to the model leads the model away from equilibrium, even in the long run…

In Summers’ words, we may have found ourselves in a situation in which the natural rate of interest – the short-term real interest rate consistent with full employment – is permanently negative … It may seem somewhat surprising that the idea of secular stagnation has not already been studied in detail in the recent literature on the liquidity trap, which does indeed already invite the possibility that the zero bound on the nominal interest rate is binding for some period of time due to a drop in the natural rate of interest. The reason for this, we suspect, is that secular stagnation does not emerge naturally from the current vintage of models in use Read more…

Categories: Uncategorized

The english recovery (?)

The ONS has published a new report on the recovery (?) of the UK economy. Some snippets:

While aggregate output has grown strongly in recent quarters, Figure 2 suggests that GDP per capita – a measure of output per person in the economy – has only recently started to recover. This difference is particularly pronounced in Panel B of Figure 2. While GDP has closed on the predownturn peak, GDP per capita remains some 6.1% below the level in Q1 2008, and is little higher than the level first achieved in early 2005.

Also, 72% of the increase of the number of self-employed was caused by an increase in the number of self-employed of 50 years and over of age, about half of these were over 65.

Also, the old and feeble also increasingly work for the young and healthy:  Read more…

Categories: Uncategorized

Unemployment: is there always a market clearing price? (Netherlands very long run edition).

April 7, 2014 4 comments

Frances Coppola has an interesting post on ‘why labour markets don’t clear‘. She points to the fact that during downturns,

the market-clearing price of labour can fall to below the minimum needed to sustain life.

When wages are at starvation level, hours worked, labour force participation rate and workforce size all decline as people become weak, ill and eventually die – or, if they can, leave for somewhere more prosperous.  Reducing the size of the workforce means that the market will eventually clear and wages start to rise again – for those who have survived.

This is the fundamental flaw in the “sticky wages” argument. In an economic downturn, the labour market cannot clear without incurring unacceptable social costs. Malnutrition, starvation, disease and death are the  consequences of freely falling wages in an economic downturn. The reason why labour markets don’t clear is because we don’t want them to.

I do not entirely agree. Let’s take a look at the Netherlands.

Unemployment

Source: Centraal Bureau voor de Statistiek. Data for the first decades are probably pretty shaky, the 1813 level must for instance have been higher. The rise around 1850 is however real, though the magnitude might have been different. Read more…

Categories: Uncategorized

Lower relative and even absolute wages did not lead to lower price levels in the Eurozone, up to 2012

Do lower relative or even absolute wages lead to a lower absolute or price level, as implied by the at least some of the versions of the ‘New Keynesian pricing Equation‘, other things, like total employment, equal? No, they don’t. At least not in the short or even the medium run. 2012 wages in Portugal and the UK (Euro price level) were about as high as in 2004. Greek wages were even lower (and continued to decline in 2013…). But the price level in these countries increased about as much as the price level in other EU countries.  This is an important fact. It means that slashing wages leads to a massive erosion of purchasing power, with, of course, dire consequences for expenditure, employment and all that. Employment won’t be equal and its decline will aggravate the slump. It’s a very Old Keynesian situation.

Priceswages

Austerity ideology states Read more…

Categories: Uncategorized

Deflation has arrived in at least five Eurozone countries, including Finland and the Netherlands

April 5, 2014 1 comment

Mario Draghi will have to push for wage increases as disinflation continues, in the Eurozone and as quite some countries are already experiencing deflation. Deflation is vicious – especially when, like in the euro zone, debt levels are high while nominal debts are highly rigid and sticky.  On the Eurozone level, there has been quite some disinflation and during the last 9 months inflation was in fact close to zero. And while I do not expect a prolonged period of average deflation as long as average nominal wages keep increasing with about 1%, even ‘lowflation’ will be devastating for the possibilities of quite some countries to pay back their debts. Especially as zero average inflation will, arithmetically, mean that quite a number of countries are actually having deflation. The only way out is not monetary policy (which takes to long to work, if it works at all) but higher wage increases, especially in the Eurozone ‘core’.

Below, data on seasonally adjusted domestic demand inflation in the Eurozone and in individual Eurozone countries. Domestic demand inflation is a broader concept than consumer prices and also includes prices of real investments and government purchases. What do these graphs show?

Deflation2

A) There has been (as we all know) quite a bit of disinflation in the Euro area (graph 1). Read more…

Categories: Uncategorized

The weird world of DSGE inflation metrics

April 4, 2014 7 comments

Whenever a DSGE economist uses phrases like ‘fundamental’, ‘deep’, ‘sound’, ‘data-rich’, ‘non-trivial’ or ‘micro-founded’ – beware. The opposite will likely be the case.

Robert King and Mark Watson point out another sorry example of this. They, literally, deconstruct the inflation variable used in a Smets-Wouters model (Smets being the head economist of the ECB) and a Gali-Gartner model, a variable  mistakenly named ‘fundamental inflation’. And this ‘fundamental inflation’ metric turns to be totally unrelated inflation as you and I know it. Some quotes:

We study two decompositions of inflation, motivated by a New Keynesian Pricing Equation. The first uses four components: lagged inflation, expected future inflation, real unit labor cost and a residual. The second uses two components: fundamental inflation (discounted expected future real unit labour cost) and a residual …

From 1999-2011 fundamental inflation fell by more than 15 percentage points, while actual inflation changed little. We discuss this discrepancy in terms of the data (a large drop in labor’s share of income) and through the lens of a canonical structural model (Smets-Wouters (2007)) … While actual inflation is essentially unchanged over the post-1999 period, the measure of fundamental inflation constructed along Gali-Gertler lines fell by nearly 15 percent and that implied by the Smets-Wouters DSGE model fell by nearly 20 percent. That is, both measures of fundamental inflation predicted large deflation over the last decade.

What happened? Labor’s share in income showed a dramatic decline – as is also shown by the familiar graphs comparing real wages per hour (stable) with production per hour (up). This is however not visible in the Smets-Wouters formula, as they mix up micro relations with macro relations – there literally is no changing ‘labor share’ possible in their world. Which rules out inflation caused by increases in profits, taxes, ‘mixed income’ of the self-employed (wich is not included in real unit labour costs), changes in the sector structure of the economy and comparable factors. As they do not take these factors into account they are in fact kind of estimating the declining labour share of income. And call this: ‘fundamental inflation’ (taking their model serious they should of course have pushed for massive increases of wages – but I’m afraid that did not happen…).

By the way – King and Watson define real unit labor costs as nominal labour costs divided by nominal income. Which is wrong.  Real unit labour costs are a derived indicator calculated by taking a series of nominal labour costs in different years and dividing this series by a series of real production in these years.

I’m preparing an article titled: ‘Metrics-meta about a meta-metric – a critical history of the concept of the price level’. However – either my table or my head may give way before it’s finished.

Categories: Uncategorized

Some links, 4/4/2014. Worried banks, What is capital, World Bank made killing mistake, Slave labour, L’amour (not?)

April 4, 2014 3 comments

Straight from the financial world (and they do not even mention Greece…):

2015 will not be the year when Spain, Italy and Portugal return to normal. When we look at the dynamics of unemployment, public and private debt, household and corporate solvency, and industrial production capacity, we see that it will take from 5 to more than 20 years to really return to normal. During this very long period, their economies will remain fragile; the Southern euro-zone countries will remain under the threat of a return of investor pessimism.

I love good writing about the relation between the concept of an economic variable and its measurement. Jamie Galbraith does an outstanding job when he discusses the concept ‘capital’. A taste, from his review of Piketty, ‘Capital in the Twenty-First century‘:

What is “capital”? To Karl Marx, it was a social, political, and legal category—the means of control of the means of production by the dominant class. Capital could be money, it could be machines; it could be fixed and it could be variable. But the essence of capital was neither physical nor financial. It was the power that capital gave to capitalists, namely the authority to make decisions and to extract surplus from the worker.

Early in the last century, neoclassical economics dumped this social and political analysis for a mechanical one. Read more…

Categories: Uncategorized

Explaining the difference between supply and demand to John Cochrane

April 2, 2014 1 comment

John Cochrane, a well-known Chicago economist, seems to think that a ‘sudden stop’, a sudden slowdown of private capital inflows into a country, leads to inflation. A ‘sudden stop’ is one manifestation of what is also known as a ‘credit crunch’. Cochrane states about these crunches:

“If we just had a credit crunch, we would expect to see stagflation–lower quantities sold, but upward pressure on prices. A credit crunch, like a broken refinery is a “supply shock.”

Sigh. Less credit does not directly lead to less supply. It leads to less demand and therewith to a pressure on prices and only subsequently, as companies can’t sell their stuff, to less supply. Read more…

Categories: Uncategorized

The UK – even a flexible labour market still is a LABOUR market

March 29, 2014 1 comment

In their book Animal Spirits George Akerlof and Robert Shiller state that wages are not set equal to the marginal productivity of labour but according to social norms – i.e. equal to (a part of) other wages. A wage is not just the atomistic market remuneration for your labour. It’s also a powerful social token of the respect you earn and a neon sign of your position in society. You’re paid for a position – not for your work. As far as I’m concerned this characteristic of wage setting is more important than the famous ‘stickiness’ of wages. The ONS recently published some information which corroborates this view of wage setting (graph): the only constant pattern in the graph below is the stunning conformity of wage developments in the UK services sector and the UK manufacturing sector – also during the disastrous decline of UK manufacturing before the 2008 devaluation. Even Schumpeterian dynamics were less strong than social norms. Mind that the UK labour market is supposed to be one of the most ‘flexible’ of the European Union.

sectoren

Categories: Uncategorized
Follow

Get every new post delivered to your Inbox.

Join 9,516 other followers