Summary: It´s meaningful to value fixed assets in the national accounts. As, however, not all transactions are capitalist market transactions and not all prices are capitalist market prices, the ´method of choice´ to value fixed assets proposed by the new national account guidelines, discounting expected future flows of income, is not as universal as suggested. Instead of one method, an amalgam of methods should be used, tailored to the specific ways fixed assets are used.
The modern national accounts, which nowadays incorporate the financial accounts or ‘flow of funds’, are the most important set of macro-economic statistics. They are based upon guidelines issued by international organizations. At this moment, there’s a new set of guidelines that has to be discussed. Previous blog posts discussing them can be found here.
Chapter 17 of the new guidelines is about how to value ‘fixed assets’. Valuing these assets yields an estimate of the value of ´fixed capital´. Knowing this value (and its components) is necessary to be able to, using a neoclassical growth model, estimate the contribution of ´fixed capital´ to the value of GDP.
In Chapter 17, GDP is implicitly understood as ´capitalist market value added´. Which it isn´t. It also encompasses non-capitalist production, for instance, the provision of public goods by the government. And by but what´s called ´Non-Profit Organizations Serving Households´ (NPISH). Also, the household sector itself produces, through its ownership of dwellings, a considerable share of GDP.
On top of this, there´s a long tradition that recognizes ´mixed income´. This is the income of self-employed. It cannot be classified as interest, profit, or wage income. Self-employed individuals generally charge one fee. They do not charge several fees for both labour costs and capital costs. Uber drivers are a marked characteristic of modern economies. They earn an income. But not a wage. Or a capital income. It´s called ´mixed income´. In a sense, using the Marxist phrase, these are people who own means of production. They have more to sell than just their labour.
The valuation of fixed capital should recognize this. Fixed capital is the value of fixed assets; fixed assets are stuff created by humankind that lasts more than a year and is used for public or private consumption or to produce other stuff. However, ´non-produced´ assets like copyright are also considered fixed assets. Economists want to value these assets. The ´method of choice´ to do this, absent a market price (and market prices are often absent), is according to the manual, to discount expected future flows of income. That´s murky business. We do not know these flows. And we do not know future interest rates. Also, these flows of ´capital income´ are dependent on wage rates and market power of companies and hence on labour and competition policies of governments. And what´s the expected future flow of non-labour income net of variable costs of a bridge? A tank? An MRI machine, which is an integral part of a hospital? There’s a lot to unpack here. To do this, we have to pose the following five questions:
- What is ‘fixed capital’ according to the national accountants?
- Who owns ‘fixed capital’ according to the national accountants?
- Where does ‘fixed capital’ come from?
- What is the difference between ‘capital’, and ‘fixed capital’?
- Is, considering the answers to these questions, the method proposed in the guidelines to value assets a sound method?
- What is fixed capital?
According to the definitions, fixed capital is the combined value of almost anything built by humankind that lasts longer than one year and is used for private or public consumption or for productive purposes.
According to current estimates (which are based on discounted future income flows only rarely), fixed capital consists mainly of buildings and other structures (including the land underlying them). Next to these, there is a long list of items like animals, machines, computers, transport equipment, ‘artistic originals protected by copyrights’, and the like. But, when we look at their value, these are all much less important than our built environment. Some items that do fit the concept are left out. Artificial hips and the like might be included, in my opinion, but aren’t. Plastic pollution (which would have a negative value) is also not included. The list is at the end of chapter 17.
Source: Centraal Bureau voor de Statistiek
Graph 1 shows the distribution of the value of assets for the Netherlands (where I live). Buildings (including the land underlying them) account for the bulk of the value of the total stock of fixed assets. Again and again, in any age (post-1800) and country, buildings are the number one fixed asset. In the Netherlands, over ¾ of the total value of assets consists of buildings. Even the combined value of all transport equipment, including planes, trains, all automobiles and bicycles, is roughly 1/14th the value of dwellings alone. Including civil engineering (roads and bridges) brings the total value of our built environment to a whopping 83% of the value of total fixed asses (The data of the Centraal Bureau voor de Statistiek (CBS) show that the combined value of trains, planes and automobiles is 2 billion less than the total value of all transport equipment. I assume that these 2 billion relate to bicycles.
Breaking down the data on buildings shows that roughly half of the total fixed capital stock consists of dwellings, which are often owned by individual households or social housing corporations (NPISH). And increasingly by organizations like BlackRock and Blackstone, but the total number of dwellings owned by such organizations is still relatively small. According to its yearly report (p. 129), the large Dutch ABP pension fund owned 84 billion euros (over 10% of its total wealth) of real estate in different countries (31 December 2024). This compares with the 1,338 billion euros that Dutch dwellings were worth in 2024, according to the CBS. Summary: fixed capital mainly consists of the value of buildings and built structures.
2. Who owns physical capital?
Graph 2 shows a breakdown of the ownership of the total value of fixed assets in Italy and Germany. The national accounts distinguish several sectors: households, the government, financial corporations (with the central bank as a spider in the web), non-financial corporations, and NPISH, such as churches, amateur soccer clubs, social housing organizations and labour unions. In the graph, NPISH are included in households.
Source: Eurostat
Households own the largest share of fixed capital – by far. Financial corporations, such as banks, mainly own financial assets but not much fixed capital. The government and non-financial corporations are also large owners. The graph does not show the distribution of this wealth. The aggregate value of fixed assets owned by households is enormous – but it will be clear that generally, renters do not own much housing wealth. According to Eurostat data, the patterns shown by Germany and Italy are roughly comparable to those in other European countries, even though average wealth per capita differ.
When we break down ownership in different classes of assets (graph 3), it shows that ownership of fixed assets of households is dominated by, not surprisingly, houses. Even then, quite a few dwellings are also owned by non-financial corporations. Again, the relatively low importance of ‘other’ fixed assets, like trucks, computers, machinery, and stocks of animals, stands out, even for the sector of non-financial corporations. The graph shows this for Italy, but differences with other countries are, according to the Eurostat dataset, limited.
Source: Eurostat
Mortgages often finance house ownership. In these cases, it is assumed that households have full ownership rights. The debts to the bank are, of course, part of the debt side of the balance sheet of the sector ‘households’ and the ownership side of the financial sector. The data in the graphs do not show net wealth per sector. Graph 3 also shows that, in Italy, ownership of non-financial corporations is dominated by buildings (and the land underlying them), too.
Summarizing: households are the most important owners of fixed capital, followed only at a distance by non-financial corporations and the government. In all cases, ownership is dominated by buildings and other built structures.
3. Where do fixed assets come from?
Fixed assets are, according to the national accounts, created by matching expenditure on the production of fixed assets. Below, some long-term graphs showing ‘gross fixed investments’ are presented. This is, afaik, the best long-term overview on this. The World Bank has an excellent dataset covering basically all countries in the world, but which for many countries only starts in or a few years after 1960. For instance, the World Bank data for Japan only begins in 1970. See Knibbe (2014) for more literature.
Sources: Central Bank of Japan, Central Bank of Sweden, Bureau of Economic Analysis, Worldbank, Centraal Bureau voor de Statistiek, Bank of England, National Bureau of Economic Research, Gallman and Rhode.
The graphs contain a lot of interesting information. But the points here are (a) that during the period of modernization, investment rates rise. And (b) much more so for latecomers to the processes of modernization and economic growth. Possibly, this last effect occurs due to the ability to catch up on new technological possibilities. Being backward initially also contributes to this phenomenon. To state this otherwise, unlike 21st-century China, 19th-century US did not have the possibility or the examples to catch up on high-speed rail. Also, the most important fixed asset, dwellings, changed during this period. This was a crucial aspect of modernization. Consider access to the sewer system. Think about using bricks, tiles, and concrete instead of wood and reed. Also, consider better ventilation and matching parking places and multiple rooms, contemporary kitchens, running water, electricity, and heating systems.
Investment percentages of over 30% are nowadays not uncommon and can be sustained for decades. The low level for the UK – an early bird when it comes to modernizing its economy, i.e., investing in railways and urbanization and the like, is still surprising and even more so when we compare it to levels in the USA and Germany or even Sweden and the Netherlands, which were both relatively late to the 19th century modern economic growth party. The earlier UK data are probably without weapons systems – the graphs show that these investments can be considerable
Investments in real estate have been high. But the preponderance of real estate , including underlying land, in the total value of fixed capital does not only originate from high levels of investment in houses and other structures. Buildings typically have lower depreciation rates than trucks or even ships. At the same time, the value of buildings, and surely of dwellings, tends to increase instead of, like the value of machinery or even patents on drugs, to decrease. Also, we use more and more built space per person. Houses themselves tend to become larger. And think also of parking spaces next to malls, factories, and offices, ten-lane highways, offices, malls, and the like.
Anyway, the current stock of fixed assets results from an increase in the rate of investment. This especially occurred during periods of ‘super growth’ such as the fifties and sixties in Europe and the USA. In recent decades, it has also occurred in China, India, Bangladesh, and other countries. These investments are not just investments ordained and financed by corporations, but also by households and the government.
4. What is the difference between ‘capital’ and ´fixed capital´
Above, we’ve tracked (investments) in ‘fixed assets’. ´Fixed assets´ has a clear definition and operationalization, and the estimated monetary value of these assets is often called´fixed capital´. However,´fixed capital´ is not the same thing as ‘capital’. What are we talking about?
We might, for educational purposes, first look at a definition by Branco Milanovic. He looks, as a kind of thought experiment, through a 19th-century Marxist lens at fixed capital. To understand this experiment, we have to realize that in 18th- and 19th-century villages and cities, various tradespeople such as the blacksmith, carpenter, butcher, baker, and miller owned their own ‘shop’. In other words, they owned and used their ‘means of production’. When Marx and Engels were young, self-employment was a thing. Well, it of course still is. But other kinds of employment were less common. The fruit of the labour of these self-employed was based on their ownership of ´fixed assets´ and ‘fixed capital’ in the sense of the national accounts. Aside of this, ´laborers´, employed to use the fixed capital owned by others, existed. These had existed for a long time. But their number increased, as did the average size of the company where they worked.
According to Milanovic, fixed capital owned by the self.employed does not count as ‘capital’ in the 19th-century Marxist sense. In the Marxist sense ‘fixed capital’ only becomes ‘capital’ when it is used to employ others who do not own any productive fixed capital themselves. And who have to do the owners´ bidding. Translating this to the national accounts, only fixed capital owned by a part of non-financial and financial corporations using paid labour and paying wages is included in the 19th-century Marxist ‘capital’.
The modern national accounts recognize this situation. The income of the self-employed is not split into labour and capital income. Instead, it is aggregated in a third category, ´mixed income´. One is also reminded of Chayanoff’s work. His research is based on numerous surprisingly good studies. These studies focus on 19th and early 20th-century Russian agricultural villages. In these villages, ‘wage labour’ and ‘wages’ were largely absent. Therefore, economists could not distinguish between wage and capital income, nor was such a distinction made by the peasants. Also, their use of ´fixed capital´ (mainly land) was not so much influenced by a tendency to accumulate but, based on their own decisions, by the family cycle. Land is used and experienced as a physical entity, not so much as a store of monetary wealth.
This is different in a capitalist society. The owners of ‘fixed capital’ used for market production do not view it as purely physical. Instead, they see it as an economic entity with a monetary value. It must be used to earn money. ´Fixed capital´ (also) becomes ´capitalist capital´. After wages are paid and depreciation is calculated, this kind of capital generates a yield. It relates to an expected future flow of income. This is true even if the assets are sold and the money is invested in financial instruments, including deposits. Shimshon Bishler and Jonathan Nitzan agree with this view with their Capital as Power (CasP) approach:
‘Accumulation, we argue, has nothing to do with what economists call ‘real capital’. It has no direct relationship to the amalgam of machines, structures and technical knowhow, nor is it related to their presumed ‘productivity’. According to CasP, capital is finance and only finance. It exists not as backward-looking stuff, but as forward-looking capitalization – the discounted present value of expected, risk-adjusted future earnings. And that’s it. Accumulation is the increase of this forward-looking magnitude.’
According to Bishler and Nitzan , in the dog-eats-dog world of the capitalist economy, companies have to strive for the highest yield, to prevent being eaten. Accumulate or perish.
Josh Mason has a comparable view. He states in ´Disgorge the cash´, arguing that accumulation has shifted from investing in factories and whatever to investing in financial items:
‘Whereas firms once borrowed to invest and improve their long-term performance, they now borrow to enrich their investors in the short-run.’
Capitalist investment is not about production. Or consumption. It´s about yields. A key argument of Mason is that the ‘internal’ discount rate of investors is higher than the expected rate of return of factories. This means that companies might even be stripped of physical assets to enable pay-outs. To increase the discounted value of ‘expected, risk-adjusted future earnings’. Significantly, according to Mason, legal, cultural and other restraints on such behavior were diminished, at least in the USA, to enable investors to act in this way. At the same time, he states that actual investment decisions might not be made by individuals. Instead, these decisions are driven by more complex political processes within corporations.
Such ideas are not new. Thomas Hauner, Branko Milanovic, and Suresh Naidu test the ´accumulate or perish´ ideas of John Hobson, Vladimir Lenin, Rosa Luxemburg, and Joseph Schumpeter from the beginning of the 20th century. These authors stated that around 1900 and later, ‘capitalists’ strove to invest surplus funds abroad to obtain higher yields. According to Hauner e.a al., the facts agree.
Interestingly, the mechanics behind this Marxist-radical concept of capitalist accumulation, ‘the risk adjusted rate of return’ used to discount expected flows of money to value ‘capital’, is also the ‘method of choice’ used to value fixed assets in the new guidelines for the national accounts (lemma 17.9).
‘In the absence of observable prices, the value of an asset may be determined by the present value of its future economic benefits. Economic theory states that in a well functioning market (suitably defined) even when prices are observable, this identity will hold.´
Also, the manual gives examples using a fixed interest rate. We all know that interest rates change. Changes in interest rates, especially for long-lived assets, lead to large estimated swings in expected value. For example, one can consult the wildly varying estimates of the discounted debts to the contributors of Dutch pension funds. These are assets of the contributors. They are based on the wildly varying administered obligatory discount rate set by De Nederlandsche Bank.
In the OECD manual on which chapter 17 is based, a graph is even shown which shows the falling rate of return on assets. At first glance this underscores the arguments of Hobson, Lenin, Luxemburg and Schumpeter. Japanese people would have done well to invest abroad.
But the point: the graph is based on the value of fixed capital for the entire economy. Not just for the market economy. Do we really expect Japanese farmers to take a second mortgage or to sell their farm when, up to 1991, the price of their land increased? Do we expect them to take part in the ´carry trade´, i.e., investing the money abroad to reap higher yields? The economists’ answer to these questions is yes. But many didn´t.
An (neoclassical) assumption of the guidelines is that the whole economy can be described as a ´well-defined, suitable functioning capitalist wage-labour and interest-dependent market´. But it isn´t. The manual assumes that all fixed assets are ‘capital’ in the Marxist-radical sense. They are owned by capitalist people and organizations. These entities will search for yield, and they have to do so to prevent annihilation. All fixed assets are assumed to function in a ´suitably defined and well-functioning´ market. It´s clear that owning your own house is not part of the rental market. The national accounts themselves show that ´mixed income´ is still an important variable. We also know that not all government investment is aimed at increasing future income flows. In other words, not all transactions are market transactions, not all assets are market assets. As dwellings are the most important single fixed asset and as dwellings are largely owned by households and, in many countries, NPISH social housing corporations, this is not a trivial example. It´s about one of the main building blocks of modern economies: the non-market use of many dwellings. We can´t and we shouldn´t define the entire economy as one big market. Not all companies are capitalist companies, and not all fixed assets are used in a capitalist market-consistent way.
Despite this, we have, according to the national accounts manual (and the OECD document on which it is based), to calculate the capital value of the fixed capital used by these entities, using an appropriate (exogenous) interest rate to discount the expected future flow of income from the means of production you own and assuming markets everywhere. To quote the OECD document:
Under these circumstances, an exogenous real rate of return may be the appropriate choice. For its estimation, one would typically turn to interest rates on financial markets and select an average of key rates that bear a link to the opportunity costs of investing in non-financial assets. Candidates for interest rates are government bonds, corporate bonds, and interest rates on corporate debt of varying maturity.
All ‘fixed capital’ is required to be ‘capitalist capital’. But we should and can loosen this restriction. Dwellings occupied by owner, the truck of the self employed plumber, the land of Phoenix park in Dublin – it´s fixed capital. But it´s not capitalist capital. There are no economically significant identifiable flows of money related to ownership of the asset. To give one more example: many households own one or more cars. Which, surely, when there are two, are idle for 95% of the time. Uber is important. However, most people do not try to earn additional money with their fixed capital. They have a second car but do not use it on the Uber platform. They have rooms to spare but they do not rent them out. They have a garden but do not use it to grow potatoes. Income-poor but asset-rich households do. But the wealthy don´t. Also, the remaining 5% of the lifespan of an automobile is often spent to enable family festivities, unpaid care work and bringing the kids to soccer clubs. Cars (and dwellings) are fixed assets with an identifiable value. But they are generally not used for market activities, not even of the self-employment type.
What to do? We should stop pretending that all activities are market activities. We must acknowledge the existence of non-market activities, even in capitalist societies. We also need to value fixed assets accordingly. There are a plethora of ways to do this (used by statistical offices worldwide). Sometimes selling prices can be used to value fixed assets. In other cases, it’s appropriate to use insurance value or second-hand prices. Construction costs might also be relevant, as in the case of coastal levees. But we should stop pretending that the world is as simple as our economic models. While we also know that in an uncertain world, interest rates and technologies will change. This means expected flows of income related to fixed assets with a given state of technology can’t be meaningfully discounted. Even when the behavior of capitalist corporations can be described by discounting expected future flows of income.
