Sunday, February 04, 2018

Protect Diversity Not Competition

I have mentioned before we should get rid of the competition watchdog, and have laws to protect diversity not competition. If we protect competition then there can "only be one", and then the competition is over: everything else is wiped out. Not a very sensible thing to do.

So following on from the previous post about political parties, a similar approach could be applied to other areas. That is limit market share to 20%, in all its variations. So having the big 2 is not acceptable but the big 5 is tolerable.

Currently when the competition watchdog investigates, the big department store or supermarket is seen to do no wrong, and is permitted to move into an area and wipe out all local business. Protecting diversity rather than competition should prevent such action.

A big supermarket may only have 15% of the market nationally, but locally it can hold 100% of the market. The latter is not acceptable, it violates the 20% rule. But what is local?

Previously, I have indicated that a circular cell 5 km in diameter can be surveyed by one person in one day, assuming they can walk at 5km/h and otherwise walk for 5 hours/day. Such cell can be traversed in 1 hour, its takes 30 minutes to get from perimeter to the centre. It takes a day to walk from the centre around the perimeter and return to the centre. Various zigzag paths can be taken around the cell to explore it in 1 day. The world's land mass contains over 7 million such cells, and with a population of around 7 billion, then 1000 people can occupy each cell. A 1 km square cell with a hub 500m x 500m, can provide home to some 5000 people in single person single storey dwellings. Such dwellings are equally suitable for 2 persons and a baby, extended to two stories and the dwellings are suitable for two adults and two children. Thus 20 thousand people in a 1km square.

Any cell greater than 5km diameter is likely to require mechanised transport, at the very minimum a handcart or bicycle.  A cell 1km in diameter is potentially the more walkable, and less dependent on machinery. Though mechanised transport is likely required to get goods in and out of such cells.

Any case it provides a starting point for setting some limits. Population limited to 200 to 4,000 people, or radial extents limits to a distance from 0.5km to 2.5km.

So if we pick a supermarket at random and locate it at the centre of a 5km diameter cell, then inside that cell there should be found, at least another 4 similar independent stores. If not then the supermarket has too large a market. A second check is the size of the population, it actually serves compared to available population: if it supplies more than 4,000 people from the one store then it is too large. That is the simple check.

An alternative check is to move the centre of the 5km cell until it captures 5 stores within its boundaries. Here on the York Peninsula in South Australia, the distance between towns is about 25 km, and each town typically has a population less than 1000 people, some have populations less than 100. The rural roads between towns typically have speeds at 100km/h, so each town is around quarter of an hour away, though a town with a commercial hub could be 1 hour to 2 hours away.

In such situation the hinterlands (catchment areas), need to be larger in diameter to capture a large enough population to make the businesses viable in the first place. Which raises the question of the minimum population required to serve for the business to be viable. The initial assumption was local population of 1000 to 20,000 people to be served by a minimum of 5 businesses: so an individual business can only serve 200 to 4000 customers.

Clearly businesses which supply carports and verandahs or sheds, serve the entire Adelaide metropolitan region, even the entire state of South Australia, and some companies the whole of Australia. Most builders have indicated they sell 40 carports/verandahs a month. Since the number is so common I doubt that is reality, more like some ideal figure dreamed up by some common source, now lost in time.

In terms of timber , most such structures take carpenter 2 days, with help of one assistant. However there may be some preliminary factory work done before site works, such as cutting, and painting. So assuming 5 days in a working week, and 4 weeks in a month, then 20 days a month: then one team can only supply 10 canopies each month, not 40. Or assuming 50 productive weeks a year, then 250 productive days in a year, then only supply 125 canopies a year, not 12*40=480. And who wants these?

For some reason houses aren't designed and built with verandahs.Taking South Australia with my rough and ready, easy to remember statistics, there is approximately 1.5 million people, with household occupancy less than 3 persons per dwelling. An old Australian Bureau of Statistics (ABS) report indicates that some 5% of dwellings each year get some form of renovation: kitchen, bathroom, extra room, outdoor living (decking and/or verandah). So first requirement for the industry would be to supply the 500,000 dwellings. Thereafter provide replacements, at the rate of 25,000 each year. So would require 200 teams, or individual businesses. Of course this can be made more complicated by taking population growth into consideration. Last time I looked, the number of dwellings were around 600,000 and occupancy around 2.8 persons.

Any case, the simple numbers puts each business serving around 7,500 people, spread around the whole of South Australia. Local business however is unlikely to be viable. As move away from the hub of Adelaide, the size of hinterland needs to change for local business, from 5km diameter to possibly over 300 km radius: a radius for the latter because most likely only serve in a small sector rather than a full circular cell. For example Kadina, in South Australia, basically serves everything to its South on the York Peninsula. Everything south of Kadina, has a both a smaller population and fewer facilities. Kadina has a population of around 5000 people, everything to its south has a population less than 1000 people. Kadina is in the Copper Coast council district, and that to the south is in the Yorke Peninsula council district, each district having a population around 12,000  people. For such region, before consider protecting competition or diversity, the first requirement is to provide service. Whilst some specialist stores are viable, more likely to find general stores providing a broad range of goods: and having 100% of the market with respect to those goods. Likewise the specialist stores also hold 100% of the local market. Local with respect to the limits of the township. People in such regions also likely to work part-time across a range of activities. This is because no single activity is likely to provide adequate income to make a living, secondly because it is difficult to get people into the regions and so people have to "wear different hats" if anything is to get done. The problem however is that makes the towns like ghost towns, as stores are closed and services only available by appointment: not at all welcoming to visiting tourists. Not only do businesses need to be in a network but so do their customers need to be in such network: because without network connections no one would know the services are available.

So we have an environment different than that in the big city. But Australia itself is an environment different than that in other industrialised nations. My typical reference is that Australia is akin to a remote mining colony on Mars. Except if were on Mars, the long wait for the bat bringing goods and services in would be more apparent, and the remote location would be better managed. Then again Australia is only remote from the west, it is not remote from the east. If we were on Mars, then we would more likely develop better relationship with the martians (assuming they may exist), than we have done with Asia.

But that is the other problem with protecting diversity, its attack from external competition. Local towns are not just under attack by national companies, they are also under attack from foreign powers. My other common point is that if land defines a nation, then land should not be permitted to be bought and sold. Land should be owned by the nation state, and the state should only grant license to occupy and use the land: not own the land.

Protecting the land is important with respect to farms and mines. The 20% limit, would restrict ownership of resources, or license to resources. So no farming business could hold license to more than 20% of the farmland, to 20% of the irrigation water resources. Further restriction to no more than 20% of milk production, nor more than 20% of beef production, etc... Similarly no mining company hold more than 20% of steel production, 20% of aluminium production.

As done above for the canopies, can work out how much work one person or a team of people can produce in one year. Secondly need to check the demand for such work over one year, and determine if that is repetitive from one year to the next, growing or decaying. Once population stabilises, with basically one person born for each person who dies, there is no need for increased house construction: unless people want multiple houses scattered all over the planet: but then once the planet is fully occupied by humans, then house construction no longer needed. However, house maintenance and/or replacement will then need to occur. If something is needed then we can get some idea of demand, compared to pointless widgets which are just made and sold just for the purpose of surviving in a market system.

Government services increasingly privatised. So once again, supplier should be restricted to 20% of the market, and the choice available locally, not just nationally. The construction of Adelaide Hospital for example would not be permitted, its expected catchment area is too large, and it has been constructed at the expense of largely ignoring rural and regional hospital services. The principle would require that South Australia has at least 5 large regional hospitals. Such would require South Australia to be divided into 5 regions, with a central hospital which can serve such region. Not 5 hospitals located based on population, but based on geographical area. The point of the state or nation is development of the geographical region, having the capability to occupy, hold and serve the region.

So we start by dividing regions progressively by 5. Thus Australia has 5 or more regions, then each of these regions further divided by 5, and then each of these in turn is further divided by 5. So South Australia is one of the national divisions, which gets divided into 5 regions.

Why 5? Because in group dynamics, two people can easily disagree and then nothing gets done, with 3 people, it can easily split into two groups, whilst get a 2/3rds majority can otherwise loose 1/3rd taking any interest. With a group of 4, there is a possibility of an even split, and thus two groups working against one another. With 5 people, need 3/5ths to create a majority, there is no even split, and whilst may have a 2/5ths opposition, its likely get two opposing groups of 1/5th each. A group of 5 is therefore the smallest odd number group which seems practical.

Geographically however, hexagons are important to filling space uniformly. A hexagon can be split into 6 segments: 6 being an even number can provide an equal split, no majority view. However hexagons, and circles form clusters of 7: an odd number and potentially better  than 5. Adopting 7 would reduce acceptable market share down to 14% limit. Each 7 cell  cluster would form a community, but persons could seek service from any of the 7 hubs. I don't have a problem with that, however most businesses with market dominance typically have market share between 20% and 30%, so adopting 20% limit doesn't impose too much of a burden on existing business except those which really do hold a majority of the market. Microsoft for example with near 100% market share of computer operating systems.

So what to do about companies like Microsoft? First back to the division of regions. So South Australia gets divided into 5 regions, most of those regions would be without any significant population, and furthermore the distance to any central hospital would be too far to travel for an emergency: hence why investment in the one Adelaide hospital is not acceptable for the state. We therefore continue dividing the area by 5 until create cells with populations and within reasonable distance of the needed facility, which could be hospital, school, shop etc... Those cells with populations should get facilities, but if the land is to be considered occupied then need to consider how to service the remote unpopulated cells assuming it is permitted to visit these cells. For example if helicopters have operational radius of 250 km, then need heliports at no more than 500 km centres if expect to be able to search the Australian interior for lost tourists. Heliports need fuel for the helicopters and other supplies for the personnel: how are those resources going to get there? It is not really appropriate to consider markets for everything, but if going to get held to ransom by lone supplier, then protecting diversity is important.

Diversity however doesn't help if there are pressures to supply at equal price. For example Engineers Australia, and Institute of Architects, and institute of building designers all have members who complain about undercutting fees, and attempt to publish guidelines to what fees should be. The fee guidelines typically take an elitist unrealistic viewpoint: not everyone gets to rip the government off,charging extortionate fees for public infrastructure. Engineering effort unlike construction effort is not proportional to the size of a structure or other system, but proportional to the complexity: if two structures have the same structural form but only vary in size, then they require the same effort to design. The engineers and architects working on the larger project will get the higher fees.

Consider that a garden shed takes the same design effort as a large industrial building. Say the garden shed costs $500, whilst the larger structure costs $1 million dollars, they both have the same structural form. The consulting fees for large projects may be defined in terms of percent fee, typically less than 5% and apparently usually around 1%. So for the garden shed the consulting fee would be $5, whilst that for the industrial building would be $10,000.

Australia's federal minimum wage is $17.20, so for the sake of argument assume work capable by one person producing drawings and calculations, no site supervision. Assume design effort capable of being completed in 1 week, of 40 hours. So labour cost is $688, the percent fee doesn't cover the cost for the garden shed, but the fee for the industrial building is 14.5 times larger than cost. However ignored operational costs of the business, assume needs of business is similar to a person and minimum rate is same as federal minimum wage: so we multiply by 2. Therefore cost is $1366, most definitely not the kind of fee the person buying a garden shed wants to pay. Whilst the fee for the industrial building is still 7.3 times larger. Is such multiplier an acceptable relativity between wage earners?

Now the cost for the small garden shed indicates that really needs to be designed once and made many times: but most manufacturers also do not find the fee acceptable. They want lower price. That therefore means the cost of the drawings and calculations need to be dropped below 40 hours: the process needs to be automated: electronic computers are needed. When automation of design takes place then the market held by the architects and engineers is displaced over to the manufacturers. The high fees they once demanded disappears: the value of their services diminishes. A service which mostly was imposed on projects, where such need was never previously required. A monopoly given by legislation, taken away by technology and market forces, a monopoly which should have never been permitted in the first place. Impose purchase of something which wasn't previous needed, and the buyers will take measures to remove such imposition. If a market imposes a need or generates a demand where  need is not necessary, then market will also seek to remove. A dynamically adaptive system will attempt to return to some lower energy level: the level before unwanted impositions.

Australian industrial awards and Federal minimum wages  give some idea of minimum costs, if buying and selling time is the only thing of concern. Thankfully at the moment industrial awards haven't been pushed up to match market rates. Thankfully because the awards tend to only go upwards, and therefore stifles new business, and favours big business with large markets. More to the point it benefits employees with big business to increase minimum wages in awards, as it can close down smaller business, destroy the competition, destroy diversity. Then once other suppliers removed the big business steps in and supplies a larger market, and potentially at higher prices.

Local community groups are better than labour and trade unions. Trade unions were historically more committed to local communities and did more social good than harm. Now most trade unions are limited companies, full trading businesses, national, and no real care about community. Communities however can fight back. Back in 70's, one local community blocked a trade union picket line. The union was potentially going to shut off the power supply, blocking access to power station: the community prevented the union members from leaving the picket line and from refreshed members reaching the line. The community could hold out longer, than employees and unionists who didn't live in the area. A community doesn't have to be held to ransom by outsiders.

Similarly communities do not have to tolerate privatisation of their resources. They can buy resources back. I recollect a documentary about water supply where one French village regained control of their local water resources after it had been privatised. Communities should have control over that which their survival is dependent on: sovereignty over dependent resources. What is the point of nations, of fortified towns, if they are permitted to be attacked by economics?

Should have free markets, open markets, no regulation? Who does this benefit? Mostly big multinationals: entities which have no geographical boundaries, and no social responsibilities, and which are increasingly more powerful (economically) than many nations. However if they are highly specialised they are not overly sustainable. If they are involved in everything, are corporations with voting shareholders, then we have a future world, governed by something other than nations and without geographical limits. Instead of being a citizen of a nation, you will be a member of an world corporation. There would have to be at least 5 such citizen corporations in the world.

People don't migrate between countries, they migrate between cities. People in cities may migrate to more remote rural regions, in the main however people in rural regions migrate to cities. People changing countries are really changing cities. It concerns the critical mass of population required to make certain goods and services available. The difference between making one combined harvester every 10 years, or making several hundred each and every year. What is the critical population required to design and produce computer chips? The population required to get started from scratch is not the same as required to start up once the technology exists.

One article I read indicated something along the lines that Britain took 200 years to industrialise, the USA with import of technology from Britain took 100 years, and Japan with technology imported from the USA took 50 years. After the second world war, redevelopment took 25 years, nations with the help of Japanese industry have taken 15 years to become industrialised. Developing technology requires people and time. With technology development takes less and less time, as all existing technology is a foundation and building block for future technology.

So geography may not be relevant to the future. Brexit for example could turn out to be an entirely pointless exercise: as corporations become increasingly more relevant than geographically bound nations. Passports and visas shouldn't be necessary, no one should need a permanent home address and all should be citizens of the world.

Protecting national boundaries from illegal immigrants is it really important? I doubt it! I would say the current problem isn't so much that illegal immigrants get into the country, but that they cannot get out. I suggest that it is easier to get out of Africa, and get to London, than it is to get out off London: one way traffic. Thus having discovered that no economic benefit to being in London, it is going to require a lot more effort to get to somewhere else. But where else: New York, Tokyo, Sydney, Paris, Berlin, Rome. Where? Is life going to be better there? Want to learn a third language? What are the 5 most common languages? Where are the 5 most popular cities?

Population doesn't need to move if local town provides for its needs. So protecting diversity also includes protecting the survival of local towns. In Greece and Italy I believe some towns have become ghost towns, where only the retired remain: the younger generation have left leaving the retired behind. The big cities stealing the population. So the 20% limit can apply to the population of cities, to the employment of people by business.

As for companies like Microsoft. Their upgrades are not upgrades they are different products, we are held to ransom, forces to adopt new product, and new way of doing things. But there are people attempting to create alternatives like FreeDOS to keep old software going, others attempting to replicate Windows XP. The propaganda machine says such software is a security risk: well it wasn't when we first bought it, and it was operating perfectly fine for the majority of people. So no real need to change.

So here's an idea. If Microsoft no longer wishes to sell or support MS DOS or Windows XP, it doesn't have to. On the other hand it shouldn't be permitted to prevent others from supporting such software. So it could be required to release its source for its abandoned software, thus FreeDOS doesn't need to reinvent the operating system: rather it can develop it in another direction. In other words microsoft releases products (it can license) development of its old products until usage of any given product drops to 20%.

... interrupted for tea and telly. [17:47]

[22:25] ...
Would Windows be a prime target for attack if it only represented 20% of computer users, and another 4 systems were available for attack. Currently has I understand the majority of users are on Windows 10, then earlier versions of windows, Apple, and various distributions of Linux: all largely determined from connections to the internet. Computers however do not need to be connected to the internet, and such computers do not need need security protections. So MS DOS is perfectly fine for crunching numbers, writing letters, as is CPM/80.

SunOS used to be able to spawn an MS DOS process in its own window. As far as I know it was not a virtual machine, it was an independent process. Early MS DOS installed on 720 kb, 3.5" floppy disks, whilst CPM/80 installed on 360 kB 3.5" floppy disks. Gigabytes of data are not required for the operating systems: so Windows 10 should be able to run separate windows with any of the earlier versions of Windows and all prior software should still be able to run. But it doesn't, and since software ceases to run when replace hardware and operating system: the software is not an upgrade it is a different product.

So amongst first things to do is impose the 20% on the types of licenses. For example could restrict OEM licences to 20% of licenses, or require at least 20% of licenses  are NON-OEM. But would need to split market into sectors. If not split into sectors likely to have the suppliers declare that the 20% has been reached by supply to governments and large corporations, and so no need to supply to individuals. By separating business from personal, the 20% limit can be imposed on each sector.

Similarly can require that 20% of hardware is supplied without any operating system, and is capable of having any operating system installed. Can require that the hardware is suitable for at least 5 different operating systems, that there are at least 5 different operating systems available.

Here the limit is placed on the product not necessarily the business. So Microsoft can license its earlier operating systems to be developed and maintained by others, so that there is a greater diversity of products in the market. It doesn't necessarily have to license or provide access to the source code. So MS DOS can be licensed as a foundation for building variants. For example FreeDOS is not entirely capable of running all earlier MS DOS software. On the other hand FreeDOS can be installed and operate on hardware which wasn't available when MS DOS was originally available. Many of the original external commands in MS DOS could be replaced by Linux tools. Linux operating systems could run MS DOS in a separate process.

Historical cars can be kept going because the parts do not require any significant infrastructure to fabricate. People can machine parts in a small workshop. Computers and other electronic hardware however are more complicated to build than is possible with tools found in the family garage.  But with 3D printers, then maybe it will one day be possible to fabricate digital chips in the home.

Any case markets can comprise of products, and should protect diversity of products. Cars for example there should be at least 5 manufacturers. And following what I said for computers, each component should have 5 potential substitutes. So 5 different, independent suppliers of tyres, 5 suppliers of spark plugs, think of a component then that component should have 5 suppliers: not just suppliers 5 different manufacturers.

If there is a definable market, and a business holds more than 20% of that market, then its share is too high. A business doesn't necessarily have to do anything, since next week, it could hold a smaller proportion of the market. The big issue is when a business holds more than 50% of the market, and other businesses hold less than 20%. In such situation may have diversity, but the situation is heading towards monopoly with the other players potentially being removed from the market. Such situation doesn't necessarily represent what people want, but rather what they can get. The other players just don't have the resources to produce quantities large enough to flood the market: they don't have the distribution network and therefore people have neither knowledge nor access to alternative products.

It is not just a matter of reducing the 50% dominance of the market down to 20%, it is a matter of boosting the other players so that they have nearer to 20% of the market. Rather than manufacturers being the problem in such situation, it maybe the wholesalers and retailers who are the problem: they are not giving other manufacturers the opportunity.

So the supermarket shelves have to have products from at least 5 different producers in each and every product category. This doesn't mean one box from 4 suppliers and a 100 boxes from one dominant supplier. The stores product mix has to include 5 suppliers with near equal demand quantities. Unless there is clear bias in the local community. In which case it can cater to the bias, but it should otherwise seek to break such bias. It should break the bias because such is not good for the future survival of the community and not good for the survival of the store: shouldn't have all eggs in one basket. To ensure security of future supply, should have a diversity of products on offer, and individuals themselves should also use a diverse range of resources rather than relying on one.

Economy of scale maybe efficient, but it is not secure. Economy of scale gives greater importance to efficiency over effectiveness. A system cannot be efficient or economical if it is not effective. The system will not be effective if it is not secure. Diversity provides security because if loose one resource there are 4 alternatives to take its place. If only have one resource and lose it, then lose everything.

It's the problem with energy. Fossil fuels are not so much the problem, the problem is that such fuels dominate the market. Adopting solar power or wind power in the place of fossil fuels will just create a different set of environmental and economical problems. We need a diversity of energy sources with no single one permitted to dominate. Coal, Gas, and Oil are not 3 independent fuel supplies, so having these 3 supplies isn't adequate diversity: we need independent alternatives. Solar and wind are likewise related, both driven by the Sun.

Whilst the 20% rule may not be entirely practical to implement, it does give a guideline to direct investigation. Smaller business doesn't need to grow to dominate the market, but it does need to prevent others from dominating the market, it does need to help ensure diversity is sustainable.


Related Posts


  1. [04/02/2018] : Original