Main body

The series in one page

The first two parts established the problem.

Tanzania's large banks are profitable, but a substantial share of their balance sheets is invested in government securities. In June 2026, commercial banks held 28.8 percent of domestic government debt. At the same time, the country has only two community banks, a TZS 15 billion minimum capital requirement for a commercial bank, widespread problems with land titles, thin credit histories, and a lending environment in which a Treasury bond can be easier and safer for a bank to hold than a loan to a small workshop.

That brings us to the harder question.

What would actually have to change for banks to finance more of the productive economy?

Japan and China are useful because neither started with the banking systems they eventually became known for. Both changed the structure of finance itself. They created different types of banks, changed how credit was directed, and built institutions that could reach businesses and households beyond the largest cities and companies.

The results were not produced by banking reform alone. Land reform, trade, industrial policy, education, infrastructure, and other factors mattered. But banking determined an important question: who could actually get the money to invest and expand?

That is where the comparison becomes relevant to Tanzania.

Japan: building a layered banking system

Before Japan's post-war economic expansion, wartime finance had concentrated credit around the state and a relatively small number of major banks.

After 1945, Japan did not simply create one large national banking system and leave it at that. It built several layers of banks, each with a different role.

Large city banks served major industrial groups. Regional banks focused on particular prefectures. Long-term credit banks raised funds through debentures and provided longer-term financing for equipment and investment. Shinkin banks and credit cooperatives collected smaller deposits and lent within local communities.

The distinction mattered.

A steel company in Osaka did not need the same type of bank as a small manufacturer in a regional town. Japan built institutions around those different needs rather than expecting one type of bank to serve everyone.

The government also managed the pace of credit expansion.

From the early 1950s until it was abolished in 1991, the Bank of Japan used what became known as window guidance. The central bank gave city banks quarterly targets for loan growth. These were not laws forcing every bank to lend a fixed amount. But banks had a strong reason to follow the guidance because the central bank controlled access to its credit while the Ministry of Finance controlled permission for new branches.

In the early decades, city-bank loan growth generally stayed within about one percentage point of the target.

The third change was perhaps the most important: government backed development finance and ordinary commercial lending were given different jobs.

Large policy projects could be handled by public institutions such as the Japan Development Bank. Commercial banks could then concentrate much of their lending on companies expanding production steel, automobiles, electronics, machinery and other industries.

Many companies also developed long term relationships with a main bank. The bank did more than issue a loan and wait for repayment. It monitored the company, participated in lending syndicates, sometimes held shares, and could help restructure a struggling borrower.

That relationship became particularly important when a company had a bad year.

The bank did not necessarily walk away at the first sign of trouble.

Japan: what happened after the system was built

By 1971, Japan already had a remarkably dense banking network.

There were 14 city banks, 61 regional banks, 71 second tier regional banks, seven trust banks and three long-term credit banks 156 banks in total alongside 483 shinkin banks and 524 credit cooperatives.

That meant more than a thousand insured deposit taking institutions serving a population and economy that were still in the process of catching up with the West.

The branch network expanded with them.

Between 1953 and 1970, city bank branches increased from 1,821 to 2,409. Regional bank branches increased from 3,602 to 4,335. Shinkin branches grew from 2,045 to 3,871.

Across all banks, the number of branches increased from 9,643 to 13,755.

The interesting part is that the local network grew faster than the city bank network.

That mattered because finance was physically close to the businesses that needed it.

The economic numbers were extraordinary.

Between 1957 and 1973, Japan's GNP grew at roughly 10 percent a year. Between 1953 and 1973, real GDP grew by about 9.1 percent annually, real GDP per person by 8 percent, and real investment by 14.9 percent.

Export volumes grew by about 13.5 percent a year.

The government had set an ambitious target with the Income Doubling Plan of 1960, aiming for annual growth of 7.2 percent so that the economy could double in a decade.

Actual growth exceeded the target.

The change in employment tells an even clearer story.

In 1953, agriculture, forestry and fishing employed 39.9 percent of workers. By 1973, that share had fallen to 13.5 percent.

Industry moved in the opposite direction, rising from 24.3 percent to 36.6 percent of employment.

Total employment grew by about 1.5 percent a year. Unemployment fell from 1.8 percent to 1.3 percent, while labour productivity increased by roughly 7.6 percent annually.

Japan was moving people out of low productivity agricultural work and into increasingly productive industrial jobs.

The catch-up with the United States was just as striking.

In 1953, Japan's GDP was roughly 13 percent of the US level, while income per person was about 23 percent of the American level.

By 1973, those figures had risen to approximately 35 percent and 68.5 percent respectively.

Japan joined the OECD in 1964.

The banking system did not create this transformation by itself. But it provided something an industrializing economy could not do without: a mechanism for turning savings into machinery, factories, exports, and jobs.

Japan also shows what can go wrong

There is an important warning here.

Japan's post-war banking model worked extremely well during the high-growth period. But some of the same relationships that supported the miracle later became a problem.

After the asset bubble burst, banks continued rolling over loans to weak borrowers instead of forcing losses to be recognized. Window guidance had already been abolished in 1991, and the old system eventually became part of a much larger bad-loan crisis.

The lesson is not that patient lending is bad.

It is that patient capital still needs discipline.

A bank that understands its borrowers and gives them time to recover can support productive businesses through difficult periods. A bank that continually rolls over loans that will never be repaid simply hides losses.

Tanzania should learn from both sides of the Japanese experience.

China: turning one bank into a banking system

China's starting point was even more concentrated.

In 1978, China essentially operated a monobank system.

The People's Bank of China issued currency, held reserves, and chanelled funds to state-owned enterprises according to the economic plan.

A factory generally did not approach a bank and negotiate a commercial loan.

It received funding through the state budget.

The reforms that followed changed that system step by step.

Rebuilding specialized banks

In 1979, the Agricultural Bank of China was restored to serve agriculture and rural industry. The Bank of China took responsibility for foreign-exchange business, while the Construction Bank focused on investment finance.

These institutions had existed in earlier forms before being absorbed into the monobank.

Now they were separated again.

Between 1983 and 1984, the People's Bank was turned into a central bank and its urban commercial operations were transferred to a new institution, the Industrial and Commercial Bank of China.

China had effectively moved from one institution doing almost everything to a two-tier banking system: one institution responsible for monetary policy and other banks responsible for lending.

That distinction sounds basic today.

At the time, it was a major change.

From government grants to bank loans

China also changed how enterprises were financed.

Beginning in 1985, funding for enterprises increasingly moved from the government budget to the banking system.

State grants, known as bokuan, were gradually replaced by loans, daikuan.

The share of funds reaching industrial enterprises as budgetary grants fell from roughly 70 percent in 1978 to about 20 percent by 1982.

Banks increasingly supplied the difference.

In December 1986, Deng Xiaoping captured the problem bluntly when he argued that Chinese banks needed to behave like actual banks rather than simply issuing currency and holding reserves.

Another important change happened at the local level.

Early in the reform period, a county branch collecting more deposits did not automatically have the freedom to lend those funds locally. As the system developed, that separation weakened. Agricultural Bank branches increasingly lent from deposits collected within their communities.

Rural credit cooperatives were also encouraged to serve township enterprises and households rather than simply passing money through the state plan.

The financial system was beginning to see the local economy as a source of both deposits and borrowers.u

China separates policy finance from commercial lending

China eventually recognized another problem: development policy and ordinary commercial lending could not remain mixed indefinitely.

In 1994, the government created three policy banks:

  • China Development Bank

  • Export-Import Bank of China

  • Agricultural Development Bank of China

The purpose was straightforward.

Policy-driven projects could be financed through dedicated institutions rather than sitting on the same balance sheets as ordinary commercial loans.

A commercial banking law followed in 1995.

This separation is particularly relevant to Tanzania.

If the government wants banks to finance a policy objective, it should be clear whether that is ordinary commercial lending or government-directed development finance. Mixing the two can make it difficult to tell whether a bank is allocating capital because a borrower is creditworthy or because the government wants the project funded.

China: what happened after credit reached the local economy

By the end of 1993, China's banking system looked nothing like the monobank of 1978.

It had four specialized state banks, nine nationwide or regional commercial banks, around 59,000 rural credit cooperatives and about 3,900 urban credit cooperatives, alongside trust, leasing and securities institutions.

Credit expanded rapidly.

Between 1979 and 1985, deposits nearly tripled and the value of bank loans increased by about 260 percent.

Household bank deposits rose from roughly 7 percent of national income to 38 percent between the late 1970s and late 1980s.

One of the most revealing numbers came from township and village enterprises (TVEs).

Loans to these businesses increased about twelvefold between 1980 and 1994, faster than the economy as a whole. The share of Agricultural Bank and rural-cooperative loans going to these enterprises rose from 13.8 percent to 33.2 percent.

Then came the jobs.

TVE employment rose from approximately 23 million people in 1977 to 52 million in 1984, and reached 106 million by 1992.

By 2010, the sector employed around 153 million people, roughly one-fifth of the Chinese workforce.

Rural industrial output grew by about 1,188 percent during the 1980s, compared with 228 percent for industry as a whole. The share of industrial output produced by rural industry increased from 13 percent in 1980 to 40 percent in 1990.

The national economy was changing just as quickly.

Between 1984 and 1992, China's GDP grew at roughly 9.6 percent a year. The non-state sector grew even faster, at approximately 17.3 percent annually, and accounted for much of the acceleration.

In 1978, about 70.5 percent of Chinese workers were still in agriculture, and 82 percent of the population lived in rural areas.

As township industries expanded, workers began leaving agriculture for factory and wage employment.

The banking reforms did not cause China's economic transformation by themselves. Agricultural reform and China's opening to international trade were fundamental.

But credit mattered because it allowed higher rural incomes and local savings to become something tangible: workshops, factories, machinery, inventories, and wages.

China also paid a price for soft lending

The Chinese experience has its own warning.

Specialized banks were initially still expected to make politically directed loans. Some of those loans were never commercially viable, and bad assets accumulated.

By the late 1990's, the government had to restructure the banks and remove large quantities of non-performing loans.

That experience helps explain why China eventually separated policy banks from commercial banks.

The lesson is straightforward:

If the government wants a loan made for policy reasons, the risk should be visible and accounted for somewhere. It should not quietly become a commercial bank's problem.

Tanzania is still near the beginning of this journey

Tanzania has around 35 commercial banks and two community banks.

A substantial share of bank assets is invested in government securities. Private-sector credit is growing, but personal lending and trade account for a significant portion of that expansion, while many smaller businesses still struggle to qualify for formal loans.

The comparison with Japan and China should not be taken to mean Tanzania can reproduce either country's growth simply by creating more banks.

It cannot.

The starting conditions are different. The economies are different. The institutions are different.

But Tanzania can ask the same structural question:

Who is the banking system designed to see?

If a bank officer can easily assess a government bond but cannot confidently assess a small manufacturing business in Mwanza, the financial system has an information problem.

If a business owner has machinery, inventory and reliable customers but cannot borrow because none of those assets carry a conventional title, the system has a collateral problem.

And if starting a regional bank requires so much capital and regulatory infrastructure that only very large investors can enter, the system has an entry problem.

Those are policy problems. They can be changed.

What Tanzania could change

The useful reforms are not slogans about "financial inclusion." They are specific changes to how the system works.

Make room for regional banks

Tanzania already has legal room to apply different capital requirements to institutions serving underserved communities.

That opportunity could be used to establish a genuine regional-bank category with staged capital requirements below the TZS 15 billion commercial-bank threshold, defined geographic limits, and clear requirements for building capital as the institution grows.

Japan deliberately protected the regional role of its banks.

China built tens of thousands of rural cooperatives because one national banking institution could not understand the financing needs of every county and township.

Tanzania does not need tens of thousands of banks.

It needs enough local institutions to make local businesses visible to the financial system.

Separate policy lending from ordinary banking

If the government wants banks to finance a particular development objective, the risk should be explicit.

A development window, guarantee fund, or dedicated institution can carry policy-driven lending, as China's policy banks eventually did.

Commercial banks can then judge ordinary borrowers on commercial terms.

That makes both sides more accountable.

Make government securities less dominant

Banks hold Treasury securities for understandable reasons.

They are liquid, relatively low risk, and useful for meeting regulatory requirements. They can also be used as collateral at the central bank.

The answer is not to make government bonds unattractive.

It is to make well-managed lending to productive businesses sufficiently viable that banks have a genuine choice.

Otherwise, the rational bank will continue buying government paper.

Fix the collateral problem

Tanzania needs a functioning movable-assets registry that allows businesses to use machinery, inventory, receivables and other productive assets as collateral.

The system should also make better use of information already being generated through mobile money, tax records, utility payments and other formal transactions.

A business should not become invisible to a bank simply because it does not own titled urban land.

Build better credit information

Japan's regional banks had something valuable that a distant lender often lacks: knowledge of the businesses in their communities. China eventually built a similar advantage through local branches and cooperatives. Tanzania can combine local banking with modern data.

A lender should be able to look at a small manufacturer's transaction history, tax records, mobile-money flows, inventory, existing obligations and repayment behaviour and make a decision based on the business as it actually operates. That is far more useful than asking only whether the owner has a title deed.

Keep discipline in the system

This may be the most important lesson from both countries. Do not confuse access to credit with good lending. Japan's later bad-loan crisis showed what happens when banks keep weak borrowers alive indefinitely. China's restructuring showed the cost of political lending that is allowed to accumulate without clear accountability.

A new Tanzanian regional bank should not become another buyer of government securities with a different name. Nor should it become a machine for rolling over loans that borrowers cannot repay. The objective is productive credit.

The real test

For an ordinary Tanzanian, the question is ultimately simple. Did the loan help someone buy a machine?

Did it finance a farmer's inputs for the next season? Did it allow a workshop to take on ten more workers?

Did it help a manufacturer increase production? Did it finance inventory that would become a sale rather than simply another consumer purchase?

Those are the kinds of uses of credit that can change an economy's productive capacity. Japan moved millions of workers from agriculture into industry during its high-growth period. China did something similar through township enterprises and rural industry. Tanzania does not need to copy either country.

But it can learn from what both discovered: economic transformation requires a financial system capable of finding productive borrowers wherever they are. Tanzania's banks are already profitable. The country already has savings, pension assets, mobile-money activity, and a growing private sector.

The missing question is increasingly difficult to ignore:

Can we redesign the banking system so that more of that money reaches the people building the next generation of Tanzanian businesses?

That is the reform worth pursuing.

Sources