## _cr04251

## Source details

**Canonical URL:** [_cr04251](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04251.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04251.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04251.pdf.json)

---

### Interest rate weight in the MCI vs. openness of the economy (1996–2003): context and key conclusions
- Widespread euroization and openness restrict scope for autonomous monetary policy.
- Authorities focus on maintaining a broadly stable kuna-euro exchange rate; combined with a relatively open capital account, this limits scope for autonomous monetary management.
- Monetary policy could still play a role because:
  - domestic and international capital markets are not perfectly integrated, and
  - the exchange rate is managed rather than fixed (actual daily fluctuations were in the range of ±4.5 percent in 1999–2003 and ±7.5 percent in 1996–2003).
- Empirical evidence (short series, structural breaks) supports limited effectiveness of monetary policy for aggregate demand management in Croatia:
  - financial conditions only weakly correlated with monetary policy stance;
  - monetary policy can influence money market interest rates but influence on lending rates is uncertain and lagged, and the link has weakened since 2001;
  - credit controls in most of 2003 illustrate this ineffectiveness.

### MCI and FCI definitions, calibration, and Croatia result
- Definitions and construction:
  - MCI = weighted average of a kuna-euro nominal exchange rate index and a money market interest rate index (both normalized: 2002 average = 100).
  - FCI = weighted average of a trade-weighted real effective exchange rate (REER, CPI-based) and a real effective lending rate (RELR). Both REER and RELR seasonally adjusted (X-12 multiplicative) and normalized to 2002 average = 100. RELR computed as weighted average of interest rates on domestic and foreign borrowing (weight = share of external debt in total private sector debt), adjusted by CPI.
  - Key calibration parameter: relative weight of interest rate vs. exchange rate in MCI (the MCI ratio). Open economies receive relatively lower interest-rate weight.
- Croatia calibration:
  - Degree of openness (average of exports and imports as percent of GDP) = about 52 percent.
  - Regression (Figure 1): y = -0.01x + 0.97, R2 = 0.88.
  - Implied interest rate weight for Croatia = 0.57 (MCI ratio of 1.3:1).
  - Interpretation: effect on demand of a 1.3 percentage point exchange rate appreciation can be offset by a 1 percent decline in interest rates.

### Empirical behavior of MCI and FCI (1997–2003)
- Constructed MCI strongly correlated with money market interest rate; kuna-euro exchange rate exhibited much lower volatility.
- Trend (HP filter): monetary policy loosened mid-2000 to mid-2002 and tightened since mid-2002.
- FCI dynamics:
  - Loosened during 1997 to early 2001.
  - Tightened from mid-2001 to mid-2002.
  - Since mid-2002 relatively stable: mild loosening in second half 2002, mild tightening in first half 2003, loosening at end-2003.
- Comparison MCI vs. FCI:
  - MCI (and money market rate) exhibited much higher volatility than FCI.
  - Correlation MCI–FCI = +0.38 for 1997–2003; slightly negative for the last three years of the sample.

### Formal tests of transmission (MCI → FCI)
- Bivariate Granger causality tests of d(MCI) → d(FCI):
  - Period 1997:1–2003:12:
    - 3 lags = 1.785 (p = 0.079)*
    - 6 lags = 1.516 (p = 0.187)
    - 9 lags = 2.154 (p = 0.040)**
  - Period 2000:1–2003:12:
    - 3 lags = 0.198 (p = 0.897)
    - 6 lags = 0.711 (p = 0.643)
    - 9 lags = 0.970 (p = 0.489)
  - Notes: */** denotes significance at 10/5/1 percent levels; “d” stands for difference. Granger tests from FCI to MCI were insignificant at the 10 percent level for reported lags.
- Impulse-response analysis (VAR, 2000–2003): weak FCI response to innovations in MCI (Cholesky one s.d. innovations ± 2 s.e.), consistent with Granger results indicating weak and generally insignificant transmission since 2000.

### Synthesis and policy-relevant observations (MCI findings)
- MCI calibration (interest weight = 0.57, MCI ratio = 1.3:1) and empirical results indicate a limited role for conventional monetary policy in shaping overall financial conditions under observed regime and openness.
- Monetary policy clearly influences money market rates but has weak and uncertain influence on broader financial conditions and lending rates, with long lags.
- Weakened transmission since 2000 and limited co-movement between MCI and FCI imply exchange-rate management and other financial sector factors can dominate transmission.
- Historical episodes (including credit controls in 2003) illustrate practical limitations of monetary instruments for aggregate demand management in Croatia.

*Source: The authors’ calculations, based on data from the CNB.*

### Robustness of MCI and FCI to weight changes
- MCI and FCI are rough indicators due to model dependency, ignored dynamics, parameter inconstancy, and nonexogeneity of regressors.
- Sensitivity to weights:
  - Varying interest-rate weight from 0.57 to 0.50 or 0.40 produces indices with pair-wise correlations of about 0.99.
  - Conclusion: a somewhat higher weight on the exchange rate (e.g., if Croatia is more euroized) would not substantially change conclusions about monetary conditions and policy stance.

*Source: The authors’ calculations, based on data from the CNB.*

### Interest rate channel of monetary transmission: money market rates, excess liquidity, and lending rates
- Money market rates and excess liquidity (daily data, Jan 2001–Feb 2004):
  - Excess liquidity (daily deviation from 30-day MA) is a leading indicator for money market rates; negatively linked to money market rates.
  - Correlation coefficients (excess liquidity vs. ZIBOR):
    - Excess liquidity = daily − 30-day MA:
      - Overnight: -0.20
      - 1W: -0.20
      - 2W: -0.18
      - 1M: -0.15
      - 3M: -0.12
      - 6M: -0.23
    - Excess liquidity = 30-day MA − 12-month MA:
      - Overnight: -0.16
      - 1W: -0.20
      - 2W: -0.24
      - 1M: -0.33
      - 3M: -0.35
      - 6M: -0.34
  - Note: “W” = week, “M” = month, “MA” = moving average. Excess liquidity defined as departure from trend of currency in circulation plus government deposits.
- Transmission from money market rates to bank lending rates:
  - Empirical finding: transmission weak.
  - 2003 episode: money market rates rose after kuna required reserves increased from 25 to 42 percent, but effective lending rate of domestic banks remained largely unaffected; spread vs. foreign rates unchanged.
  - Granger causality tests (money market rate → lending rate, sample/lags):
    - 1997:1–2003:12, 3 lags: 2.411 (0.074)*; 6 lags: 1.907 (0.093)*; 9 lags: 2.831 (0.008)***
    - d(money market rate) → d(lending rate), 1997:1–2003:12, 9 lags: 2.189 (0.037)**
    - 2001:1–2003:12, money market rate → lending rate: 3 lags: 0.219 (0.882); 6 lags: 0.847 (0.545); 9 lags: 1.038 (0.440)
    - Notes: significance denoted */**/*** for 10/5/1 percent; “d” = difference.
  - Interpretation:
    - Link present in 1997–2003 but disappears when restricting to 2001–2003 (post liberalization).
  - VAR results:
    - For Euro Area (Angeloni and Ehrmann, 2003): maximum impact on lending rates in ~5 months, maximum impact ~0.8 of original shock.
    - For Croatia: VAR suggests in 5 months the impact of money market rates on lending rates ≈ 0.15 of original shock.
    - Re-estimated since 2001: response of domestic lending rates virtually zero.
    - One st. dev. innovation ≈ 3.7 percentage point increase in money market rates (note in Figure 7).

*Source: The authors’ calculations, based on data from the CNB.*

### Short-term capital flows and interest rate differentials
- Correlation between interest rate differentials (domestic lending rate − euro-area lending rate) and short-term capital flows:
  - Sample 1997:1–2003:4: 0.11
  - Sample 1997:1–2001:2: 0.03
  - Sample 2001:3–2003:4: 0.24
- Interpretation:
  - Near nil before removal of restrictions in 2001; about 1/4 since then (insignificant at 10 percent).
  - Suggests interest rate channel may have weakened further; more data needed.

*Source: The authors’ calculations, based on data from the CNB.*

### Other monetary transmission channels
- Possible channels besides interest rate and exchange rate:
  - Equity price channel, credit channel, balance sheet channel (external finance premium depends on borrowers’ financial position).
- Relevance for Croatia:
  - Equity price channel likely limited (small equity market).
  - Real estate market more important, but so far driven largely by non-monetary factors.
  - Credit channel could be substantial given banks’ dominant role, but only if banks cannot find funding abroad.
  - Preliminary bank-by-bank calculations suggest banks—especially foreign-owned—can find funding abroad; thus the credit channel does not seem very substantial; further research warranted.

*Source: The authors’ calculations, based on data from the CNB.*

### Experience with credit controls (2003) and effects on credit and financial soundness
- CNB introduced credit controls in January 2003; abolished end-2003.
- Effects:
  - Bank credit decelerated in 2003; households less able to access foreign borrowing.
  - Enterprises switched borrowing to foreign banks and used leasing/other financing.
  - External borrowing in 2003 ≈ 2.5 times higher than in 2002; share of external debt in corporate investment financing rose in 2003.
  - After abolition (early 2004), credit growth did not bounce back:
    - Seasonally adjusted credit growth first five months of 2004 (after adjusting for exchange rate movements) indicates an annualized rate of 14½ percent, about the same as in 2003 — suggesting deceleration since spring 2003 was largely demand-driven.
- Selected entries (Croatia: Credit Growth Before and After the Controls, 2000–03):
  - Total lending (% change y/y)1/: 2000: 5.7; 2001: 20.8; 2002: 24.6; 2003: 20.3
  - Domestic bank borrowing (contribution in % points): 2000: 5.7; 2001: 15.9; 2002: 20.8; 2003: 10.7
  - Foreign borrowing (without leasing): 2000: 5.6; 2001: -3.8; 2002: 2.3; 2003: 5.1
  - Leasing: 2000: 0.2; 2001: 2.6; 2002: 1.9; 2003: 3.8
  - Adjustment for write-offs: 2000: -5.9; 2001: 6.1; 2002: -0.5; 2003: 0.6
  - Corporate sector lending (% change y/y): 2000: 3.9; 2001: 14.0; 2002: 17.4; 2003: 16.2
    - Domestic bank borrowing contribution: 2000: 1.0; 2001: 10.7; 2002: 11.9; 2003: 2.4
    - Foreign borrowing (without leasing): 2000: 7.3; 2001: -5.1; 2002: 3.2; 2003: 7.7
  - Household sector lending (% change y/y): 2000: 11.4; 2001: 41.7; 2002: 42.1; 2003: 28.5
    - Domestic bank borrowing contribution: 2000: 21.0; 2001: 31.8; 2002: 42.5; 2003: 27.6
    - Foreign borrowing (without leasing): 2000: 0.2; 2001: 0.1; 2002: 0.2; 2003: 0.0
  - Memorandum: Composition of corporate financing (in %), Domestic borrowing (net flow): 2000: 2.8; 2001: 25.8; 2002: 24.4; 2003: 4.8
  - Memorandum: Foreign borrowing (net flow): 2000: 20.2; 2001: -4.1; 2002: 12.1; 2003: 27.1
  - Memorandum: Other (e.g., reinvested profits): 2000: 77.0; 2001: 78.3; 2002: 63.5; 2003: 68.1
  - Note: 1/ Total domestic and external borrowing by the non-government sector in Croatia.
- Financial soundness impacts:
  - Best corporate clients redirected to foreign banks.
  - Rapid growth of unsupervised leasing companies; their contribution to overall lending growth increased in 2003.
  - Transparency of monetary and banking statistics deteriorated as banks engaged in circumvention activities (asset swaps, collateralization, accelerated write-offs).
  - Table 5 calculations attempt to approximate impact of write-offs on total credit data.

*Source: Authors' calculations based on data from CNB and the Central Statistics Bureau.*

### Transition: determinants of lending rates and spreads (Chapter II introduction)
- Rapid increase in bank credit driven by perceived profit opportunities (high lending interest rates and relatively low default rates).
- Interest rate spreads:
  - Declining but still somewhat higher than other CEEB countries (about 2–4 percentage points above CEEB average historically).
  - Within-system variability (September 2003):
    - Local currency deposit rates: 0.4 to 5.9 percent.
    - Lending rates: 6.1 to 26.8 percent.
    - Spreads: 4.7 to 21.0 percent.

### Decomposition of interest rate spreads (accounting identity)
- Decomposition identity preserved exactly:
  iL – iD ≡ o + l + iD r/(1 – r) + d + p + τ [p – o  − l  − iD r/(1 – r)– d] ,      (1)
  where:
  - iL = average interest rate on loans
  - iD = average interest rate on deposits
  - o = overhead costs relative to loans
  - l = loan loss provisions net of recoveries relative to loans
  - r = reserve requirement rate
  - d = deposit protection premium relative to loans
  - p = profit margin (pre-tax profit relative to loans)
  - τ = tax rate on profits
- Profit margins calculated as residual after other components.

### Results from accounting decomposition (aggregate sector, 2001–2003)
- Selected figures (percentage points):
  - Interest rate spread (All loans): 2001 = 9.6 ; 2003 = 10.1
  - Overhead costs (All loans): 2001 = 3.5 ; 2003 = 2.5
  - Loan loss provisioning (All loans): 2001 = 0.1 ; 2003 = 0.6
  - Reserve requirement (All loans): 2001 = 0.0 ; 2003 = 0.3
  - Deposit protection (All loans): 2001 = 0.5 ; 2003 = 0.4
  - Pre-tax profit (All loans): 2001 = 5.6 ; 2003 = 6.6
  - Tax (All loans): 2001 = 0.2 ; 2003 = 0.2
  - Profit margin (All loans): 2001 = 5.4 ; 2003 = 6.5
  - Memorandum item: ROA (after-tax) = 1.3
- Overhead costs:
  - ≈ 2.5 percentage points of spread in 2003; more than one-third of spread.
  - Above advanced-economy average ≈ 2 percentage points.
  - Overhead costs to loans declined ≈ 1 percentage point between 2001 and 2003.
- Loan loss provisioning: 0.4 percentage points in 2003, reflecting relatively low NPLs.
- Reserve requirements: impact ≈ 0.2 percentage points in 2003; reserves remunerated at 25 basis points below EURIBOR.
- Deposit protection premia: ≈ 0.4 percentage points in 2003; relatively high.
- Profit margins:
  - Pre-tax profit on enterprise lending in 2003 substantially lower than average return on banks’ assets.
  - Pre-tax return on household lending > 4 times average pre-tax return on assets.
- Foreign greenfield banks: lower spreads and profit margins; specialization on lower risk corporate clients.

### Bank productivity (2002, thousands of U.S. dollar)
- Assets per Employee:
  - Croatia: 1,413
  - Emerging markets: 2,040
- Loans per Employee:
  - Croatia: 762
  - Emerging markets: 910
- Deposits per Employee:
  - Croatia: 1,002
  - Emerging markets: 1,620
- Net Interest Income per Employee:
  - Croatia: 24
  - Emerging markets: 60

### Bank-by-bank regressions: methodology and main empirical findings
- Data/sample: panel monthly data for 46 Croatian banks, July 1999–December 2003 (≈2,100 observations, 54 cross-sections). Estimators: SUR and GLS (cross-section heteroscedasticity).
- Explanatory power: estimates account for ≈ 90 percent of bank-by-bank variability in lending spreads.
- Main findings (preserve numeric magnitudes where given):
  - Bank size: doubling balance sheet → lending rates typically lower by about 0.3–0.6 percentage points (using log(2) ≈ 0.69 and coefficient scaling).
  - Market share: evidence that lending rates/spreads increase with market share (possible market power).
  - Asset quality (NPLs): higher NPL ratio → higher lending rates and spreads.
  - Liquidity: more liquid banks → lower lending rates/spreads.
  - Capital adequacy: higher capital adequacy → lower lending rates but even lower deposit rates, resulting in higher spreads versus lower-capital banks.
  - Ownership: foreign greenfield banks → generally lower interest rate spreads.
  - Credit growth determinants:
    - Lower for banks with higher lending rates.
    - Lower for larger banks and banks with higher NPL ratios.
    - Significant relationships with bank liquidity and capital adequacy.
    - Greenfield banks growing faster controlling for other factors.
    - Higher treasury-bill rates related to lower credit growth.
  - Interest rate transmission: changes in lending rates positively related to money market and deposit rates with lags; significant AR(1) coefficient reflects persistence.
- Selected regression statistics (Table 3 highlights):
  - No. of observations: 2,140 (specifications vary: 2,140 / 2,140 / 2,109 / 2,109 / 2,144 / 2,144 / 2,144).
  - R2 (weighted): reported as 0.91, 0.94, 0.86, 0.99; R2 unweighted varies across specs.
  - Selected coefficients (significance indicated):
    - Log(assets): -0.84**, -0.86**, -0.68**
    - NPLs in total loans: 5.78**, 2.76**, 6.17**, 5.39**, 3.97**, 4.21**, -5.97**
    - Greenfield bank: -3.02**, -2.18**, -0.60**, -0.53**, -3.38**, -2.31, 1.53**
    - AR(1): 0.85**, 0.82**, 0.80**, 0.55**, 0.82**, 0.79**, 0.11**
    - Domestic T-bill rate: 0.11**, 0.08**, 0.07**, 0.12**, 0.12**, -0.12**
  - Significance notation in Table 3: * = 1 percent, ** = 5 percent, + = 10 percent.

### Policy implications from bank-level analysis
- Credit information and supervision:
  - High NPLs → higher lending rates; need improved credit information sharing and stronger prudential supervision.
  - Low liquidity and low capital adequacy associated with higher lending rates; reinforces prudential recommendations.
- Competition policy:
  - Significant economies of scale imply bank growth (including mergers and acquisitions) could reduce spreads via lower overhead costs.
  - Permissive M&A policy should be accompanied by strict enforcement against anti-competitive behavior (abuse of dominant position).
- Reserve requirements:
  - Reserve requirement rate = 19 percent.
  - Lowering reserve requirements would have limited impact on lending rates (~0.2 percentage points) because reserves are remunerated (remuneration = 25 basis points below EURIBOR).

*Source: Prepared by Martin Čihák, with inputs from Tomislav Galac and Danijela Mladinovič (as presented in the supplied content).*

### Employment protection, labor market performance, and the shadow economy
- Stylized facts (labor market):
  - Stringent employment protection increases long-term unemployment and reduces labor turnover; problem amplified by transition and war shocks.
  - Labor force survey-based unemployment ≈ 15 percent for past five years.
  - Share of long-term unemployment ≈ 55 percent (compared with CEECs ≈ 45 percent).
  - Labor force participation ≈ 50 percent.
  - Job reallocation: firms terminate ≈ 5 percent of jobs yearly vs. 10–11 percent in other CEECs.
  - Job creation ≈ 3½ percent vs. 7–11 percent in other CEECs.
- Labor costs, wages, and unemployment benefits:
  - Gross wages in manufacturing higher than most CEECs (World Bank, 2003); unit labor costs relatively strong among CEECs.
  - Unemployment benefit: flat rate max ≈ one-fourth of average wage; replacement ratio relatively low; duration capped at 312 days; coverage < 20 percent since mid-1990s.
  - Conclusion: disincentives from benefits likely modest.
- Employment Protection Legislation (EPL):
  - Croatia among strictest in CEECs; stricter than most EU-15 and other CEECs.
  - Individual dismissals costly (long notice, high severance); collective dismissals difficult; fixed-term employment restricted until recently.
  - Share of employment by SMEs = 46 percent in Croatia vs. well over 50 percent in major CEECs.
  - July 2003 labor law amendment (effective beginning of 2004) lowered Croatia’s EPL index by 23 percent via:
    - Relaxing fixed-term contract restrictions;
    - Easing pre-conditions for valid dismissals;
    - Shortening advanced notice from 6 to 3 months;
    - Reducing severance from one-half to one-third of monthly pay;
    - Relaxing definition of mass lay-offs.
- Policy implication: relaxing EPL can improve flexibility, increase employment in official sector, and promote SME-driven job creation.
- Employment protection and shadow economy:
  - Conceptual links: stringent EPL raises official labor costs → firms/workers have incentive to operate in informal sector.
  - Cross-country correlation: positive relationship between EPL strictness and shadow economy size (percent of GDP) across sample of 27 countries.
- OLS regression on shadow economy (Dependent: Shadow Economy % of GDP; Observations = 27; R-Bar-Squared = 0.607):
  - Estimates:
    - Constant: Coef. 56.814; S.E. 16.338; T-Ratio 3.477; Prob. [0.002] (***)
    - Log GDP per capita: Coef. -6.017; S.E. 1.443; T-Ratio -4.171; Prob. [0.000] (***)
    - EPL: Coef. 4.234; S.E. 1.387; T-Ratio 3.052; Prob. [0.006] (***)
    - Tax wedge on labor income: Coef. 0.046; S.E. 0.099; T-Ratio 0.466; Prob. [0.646]
    - Business regulations: Coef. 1.477; S.E. 1.294; T-Ratio 1.142; Prob. [0.266]
  - Interpretation:
    - Higher per capita income → smaller shadow economy (significant).
    - EPL coefficient positive and highly significant: stricter employment protection associated with larger shadow economy, ceteris paribus.
    - Tax wedge and business regulation index not statistically significant in this specification.
- Policy implication: relaxing employment protection can expand the tax base and boost productivity by reducing incentives for the shadow economy; July 2003 labor law amendments are an important step.

*https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04251.pdf*

### Selected statistical appendix highlights (exact figures preserved)
- Real GDP (percentage changes): 2000 2.9; 2001 4.4; 2002 5.2; 2003 4.3; 2004 3.7; 2005 Proj. 4.1
- Gross fixed capital formation (% changes): 2000 -3.8; 2001 7.1; 2002 12.0; 2003 16.8; 2004 5.8; 2005 Proj. 5.4
- Total exports (millions U.S.$): 1996 4,512; 1997 4,171; 1998 4,541; 1999 4,302; 2000 4,432; 2001 4,666; 2002 4,904; 2003 6,164
- Total imports (millions U.S.$): 1996 7,788; 1997 9,104; 1998 8,383; 1999 7,799; 2000 7,887; 2001 9,147; 2002 10,722; 2003 14,199
- Total external debt (millions U.S.$): 1996 5,308; 1997 7,452; 1998 9,683; 1999 9,978; 2000 11,055; 2001 11,317; 2002 15,421; 2003 23,570; 2004 Mar. 24,199
- Consolidated general government balance (percent of GDP, GFS 1986 basis): 1998 -3.5; 1999 -8.2; 2000 -6.5; 2001 -6.7; 2002 -5.2; 2003 -6.3
- Central bank reserves (selected, 1996–2003 Mar., in millions): Reserves: 1996 4,410; 1997 5,046; 1998 5,908; 1999 8,988; 2000 10,589; 2001 15,003; 2002 20,373; 2003 Mar. 30,014

### Monetary and banking system tables (selected balance sheet and interest rate excerpts)
- Total assets (levels, millions of kuna) series:
  - 67,483; 85,309; 93,326; 93,251; 109,949; 142,606; 165,622; 172,475; 175,799; 184,739; 195,278; 199,479
- Claims on other domestic sectors (levels, millions of kuna):
  - Total claims on other domestic sectors: 33,690; 48,592; 59,597; 55,400; 60,364; 74,284; 96,218; 101,018; 102,609; 105,689; 110,374; 111,940
  - Claims on enterprises: 26,929; 35,487; 41,225; 35,244; 35,891; 42,882; 51,723; 53,023; 52,021; 52,172; 53,810; 54,823
  - Claims on households: 6,615; 12,796; 17,717; 19,250; 23,298; 30,122; 43,073; 46,687; 49,309; 52,243; 55,001; 55,537
- Deposit and liability categories (levels, millions of kuna; selected series):
  - Demand deposits: 7,007; 8,424; 7,809; 7,891; 11,386; 15,181; 21,166; 19,973; 22,188; 22,071; 23,315; 21,559
  - Foreign currency deposits: 21,817; 31,278; 37,971; 36,966; 46,902; 71,837; 72,055; 74,069; 71,104; 75,416; 76,035; 74,070
  - Capital accounts: 15,441; 17,027; 19,786; 21,975; 24,953; 25,455; 26,323; 26,526; 26,120; 26,809; 27,390; 27,493
- Restricted and blocked deposits (levels, millions of kuna):
  - 8,224; 5,852; 4,196; 3,434; 2,550; 1,601; 1,680; 1,786; 1,858; 1,939; 1,709; 2,037
  - Households' blocked f/c deposits: 7,171; 4,574; 3,419; 2,743; 1,695; 770; 319; 258; 242; 177; 168; 111
- Interest Rates (selected monthly entries, percent, annualized, weighted averages):
  - 1996 Dec: CNB Bills (35 days) 8.01; Credits in Kuna 8.46; Deposits in Kuna 18.97; Credits in Kuna Indexed to f/c 19.50; Credits in f/c 4.15; Deposits in Kuna Indexed to f/c 9.46; Deposits in f/c 5.09
  - 1999 Dec: 10.51 3.54 12.53 6.75 4.27 6.62 4.23
  - 2000 Dec: 6.71 0.45 10.74 7.70 3.40 5.54 3.47
  - 2001 Dec: 3.4 9.51 9.29 5.94 2.76 4.58 2.60
  - 2002 Dec: 2.1 10.91 8.25 5.91 1.55 2.92 2.55
  - 2003 Dec: - 11.45 7.07 5.62 1.66 3.48 2.22
  - 2004 Mar: - 11.76 7.48 4.58 1.98 3.70 2.38

*Source: _cr04251 (IMF PDF content provided).*

### 1. Interest Rate Weight in the MCI vs. Openness of the Economy, 1996–2003.....................  6

### 1. Interest Rate Weight in the MCI vs. Openness of the Economy, 1996–2003

### A. Context and key conclusions
- The widespread euroization and openness of the Croatian economy restrict the scope for autonomous monetary policy.  
- The authorities focus on maintaining a broadly stable kuna-euro exchange rate; combined with a relatively open capital account, this limits scope for autonomous monetary management.  
- Monetary policy could still play a role if domestic and international capital markets are not perfectly integrated and because the exchange rate is managed rather than fixed (actual daily fluctuations were in the range of ±4.5 percent in 1999–2003 and ±7.5 percent in 1996–2003).  
- Empirical studies are few and inconclusive because of short time series and structural breaks. Recent evidence analyzed here supports that monetary policy is not an effective tool for aggregate demand management in Croatia: financial conditions are only weakly correlated with the monetary policy stance, monetary policy can influence money market interest rates but its influence on lending rates is uncertain and lagged, and the link has weakened since 2001. The experience with credit controls in most of 2003 illustrates this ineffectiveness.

### B. MCI and FCI definitions and calibration
- MCI (Monetary Conditions Index): weighted average of a kuna-euro nominal exchange rate index and a money market interest rate index (both normalized: 2002 average = 100).  
- FCI (Financial Conditions Index): weighted average of a trade-weighted real effective exchange rate (REER, CPI-based) and a real effective lending rate (RELR). Both REER and RELR are seasonally adjusted (X-12 multiplicative) and normalized to 2002 average = 100. RELR is computed as a weighted average of interest rates on domestic and foreign borrowing (weight = share of external debt in total private sector debt), adjusted by CPI.  
- The key calibration parameter is the relative weight of the interest rate vs. the exchange rate in the MCI (the MCI ratio). Open economies receive relatively lower interest-rate weight in MCIs.

### C. Calibration result for Croatia
- Degree of openness for Croatia (average of exports and imports as percent of GDP) is about 52 percent.  
- Regression from Figure 1 (Interest Rate Weight in the MCI vs. Openness): y = -0.01x + 0.97, R2 = 0.88.  
- Implied interest rate weight for Croatia: 0.57 (an MCI ratio of 1.3:1). Interpretation: the effect on demand of a 1.3 percentage point exchange rate appreciation can be offset by a 1 percent decline in interest rates.

### D. Empirical behavior of MCI and FCI, 1997–2003
- The constructed MCI is strongly correlated with the money market interest rate, reflecting much lower volatility in the kuna-euro exchange rate.  
- Trend in MCI (Hodrick-Prescott filtered): monetary policy loosening during mid-2000 to mid-2002 and tightening since mid-2002.  
- FCI dynamics (1997–2003):  
  - Financial conditions loosened during 1997 to early 2001.  
  - Financial conditions tightened from mid-2001 to mid-2002.  
  - Since mid-2002 financial conditions were relatively stable, with a mild loosening in the second half of 2002, mild tightening in the first half of 2003, and another loosening at the end of 2003.  
- Comparison of MCI and FCI:  
  - MCI (and money market rate) exhibited much higher volatility than FCI.  
  - Correlation between MCI and FCI: +0.38 for 1997–2003; slightly negative for the last three years of the sample.

### E. Formal tests of transmission (MCI → FCI)
- Bivariate Granger causality tests of d(MCI) → d(FCI):  
  - Period 1997:1–2003:12: 3 lags = 1.785 (p = 0.079)*; 6 lags = 1.516 (p = 0.187); 9 lags = 2.154 (p = 0.040)**.  
  - Period 2000:1–2003:12: 3 lags = 0.198 (p = 0.897); 6 lags = 0.711 (p = 0.643); 9 lags = 0.970 (p = 0.489).  
  - Notes: */** denotes significance at 10/5/1 percent level, respectively. “d” stands for difference. Granger tests from FCI to MCI were insignificant at the 10 percent level for the reported lags.  
- Impulse-response analysis (VAR, 2000–2003) shows a weak response of the FCI to innovations in the MCI (responses plotted as Cholesky one s.d. innovations ± 2 s.e.), consistent with Granger results indicating weak and generally insignificant transmission since 2000.

### F. Synthesis and policy-relevant observations
- The MCI calibration (interest weight = 0.57, MCI ratio = 1.3:1) and empirical results indicate a limited role for conventional monetary policy in shaping overall financial conditions in Croatia under the observed regime and historical openness.  
- Monetary policy shows clear influence on money market rates but weak and uncertain influence on broader financial conditions and lending rates, with long lags.  
- The weakened transmission since 2000 and the limited co-movement between MCI and FCI imply that exchange-rate management and other financial sector factors can dominate the transmission mechanism.  
- Historical episodes (including credit controls in 2003) illustrate practical limitations of monetary instruments for aggregate demand management in the Croatian context.

*Source: The authors’ calculations, based on data from the CNB.*

### 14.      The calculated MCI and FCI are relatively robust to changes in the weights.

### _cr04251 - 14.      The calculated MCI and FCI are relatively robust to changes in the weights.

### Robustness of MCI and FCI
- The MCI and the FCI are only rough indicators of monetary and financial conditions due to model dependency, ignored dynamics, parameter inconstancy, and nonexogeneity of regressors.
- Results are not very sensitive to changes in the weights of the interest rate and exchange rate:
  - Varying the weight on interest rate from the chosen number of 0.57 to 0.50 or 0.40 creates indices with pair-wise correlations of about 0.99.
  - Conclusion: a somewhat higher weight on the exchange rate (e.g., if Croatia is more euroized) would not lead to substantially different conclusions about monetary conditions and policy stance.

*Source: The authors’ calculations, based on data from the CNB.*

### Interest rate channel of monetary transmission
- MCI and FCI are based on two traditional channels: interest rate channel and exchange rate channel.
- In Croatia’s pegged exchange rate regime, a substantial part of transmission works through changes in money market rates (money market rates used as a proxy for monetary policy stance).
- The section examines transmission from policy rates to lending rates; the next section examines other channels, particularly the credit channel.

### Money market rates and excess liquidity (daily data, Jan 2001–Feb 2004)
- Excess liquidity (daily deviation from a 30-day moving average) is a leading indicator for money market interest rates; excess liquidity is negatively linked to money market rates.
- Correlation coefficients (liquidity and interest rate based on daily data), Zagreb Interbank Offer Rate (ZIBOR):
  - Excess liquidity defined as daily minus 30-day MA:
    - Overnight: -0.20
    - 1W: -0.20
    - 2W: -0.18
    - 1M: -0.15
    - 3M: -0.12
    - 6M: -0.23
  - Excess liquidity defined as 30-day MA minus 12-month MA:
    - Overnight: -0.16
    - 1W: -0.20
    - 2W: -0.24
    - 1M: -0.33
    - 3M: -0.35
    - 6M: -0.34
- Note: “W” stands for week, “M” stands for month, and “MA” stands for moving average. Excess liquidity is defined as departure from a trend of currency in circulation plus government deposits.

*Source: The authors’ calculations, based on data from the CNB.*

### Transmission from money market rates to bank lending rates
- Empirical finding: transmission from money market rates to bank lending rates is weak.
- Observed episode: money market rates rose during 2003 in response to hikes in the kuna portion of required reserves (from 25 to 42 percent), but the effective lending rate of domestic banks remained largely unaffected and so did the spread between this and foreign interest rates.
- Granger causality tests (Table 3) — F-statistics with p-values in parentheses; sample periods and lags:
  - money market rate → lending rate, 1997:1–2003:12:
    - 3 lags: 2.411 (0.074)*
    - 6 lags: 1.907 (0.093)*
    - 9 lags: 2.831 (0.008)***
  - d(money market rate) → d(lending rate), 1997:1–2003:12:
    - 3 lags: 1.893 (0.138)
    - 6 lags: 1.632 (0.153)
    - 9 lags: 2.189 (0.037)**
  - money market rate → lending rate, 2001:1–2003:12:
    - 3 lags: 0.219 (0.882)
    - 6 lags: 0.847 (0.545)
    - 9 lags: 1.038 (0.440)
  - Note: */**/*** denotes significance at 10/5/1 percent level, respectively. “d” stands for difference.
- Interpretation:
  - There appears to be a link in 1997–2003, but this includes the period before full capital flow liberalization in 2001. The link disappears if only 2001–2003 is considered.
- VAR model results:
  - For the Euro Area, Angeloni and Ehrmann (2003) estimate maximum impact on lending rates in about 5 months, maximum impact about 0.8 of the original shock to money market rates.
  - For Croatia, the VAR suggests that in 5 months the impact of money market rates on lending rates is about 0.15 of the original shock.
  - If impulse-response functions are re-estimated for the period since 2001, the response of domestic lending rates is virtually zero.
  - One st. dev. innovation equals about 3.7 percentage point increase in money market rates (note in Figure 7).

*Source: The authors’ calculations, based on data from the CNB.*

### Short-term capital flows and interest rate differentials
- Correlation between interest rate differentials and short-term capital flows (1997–2003) is weak and insignificant:
  - Sample 1997:1–2003:4: 0.11
  - Sample 1997:1–2001:2: 0.03
  - Sample 2001:3–2003:4: 0.24
- Note: interest rate differentials defined as difference between domestic lending rate and euro-area lending rate.
- Interpretation:
  - Correlations near nil before removal of restrictions in 2001; about 1/4 since then (insignificant at 10 percent level).
  - Suggests the interest rate channel may have weakened further; more data needed to verify.

*Source: The authors’ calculations, based on data from the CNB.*

### Other monetary transmission channels
- Channels that may operate besides interest rate and exchange rate:
  - Equity price channel (valuation changes affecting investment and consumption).
  - Credit channel (reduction in supply of bank credit).
  - Balance sheet channel (external finance premium depends on borrowers’ financial position; e.g., monetary restriction → decline in real estate prices → lower collateral → reduced demand for credit).
- Relevance for Croatia:
  - Equity price channel likely limited given small size of the equity market.
  - Real estate market plays a more important role, but real estate prices have so far been driven largely by non-monetary factors.
  - Credit channel could be substantial given banks’ dominant role, but only if banks cannot react to restrictive monetary policy by finding funding abroad.
  - Preliminary bank-by-bank calculations suggest banks—especially foreign-owned banks—have capability to find funding sources abroad; therefore, the credit channel does not seem very substantial, though further research is warranted.

*Source: The authors’ calculations, based on data from the CNB.*

### Experience with credit controls (2003) and effects on credit and financial soundness
- CNB introduced credit controls in January 2003 in response to booming credit and mounting external imbalance; controls abolished at end-2003.
- Effects on credit growth and sectoral borrowing:
  - Bank credit decelerated in 2003; possible impact on household spending (households less able to access foreign borrowing).
  - Enterprises were able to switch borrowing to foreign banks and use leasing and other financing.
  - External borrowing in 2003 was about 2.5 times higher than in 2002; the share of external debt in financing corporate investment rose in 2003.
  - After abolition of controls at beginning of 2004, credit growth did not bounce back: seasonally adjusted credit growth for first five months of 2004 (after adjusting for exchange rate movements) indicates an annualized rate of 14½ percent, about the same as in 2003—suggesting the deceleration since spring 2003 was largely demand-driven.
- Table 5 — Croatia: Credit Growth Before and After the Controls, 2000–03 (selected entries):
  - Total lending (% change y/y) 1/: 2000: 5.7; 2001: 20.8; 2002: 24.6; 2003: 20.3
  - Domestic bank borrowing (contribution in % points) 2000: 5.7; 2001: 15.9; 2002: 20.8; 2003: 10.7
  - Foreign borrowing (without leasing) 2000: 5.6; 2001: -3.8; 2002: 2.3; 2003: 5.1
  - Leasing 2000: 0.2; 2001: 2.6; 2002: 1.9; 2003: 3.8
  - Adjustment for write-offs 2000: -5.9; 2001: 6.1; 2002: -0.5; 2003: 0.6
  - Corporate sector lending (% change y/y) 2000: 3.9; 2001: 14.0; 2002: 17.4; 2003: 16.2
    - Domestic bank borrowing contribution 2000: 1.0; 2001: 10.7; 2002: 11.9; 2003: 2.4
    - Foreign borrowing (without leasing) 2000: 7.3; 2001: -5.1; 2002: 3.2; 2003: 7.7
  - Household sector lending (% change y/y) 2000: 11.4; 2001: 41.7; 2002: 42.1; 2003: 28.5
    - Domestic bank borrowing contribution 2000: 21.0; 2001: 31.8; 2002: 42.5; 2003: 27.6
    - Foreign borrowing (without leasing) 2000: 0.2; 2001: 0.1; 2002: 0.2; 2003: 0.0
  - Memorandum: Composition of corporate financing (in %), Domestic borrowing (net flow): 2000: 2.8; 2001: 25.8; 2002: 24.4; 2003: 4.8
  - Memorandum: Foreign borrowing (net flow): 2000: 20.2; 2001: -4.1; 2002: 12.1; 2003: 27.1
  - Memorandum: Other (e.g., reinvested profits): 2000: 77.0; 2001: 78.3; 2002: 63.5; 2003: 68.1
  - Note: 1/ Total domestic and external borrowing by the non-government sector in Croatia.
- Limits’ impact on financial sector soundness:
  - Some of domestic banks’ best corporate clients were redirected to foreign banks.
  - Limits encouraged rapid growth of unsupervised and unregulated leasing companies; their contribution to overall lending growth increased in 2003.
  - Transparency of monetary and banking statistics deteriorated as banks engaged in activities to circumvent limits (asset swaps, collateralization, accelerated write-offs of nonperforming loans).
  - Table 5 calculations attempt to approximate impact of write-offs on total credit data.

*Source: Authors' calculations based on data from CNB and the Central Statistics Bureau.*

### Transition to determinants of lending rates and spreads (Chapter II introduction)
- Understanding determinants of banks’ lending rates is crucial for macroeconomic and financial surveillance.
- Rapid increase in bank credit to the private sector in Croatia was driven by perceived profit opportunities in lending (high lending interest rates and relatively low default rates).
- Interest rate spreads in Croatia:
  - Have been declining but remain somewhat higher than in other Central, Eastern European, and Balkan (CEEB) countries.
  - Lower spreads generally observed for foreign greenfield banks, banks with lower share of NPLs, banks that are more liquid and have higher capital adequacy, and in large banks (though some evidence that spreads increase with market share).
- Policy implications highlighted:
  - Impact from NPLs underscores need for credit information sharing and strong prudential regulations.
  - Competition policy should enforce competition laws while permitting banks to take advantage of economies of scale.

*Source: Prepared by Martin Čihák, with inputs from Tomislav Galac and Danijela Mladinovič (as presented in the supplied content).*

### 28.      Similarly to other CEEB

### _cr04251 - 28.      Similarly to other CEEB

### Interest rate spreads: overview and cross-country context
- Interest rate spreads in Croatia are higher than in advanced economies but have declined since the 1990s toward advanced-economy levels.
- Throughout the adjustment period, interest rate spreads in Croatia have been about 2–4 percentage points above the CEEB average.
- Interest rate spread definition: the difference between average interest rates on non-bank client loans and deposits. For Croatia, pre-2002 figures are the author’s estimates to adjust for a methodological break in 2002.
- Observed within-system variability (September 2003 example):
  - Local currency deposit rates varied from 0.4 to 5.9 percent.
  - Lending rates ranged from 6.1 to 26.8 percent.
  - Spreads varied from 4.7 to 21.0 percent.

### Decomposition of interest rate spreads (accounting identity)
- Decomposition identity (preserved exactly as presented):
  iL – iD ≡ o + l + iD r/(1 – r) + d + p + τ [p – o  − l  − iD r/(1 – r)– d] ,      (1)
  where:
  - iL = average interest rate on loans
  - iD = average interest rate on deposits
  - o = overhead costs relative to loans
  - l = loan loss provisions net of recoveries relative to loans
  - r = reserve requirement rate
  - d = deposit protection premium relative to loans
  - p = profit margin (pre-tax profit relative to loans)
  - τ = tax rate on profits
- Profit margins are calculated as a residual after accounting for the other components.

### Results from the accounting decomposition (aggregate sector, 2001–2003)
- Data source: aggregate balance sheets and income statements for the Croatian banking sector, 2001–2003.
- Key observations (2003 and changes 2001–2003):
  - Overhead costs
    - About 2.5 percentage points of the spread in 2003.
    - Represented more than a third of the spread in 2003.
    - Above advanced-economy average of about 2 percentage points and somewhat above emerging market average.
    - Ratio of overhead costs to loans declined by about 1 percentage point between 2001 and 2003 due to cost-cutting and rapid credit growth.
  - Loan loss provisioning
    - Accounted for 0.4 percentage points of the spread in 2003.
    - Reflects relatively low percentage of nonperforming loans (NPLs).
  - Reserve requirements
    - Impact on spreads about 0.2 percentage points in 2003.
    - Reflects low interest rates on deposits and positive remuneration rate (remuneration set at 25 basis points below the EURIBOR rate).
  - Deposit protection premia
    - About 0.4 percentage points of the spread in 2003.
    - Relatively high compared with other countries; reflects high cost of past bank failures.
  - Profit margins
    - Pre-tax profit on enterprise lending in 2003 was substantially lower than the average return on banks’ assets, making further enterprise lending relatively unattractive.
    - Pre-tax return on household lending was more than 4 times the average pre-tax return on assets.
  - Foreign greenfield banks
    - Have somewhat lower interest rate spreads and lower profit margins, likely reflecting specialization on lower risk corporate clients.

- Selected numeric items from Table 1 (Croatia: Decomposition of Interest Rate Spreads, 2001–03, in percentage points):
  - Interest rate spread (All loans): 2001 = 9.6 ; 2003 = 10.1
  - Overhead costs (All loans): 2001 = 3.5 ; 2003 = 2.5
  - Loan loss provisioning (All loans): 2001 = 0.1 ; 2003 = 0.6
  - Reserve requirement (All loans): 2001 = 0.0 ; 2003 = 0.3
  - Deposit protection (All loans): 2001 = 0.5 ; 2003 = 0.4
  - Pre-tax profit (All loans): 2001 = 5.6 ; 2003 = 6.6
  - Tax (All loans): 2001 = 0.2 ; 2003 = 0.2
  - Profit margin (All loans): 2001 = 5.4 ; 2003 = 6.5
  - Memorandum item: ROA (after-tax) = 1.3

- Bank productivity (Table 2, 2002, in thousands of U.S. dollar):
  - Assets per Employee
    - Croatia: 1,413
    - Emerging markets: 2,040
  - Loans per Employee
    - Croatia: 762
    - Emerging markets: 910
  - Deposits per Employee
    - Croatia: 1,002
    - Emerging markets: 1,620
  - Net Interest Income per Employee
    - Croatia: 24
    - Emerging markets: 60

### Bank-by-bank regressions: methodology and main empirical findings
- Data and sample:
  - Panel of monthly data for 46 Croatian banks from July 1999 to December 2003 (roughly 2,100 observations in 54 cross-sections).
  - Two estimation methods used to address heteroscedasticity: seemingly unrelated regressions (SUR) and feasible generalized least squares assuming cross-section heteroscedasticity (GLS).
- Overall explanatory power:
  - The estimates account for about 90 percent of bank-by-bank variability in lending spreads.
- Main empirical findings from regressions (select results and numeric magnitudes preserved where given):
  - Bank size
    - Larger banks generally have lower lending rates and lending spreads, reflecting economies of scale.
    - If a bank’s balance sheet doubles, estimates suggest lending rates are typically lower by about 0.3–0.6 percentage points (derived using log(2) ≈ 0.69 and coefficient scaling).
  - Market share
    - Some evidence that lending rates and spreads increase with market share, indicating potential market power; coefficient significance varies across regressions.
  - Asset quality (NPLs)
    - Lending rates and spreads increase with the NPL ratio.
  - Liquidity
    - Banks with more liquidity have lower lending rates and spreads; the impact is somewhat smaller than that of NPLs but generally negative.
  - Capital adequacy
    - Banks with higher capital adequacy have lower lending rates, but even lower deposit rates, resulting in higher spreads compared with banks with lower capital adequacy.
  - Ownership
    - Foreign greenfield banks have generally lower interest rate spreads than domestic banks (formal panel regression confirms earlier literature).
  - Credit growth determinants
    - Credit growth is lower for banks with higher lending rates.
    - Credit growth is lower for larger banks and for banks with higher NPL ratios.
    - Significant relationships exist between credit growth and bank liquidity and capital adequacy.
    - Greenfield banks have been growing faster than other banks, controlling for other factors.
    - Higher treasury-bill rates are related to lower credit growth.
  - Interest rate transmission
    - Changes in lending rates are positively related to changes in money market and deposit rates.
    - Transmission operates with lags, reflected in a significant AR(1) coefficient.
- Regression sample/statistics and selected coefficients (Table 3 highlights):
  - Number of observations: 2,140 (or 2,144 in some specifications); No. of obs. listed as 2,140 / 2,140 / 2,109 / 2,109 / 2,144 / 2,144 / 2,144 for different specs.
  - R2 (weighted) reported as 0.91, 0.94, 0.86, 0.99 in various specifications; R2 unweighted reported as 0.76, 0.76, 0.75, 0.69, 0.71, 0.70, 0.03 across specifications.
  - Selected estimated coefficients (with significance):
    - Log(assets): coefficients include -0.84**, -0.86**, -0.68** in several specifications.
    - NPLs in total loans: coefficients include 5.78**, 2.76**, 6.17**, 5.39**, 3.97**, 4.21**, -5.97** (the negative appears in the credit growth regression).
    - Greenfield bank: -3.02**, -2.18**, -0.60**, -0.53**, -3.38**, -2.31, 1.53** (varies by dependent variable).
    - AR(1): 0.85**, 0.82**, 0.80**, 0.55**, 0.82**, 0.79**, 0.11** across specifications.
    - Domestic T-bill rate: 0.11**, 0.08**, 0.07**, 0.12**, 0.12**, -0.12** in different specifications.
  - Significance notation used in Table 3:
    - * significant at 1 percent level
    - ** significant at 5 percent level
    - + significant at 10 percent level

### Policy implications
- Credit information and supervision
  - Banks with high NPLs have higher lending rates, underscoring the need for improved credit information sharing and stronger prudential supervision to improve asset quality.
  - Banks with low liquidity and low capital adequacy have higher lending rates, reinforcing the prudential supervision recommendation.
- Competition policy
  - Significant economies of scale imply that growth in bank size is likely to reduce spreads through savings in overhead costs (overhead costs account for the largest part of spreads).
  - Competition policy that allows banks to grow, including through mergers and acquisitions, can help decrease lending rates and spreads.
  - Because banks with large market shares may exert market power and increase profit margins, permissive merger and acquisition policy should be accompanied by strict enforcement of rules against anti-competitive behavior (such as abuse of dominant position).
- Reserve requirements
  - The reserve requirement rate is high at 19 percent.
  - Lowering reserve requirements would not have a major impact on lending rates—only about 0.2 percentage points—because reserves are remunerated (remuneration rate set at 25 basis points below the EURIBOR rate).

*Source: _cr04251 - 28.      Similarly to other CEEB (IMF chapter/section content provided).*

### 35.      This chapter discusses the economic implications of employment protection in

### 35.      This chapter discusses the economic implications of employment protection in Croatia.

### B. Employment Protection and Labor Market Performance in Croatia: Stylized Facts
- Main interpretation:
  - Stringent employment protection increases the incidence of long-term unemployment and makes labor turnover difficult, a problem amplified by severe negative shocks such as the transition from a planned to a market economy and war experiences in the 1990s.
- Labor market outcomes and indicators:
  - Labor force survey-based unemployment rate: hovering around 15 percent for the past five years.
  - Share of long-term unemployment in total unemployment: hovering around 55 percent (compared with major CEECs averaging around 45 percent).
  - Labor force participation rate: remained low at around 50 percent.
  - Unemployment particularly high and participation low among the young.
- Job reallocation and firm-level dynamics:
  - Croatian firms yearly terminate about 5 percent of all jobs, compared with the job destruction rate of 10–11 percent in other CEECs.
  - Job creation rate in Croatia: only 3½ percent, compared with 7-11 percent in other CEECs.
  - These figures indicate sluggish job reallocation and less intensive enterprise restructuring relative to leading CEEC reformers.
- Labor costs, wages, and unemployment benefits:
  - Gross wages in manufacturing: higher than in most other CEECs (World Bank, 2003).
  - Economy-wide unit labor costs: Croatia has held a relatively strong position among CEECs in recent years.
  - Gross wage in relation to GDP per employee: Croatian workers are not overpaid compared with other CEECs.
  - Unemployment benefit system:
    - Flat rate benefit with fixed maximum amount only about one-fourth of the average wage.
    - Replacement ratio: comparison shows the replacement ratio in Croatia is relatively low.
    - Duration of benefit payment: capped at 312 days.
    - Benefit coverage rate: has been below 20 percent since the mid-1990s.
    - Reasons for low coverage: (i) high unemployment among new entrants who do not qualify; (ii) large proportion of long-term unemployed no longer eligible.
  - Conclusion on benefits: low replacement rate, moderate duration, and limited coverage imply labor supply disincentives related to unemployment benefits are likely modest in Croatia.
- Employment Protection Legislation (EPL) and business environment:
  - Employment protection in Croatia is among the strictest in CEECs; EPL index values show Croatia even stricter than most of EU-15 and other CEECs (Table 1).
  - Individual dismissals costly due to long advanced notice period and high severance pay; collective dismissals difficult due to overly inclusive definition of collective redundancy.
  - Fixed-term employment was restricted until recently, limiting a channel to circumvent high termination costs.
  - Croatia ranks poorly on “cooperation in labor-employer relations” in World Economic Forum’s “Quality of the National Business Environment Rank”.
  - Share of employment by SMEs (proxy for new businesses): 46 percent in Croatia, compared to well over 50 percent in major CEECs.
- Labor law amendment (July 2003; effective beginning of 2004):
  - Lowered Croatia’s EPL index by 23 percent.
  - Main changes:
    - Relaxing restrictions on the use of fixed-term contracts.
    - Easing the pre-conditions for valid dismissals.
    - Shortening the advanced notice period from 6 to 3 months.
    - Reducing the amount of severance pay from half to one-third of the monthly pay.
    - Relaxing the definition of mass lay-offs.
- Policy implication from stylized facts:
  - Strict employment protection is likely to have discouraged entry or expansion of new businesses and contributed to poor labor market performance; relaxing EPL is an important step to improve flexibility, employment in the official sector, and SME-driven job creation.

### C. Employment Protection and the Shadow Economy
- Conceptual links:
  - Stringent employment protection increases labor costs in the official economy and provides incentives for firms to move to or remain in the informal sector to lower labor costs.
  - Labor costs shifted onto employees can also provide workers with incentives to work in the shadow economy.
  - Empirical evidence (Schneider and Pöll, 1999) supports these mechanisms in firm-level data for Germany.
- Cross-country correlation:
  - A positive correlation exists between strictness of employment protection (EPL index) and the size of the shadow economy (percent of GDP) across 20 OECD countries and 7 CEECs (sample limited to 27 countries because EPL index exists for a smaller set).
  - Shadow economy estimates used: Schneider (2002) averages (2000/01 for 22 transition economies and 2001/02 for 21 OECD countries).
- Other determinants of the shadow economy:
  - Tax and social security burden (tax wedge on labor income) is commonly cited as a major factor increasing incentives to work in the shadow economy.
  - Business regulations also affect the size of the shadow economy.
  - Higher per capita income is associated with weaker incentives to work in the unofficial sector due to better infrastructure and public services.
- OLS regression results (Table 2) on determinants of the shadow economy (Dependent variable: Shadow Economy as a percent of GDP; Number of Observations=27; R-Bar-Squared=0.607):
  - Regressors and reported estimates:
    - Constant: Coefficient 56.814; Standard Error 16.338; T-Ratio 3.477; Prob. [0.002] (*** indicates significance at 1 percent level)
    - Log GDP per capita: Coefficient -6.017; Standard Error 1.443; T-Ratio -4.171; Prob. [0.000] (***)
    - EPL: Coefficient 4.234; Standard Error 1.387; T-Ratio 3.052; Prob. [0.006] (***)
    - Tax wedge on labor income: Coefficient 0.046; Standard Error 0.099; T-Ratio 0.466; Prob. [0.646]
    - Business regulations: Coefficient 1.477; Standard Error 1.294; T-Ratio 1.142; Prob. [0.266]
  - Interpretation:
    - Per capita income coefficient is significantly negative (higher income → smaller shadow economy).
    - EPL coefficient is positive and highly significant: less employment protection is correlated with a lower share of the shadow economy even after controlling for other factors.
    - Contrary to many studies, in this specification neither the tax wedge on labor income nor the business regulation index is statistically significant.
- Conclusions and policy implications:
  - Empirical tests indicate employment protection is correlated with the size of the shadow economy across countries.
  - Strictness of employment protection plays a more important role in explaining cross-country differences in the size of the shadow economy than the tax burden on labor income or business regulations in this analysis.
  - Policy implication: relaxing employment protection can help enhance employment in the official sector, expand the tax base, and boost productivity by reducing incentives to operate in the shadow economy; the July 2003 labor law amendments (effective beginning of 2004) that lowered Croatia’s EPL index by 23 percent are an important step in this direction.

*https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04251.pdf*

### REFERENCES

### _cr04251 - REFERENCES

### References (Selected bibliographic entries)
- Blanchard, Olivier, 2000, “The Economics of Unemployment: Shocks, Institutions, and Interactions,” Lionel Robbins Lectures 1-3, Unpublished, London School of Economics.
- Farrell, Diana, 2004, “The Hidden Danger of the Informal Economy,” McKinsey Quarterly, 2004, No. 3.
- Feige, Edgar and Urban, Ivica, 2003, “Estimating the Size and Growth of Unrecorded Economic Activity in Transition Countries: A Re-evaluation of Electric Consumption Method Estimates and their Implications,” Institute of Public Finance, Zagreb.
- Giles, David, 1999, “Measuring the Hidden Economy: Implications for Econometric Modeling,” Economic Journal, Vol 109, No. 456, pp. 370-380.
- Johnson, Simon, Kaufmann, Daniel, and Zoido-Lobaton, Pablo, 1998, “Corruption, Public Finances and the Unofficial Economy,” World Bank, Mimeo.
- Jurajda, Štěpán, and Katherine Terrell, 2002, “What Drives the Speed of Job Reallocation During Episodes of Massive adjustment?” Discussion Paper No. 601, (Germany: IZA, Bonn, October).
- Loayza, Norman, 1996, “The Economics of the Informal Sector: A Simple Model and Some Empirical Evidence from Latin America,” Carnegie-Rochester Conference Series Public Policy, 45, pp. 129-162.
- Rutkowski, Jan, 2003, “Does Strict Employment Protection Discourage Job Creation?: Evidence from Croatia,” World Bank, Mimeo.
- Schneider, Friedrich, 2002, “The Size and Development of the Shadow Economies and Shadow Economy Labor Force of 22 Transition and 21 OECD Countries: What do we really know?,” Institute of Public Finance, Zagreb.
- Schneider, Friedrich and Enste, Dominik, 2000, “Shadow Economies: Size, Causes, and Consequences,” Journal of Economic Literature, Vol 108, pp. 77-114.
- Tanzi, Vito, 1999, “Uses and Abuses of Estimates of the Underground Economy,” Economic Journal, Vol 109, No. 456, pp. 338-340.
- Thomas, James, 1999, “Quantifying the Black Economy: ‘Measurement without Theory’ Yet Again?,” Economic Journal, Vol 109, No. 456, pp. 381-389.
- Tokman, Victor, 1990, “The Informal Sector in Latin America: 15 Years Later,” In The Informal Sector Revised, Center Seminars, OECD, Paris.
- World Bank, 2002, “Transition: The First Ten Years. Analysis and Lessons from Eastern Europe and the Former Soviet Union,” Washington, D.C.

### Statistical Appendix — Key tables and exact figures (selected)
- Table A1. Croatia: GDP by Expenditure Category, 2000-05 (Percentage changes / contributions / implicit deflators)
  - Real GDP: 2000 2.9, 2001 4.4, 2002 5.2, 2003 4.3, 2004 3.7, 2005 Proj. 4.1
  - Domestic demand: 2000 -0.3, 2001 5.4, 2002 10.6, 2003 2.0, 2004 3.7, 2005 Proj. 3.2
  - Consumption: 2000 3.6, 2001 2.0, 2002 5.4, 2003 3.2, 2004 2.9, 2005 Proj. 3.3
    - Private 1/: 2000 4.2, 2001 4.4, 2002 7.5, 2003 4.1, 2004 3.5, 2005 Proj. 4.5
    - Government 2/: 2000 2/2.0, 2001 -4.2, 2002 -0.3, 2003 -0.4, 2004 0.6, 2005 Proj. -1.4
  - Gross fixed capital formation: 2000 -3.8, 2001 7.1, 2002 12.0, 2003 16.8, 2004 5.8, 2005 Proj. 5.4
    - Private 1/: 2000 1/10.0, 2001 5.3, 2002 12.4, 2003 11.5, 2004 8.6, 2005 Proj. 8.7
    - Government 2/: 2000 2/-32.9, 2001 13.1, 2002 10.9, 2003 42.2, 2004 -4.9, 2005 Proj. -8.6
  - Exports: 2000 12.0, 2001 8.6, 2002 3.1, 2003 18.2, 2004 3.9, 2005 Proj. 4.5
  - Imports: 2000 3.7, 2001 10.1, 2002 13.9, 2003 11.0, 2004 3.8, 2005 Proj. 2.8
  - Nominal GDP: 2000 152,519; 2001 165,639; 2002 179,390; 2003 193,067; 2004 205,747; 2005 Proj. 220,636
  - Notes: 1/ Includes public enterprises. 2/ Due to the switch from GFS1986 to GFS2001, there is a break in series between 2002 and 2003. 3/ Includes statistical discrepancy.
  - Sources: Croatian National Bank, Ministry of Finance, Central Statistics Bureau, and staff estimates

- Table A2. Croatia: Trends in Industrial Production, 1996-2004 (2000=100)
  - Total Industry index: 1996 90.0; 1997 96.2; 1998 99.7; 1999 98.3; 2000 100.0; 2001 105.8; 2002 109.4; 2003 114.8
  - Goods, Except Capital: 1996 77.3; 1997 90.5; 1998 93.8; 1999 102.5; 2000 100.0; 2001 100.7; 2002 99.9; 2003 104.3
  - Consumer Durable: 1996 95.6; 1997 98.3; 1998 97.4; 1999 95.6; 2000 100.0; 2001 104.7; 2002 110.4; 2003 115.5
  - Source: Central Bureau of Statistics.

- Table A3. Croatia: Price Developments, 1996-2004 (Index, Dec. 1994=100; Rate of Growth Period / Previous Year)
  - Consumer Prices (annual indices): 1996 105.6; 1997 109.4; 1998 115.6; 1999 120.4; 2000 127.9; 2001 134.2; 2002 136.5; 2003 138.9
  - Producer Prices (annual indices): 1996 101.5; 1997 104.0; 1998 102.8; 1999 105.5; 2000 115.7; 2001 119.8; 2002 119.4; 2003 121.6
  - Selected monthly Consumer Price entries (2001–2004): 2001 Jan 132.1; Feb 132.8; Mar 132.5; Apr 134.4; May 135.6; Jun 135.3; Jul 134.0; Aug 134.9; Sep 135.2; Oct 134.5; Nov 134.4; Dec 134.9. 2004 Jan 141.1; Feb 140.9; Mar 141.0; Apr 141.3; May 142.4.
  - Source: Croatian National Bank.

- Table A4. Croatia: Indices of Real Net Wages and Salaries Per Employee, 1999-2003 (1997=100)
  - Total: 1999 116.6; 2000 117.0; 2001 118.9; 2002 120.4; 2003 122.7
  - Manufacturing (D): 1999 109.3; 2000 113.3; 2001 119.5; 2002 122.1; 2003 125.1
  - Transport, storage and communication (I): 1999 115.9; 2000 118.9; 2001 123.8; 2002 128.4; 2003 133.0
  - Source: Central Bureau of Statistics.

- Table A5. Croatia: Composition of Employment, 1999-2003 (In thousands)
  - Total employed: 1999 1,349; 2000 1,339; 2001 1,352; 2002 1,363; 2003 1,398
  - Manufacturing (D): 1999 264; 2000 256; 2001 252; 2002 247; 2003 248
  - Wholesale and retail trade (G): 1999 153; 2000 154; 2001 159; 2002 165; 2003 177
  - Notes: 1/ Includes active insured persons - private farmers measured mid-year. For 1999 and 2000 data are measured by end-year. 2/ Refers to persons employed in crafts and trades as well as free-lancers during mid-year. For 1999 and 2000 data are measured by end-year.
  - Source: Central Bureau of Statistics.

- Table A6. Croatia: Exports by Destination, 1996-2003 (In millions of U.S. dollars)
  - Total exports: 1996 4,512; 1997 4,171; 1998 4,541; 1999 4,302; 2000 4,432; 2001 4,666; 2002 4,904; 2003 6,164
  - EU countries: 1996 2,303; 1997 2,074; 1998 2,161; 1999 2,110; 2000 2,416; 2001 2,526; 2002 2,584; 2003 3,363
  - Sources: Central Bureau of Statistics and the Fund staff estimates.
  - Notes: 1/ Data have not been revised in line with the 1998 balance of payments compilation methodology. 2/ Countries of the former USSR includes 14 countries. It does not include Belarus. 3/ Developing Middle East countries refer to the OPEC countries excluding Indonesia amd Venezuela.

- Table A7. Croatia: Tourism—Overnight Stays, 1996-2004 (In thousands)
  - Total overnight stays: 1996 21,455; 1997 30,314; 1998 31,287; 1999 26,563; 2000 38,406; 2001 43,404; 2002 44,692; 2003 46,647
  - 2003 monthly examples: Jul 13,022 (Total), Domestic 1,242, Foreign 11,780
  - Source: Central Bureau of Statistics.

- Table A8. Croatia: Imports by Origin, 1996-2003 (In millions of U.S. dollars)
  - Total imports: 1996 7,788; 1997 9,104; 1998 8,383; 1999 7,799; 2000 7,887; 2001 9,147; 2002 10,722; 2003 14,199
  - EU countries: 1996 4,625; 1997 5,412; 1998 4,980; 1999 4,415; 2000 4,368; 2001 5,224; 2002 5,984; 2003 8,032
  - Source: Central Bureau of Statistics.
  - Note: 1/ Data have not been revised in line with the 1998 balance of payments compilation methodology.

- Table A9. Croatia: External Debt, 1996-2004 (In millions of U.S. dollars, unless otherwise stated)
  - Portfolio Investments (total): 1996 1,462; 1997 1,955; 1998 2,057; 1999 2,571; 2000 3,180; 2001 3,732; 2002 4,525; 2003 6,124; 2004 Mar. 6,296
    - Bonds (subset): 1996 1,462; 1997 1,955; 1998 2,048; 1999 2,554; 2000 3,170; 2001 3,704; 2002 4,525; 2003 6,083; 2004 Mar. 6,252
  - Other Investments (total): 1996 3,845; 1997 5,497; 1998 7,626; 1999 7,407; 2000 7,875; 2001 7,585; 2002 10,897; 2003 17,445; 2004 Mar. 17,903
  - Total external debt (1+2): 1996 5,308; 1997 7,452; 1998 9,683; 1999 9,978; 2000 11,055; 2001 11,317; 2002 15,421; 2003 23,570; 2004 Mar. 24,199
  - Sources: Croatian National Bank; and Fund staff estimates.
  - Note: 1/ Excludes nonreported principal payments. Includes short-term credits and currency and deposits.

- Table A10. Croatia: Consolidated General Government Fiscal Operations by Economic Category, 1998-2003 (In percent of GDP, GFS 1986 basis)
  - Revenue and grants: 1998 51.1; 1999 48.4; 2000 46.2; 2001 44.0; 2002 44.5; 2003 44.3
  - Consolidated general government balance: 1998 -3.5; 1999 -8.2; 2000 -6.5; 2001 -6.7; 2002 -5.2; 2003 -6.3
  - Primary balance: 1998 -1.9; 1999 -6.5; 2000 -4.5; 2001 -4.5; 2002 -3.1; 2003 -4.2
  - Sources: Ministry of Finance and staff estimates.
  - Notes: 1/ On a GFS 1986 basis and with subnational government consisting of the 53 largest local governments. 2/ In 2000, includes 0.5 percent of GDP in back taxes.

- Table A11. Croatia: Consolidated General Government Financial Operations by Economic Category, 2002-03 (In percent of GDP, GFS 2001 basis)
  - PREL. REVENUE: 2002 46.3; 2003 46.4
  - Taxes: 2002 28.2; 2003 27.9
  - Social security contributions: 2002 14.0; 2003 14.2
  - EXPENSE: 2002 46.4; 2003 46.5
  - Compensation of employees: 2002 12.4; 2003 12.7
  - Acquisition of non-financial assets (investment): 2002 4.4; 2003 5.9
  - OVERALL BALANCE: 2002 -5.0; 2003 -6.3
  - Sources: Ministry of Finance and staff estimates.
  - Note: 1/ On a GFS 2001 basis. There may be differences from historical data, which were on a GFS 1986 basis.

- Table A12. Croatia: HBOR Operations by Economic Category, 1999-2003 (In percent of GDP, GFS 1986 basis)
  - Revenue and grants (1999–2003): 0.2, 0.3, 0.3, 0.3, 0.4
  - Expenditure and net lending (1999–2003): 0.5, 0.6, 0.9, 1.0, 1.4
  - Lending minus repayments: 1999 0.4; 2000 0.5; 2001 0.8; 2002 0.8; 2003 1.3
  - Balance: 1999 -0.3; 2000 -0.3; 2001 -0.6; 2002 -0.7; 2003 -1.1
  - Sources: Ministry of Finance, HBOR, and staff estimates.
  - Note: 1/ Unconsolidated before corrections for central budgetary transactions.

- Table A13. Croatia: Debt Stock of Consolidated General Government, 1997-2003 (In percent of GDP)
  - Debt Stock: 1997 26.7; 1998 25.7; 1999 33.0; 2000 39.3; 2001 40.1; 2002 39.8; 2003 41.5
  - Domestic / External split (selected years): 1997 Domestic 12.5; External 14.1. 2003 Domestic 17.1; External 24.3
  - Guarantees Stock: 1997 1.5; 1998 6.2; 1999 7.0; 2000 9.7; 2001 10.2; 2002 10.5; 2003 11.4
  - Total debt and contingent liabilities: 1997 31.6; 1998 36.9; 1999 45.7; 2000 50.4; 2001 50.8; 2002 50.7; 2003 53.2
  - Sources: Croatian Central Bank, Ministry of Finance, and staff estimates.

- Table A14. Croatia: Selected Public Enterprises, 2000-03 (In thousands of kuna unless otherwise specified)
  - Croatian Railways (Operating balance): 2000 -304, 2001 211, 2002 -426, 2003 025, 2003 -836, 2003 546, 2003 -779, 2003 568
    - Net indebtedness (selected): 2000 343, 2001 550, 2002 852, 2003 2,276, 2003 876, 2003 1,121, 2003 199, 2003 2,647
    - Number of employees (selected): 2000 191, 2001 821, 2002 842, 2003 816, 2003 345, 2003 153, 2003 75
  - Croatian Electricity Company (Operating balance): 2000 -439, 2001 660, 2002 -440, 2003 758, 2003 -123, 2003 138, 2003 920, 2003 9
    - Net indebtedness (selected): 2000 655, 2001 180, 2002 371, 2003 1,057, 2003 1,146, 2003 10,514, 2003 159, 2003 15
    - Number of employees (selected): 2000 1590, 2001 5158, 2002 4915, 2003 025, 2003 149, 2003 31
  - Total (aggregate lines): Operating balance (selected) -463, 363, 127, 453, 646, 467, 9....; In percent of GDP -0.3, 0.8, 0.3....; Net indebtedness (selected) 1,268, 651, 155, 542, 022, 393, 720....; In percent of GDP 0.8, 0.9, 1.3....
  - Sources: Ministry of Finance and staff calculations.
  - Note: Table entries include mixed formatting consistent with source.

- Central bank reserves and asset entries (selected series, 1996–2002 Mar.)
  - Assets: Reserves: 1996 4,410; 1997 5,046; 1998 5,908; 1999 8,988; 2000 10,589; 2001 15,003; 2002 20,373; 2002 Mar. 19,993; 2002 Jun. 21,492; 2002 Sep. 23,679; 2002 Dec. 26,784; 2003 Mar. 30,014
  - Reserves in f/c (foreign currency): 1996 4,410; 1997 5,046; 1998 4,240; 1999 4,353; 2000 5,098; 2001 9,306; 2002 13,340; 2002 Mar. 12,439; 2002 Jun. 13,516; 2002 Sep. 16,845; 2002 Dec. 20,103; 2003 Mar. 20,040
  - Reserves in kuna (selected): 1998 1,668; 1999 4,635; 2000 5,491; 2001 5,697; 2002 7,034; 2002 Mar. 7,555; 2002 Jun. 7,976; 2002 Sep. 6,834; 2002 Dec. 6,680; 2003 Mar. 9,975

*Source: _cr04251 - REFERENCES (IMF PDF content provided)*

### 2. Foreign assets

### 2. Foreign assets

### Foreign assets (levels, in millions of kuna)
- 12,550
- 16,186
- 12,763
- 12,400
- 19,710
- 32,808
- 25,978
- 26,951
- 27,101
- 31,466
- 35,383
- 35,176

### Claims on central government 2/
- Total claims on central government:
  - 16,693
  - 15,239
  - 14,864
  - 16,264
  - 19,055
  - 20,060
  - 21,918
  - 22,935
  - 23,243
  - 22,509
  - 21,544
  - 21,096
- 3.1 Bonds arising from blocked f/c deposits:
  - 8,291
  - 6,714
  - 5,802
  - 5,420
  - 4,484
  - 3,420
  - 2,473
  - 2,047
  - 1,999
  - 1,518
  - 1,532
  - 994
- 3.2 Big bonds:
  - 2,438
  - 2,292
  - 2,103
  - 1,322
  - 1,476
  - 1,659
  - ...
  - ...
  - ...
  - ...
  - ...
  - ...
- 3.3 Other claims:
  - 8,402
  - 8,524
  - 9,062
  - 10,845
  - 14,571
  - 16,640
  - 19,444
  - 20,888
  - 21,244
  - 20,991
  - 20,012
  - 20,102

### Claims on other domestic sectors
- Total claims on other domestic sectors:
  - 33,690
  - 48,592
  - 59,597
  - 55,400
  - 60,364
  - 74,284
  - 96,218
  - 101,018
  - 102,609
  - 105,689
  - 110,374
  - 111,940
- 4.1 Claims on local government:
  - 145
  - 309
  - 654
  - 906
  - 1,175
  - 1,280
  - 1,422
  - 1,307
  - 1,279
  - 1,274
  - 1,563
  - 1,580
- 4.2 Claims on enterprises:
  - 26,929
  - 35,487
  - 41,225
  - 35,244
  - 35,891
  - 42,882
  - 51,723
  - 53,023
  - 52,021
  - 52,172
  - 53,810
  - 54,823
- 4.3 Claims on households:
  - 6,615
  - 12,796
  - 17,717
  - 19,250
  - 23,298
  - 30,122
  - 43,073
  - 46,687
  - 49,309
  - 52,243
  - 55,001
  - 55,537

### Claims on other banking institutions
- -
- -
- 0
- 45
- 69
- 170
- 219
- 214
- 209
- 456
- 432
- 600

### Claims on other financial institutions
- 140
- 247
- 194
- 154
- 162
- 281
- 915
- 1,364
- 1,145
- 941
- 762
- 652

### Total assets (1+2+3+4+5+6)
- 67,483
- 85,309
- 93,326
- 93,251
- 109,949
- 142,606
- 165,622
- 172,475
- 175,799
- 184,739
- 195,278
- 199,479

### Liabilities

### Deposit and liability categories (levels, in millions of kuna)
- 1. Demand deposits:
  - 7,007
  - 8,424
  - 7,809
  - 7,891
  - 11,386
  - 15,181
  - 21,166
  - 19,973
  - 22,188
  - 22,071
  - 23,315
  - 21,559
- 2. Savings and time deposits:
  - 3,387
  - 5,599
  - 5,684
  - 5,398
  - 7,651
  - 10,213
  - 13,001
  - 15,057
  - 15,826
  - 18,263
  - 18,371
  - 19,678
- 3. Foreign currency deposits:
  - 21,817
  - 31,278
  - 37,971
  - 36,966
  - 46,902
  - 71,837
  - 72,055
  - 74,069
  - 71,104
  - 75,416
  - 76,035
  - 74,070
- 4. Bonds and money market instruments:
  - 128
  - 134
  - 154
  - 437
  - 478
  - 318
  - 216
  - 154
  - 263
  - 643
  - 598
  - 396
- 5. Foreign liabilities:
  - 12,467
  - 13,807
  - 16,177
  - 17,209
  - 17,810
  - 21,858
  - 35,023
  - 37,457
  - 40,456
  - 41,442
  - 49,932
  - 52,252
- 6. Central government and funds' deposits:
  - 1,721
  - 6,875
  - 7,298
  - 5,829
  - 6,730
  - 5,635
  - 6,095
  - 5,516
  - 5,253
  - 5,269
  - 5,283
  - 5,219
- 7. Credit from central bank:
  - 268
  - 34
  - 1,049
  - 1,139
  - 329
  - 17
  - 18
  - 14
  - 14
  - 343
  - 969
  - 14
- 8. Restricted and blocked deposits:
  - 8,224
  - 5,852
  - 4,196
  - 3,434
  - 2,550
  - 1,601
  - 1,680
  - 1,786
  - 1,858
  - 1,939
  - 1,709
  - 2,037
  - Of which: Households' blocked f/c deposits:
    - 7,171
    - 4,574
    - 3,419
    - 2,743
    - 1,695
    - 770
    - 319
    - 258
    - 242
    - 177
    - 168
    - 111
- 9. Capital accounts:
  - 15,441
  - 17,027
  - 19,786
  - 21,975
  - 24,953
  - 25,455
  - 26,323
  - 26,526
  - 26,120
  - 26,809
  - 27,390
  - 27,493
- 10. Other items (net):
  - -2,977
  - -3,720
  - -6,797
  - -7,026
  - -8,839
  - -9,508
  - -9,956
  - -8,076
  - -7,283
  - -7,456
  - -8,324
  - -3,239

### Total liabilities (1+2+3+4+5+6+7+8+9+10)
- 67,483
- 85,309
- 93,326
- 93,251
- 109,949
- 142,606
- 165,622
- 172,475
- 175,799
- 184,739
- 195,278
- 199,479

### Interest Rates (Table A16 excerpt)
- Table covers "Interest Rates on CNB Bills (35 days), Interest Rates on Credits in Kuna, Interest Rates on Deposits in Kuna, Interest Rates on Credits in Kuna Indexed to f/c, Interest Rates on Credits in f/c, Interest Rates on Deposits in Kuna Indexed to f/c, Interest Rates on Deposits in f/c" with values reported as monthly weighted average; in percent, annualized.
- Selected monthly entries (values in percent):
  - 1996 Dec: 8.01 8.46 18.97 19.50 4.15 9.46 5.09
  - 1997 Dec: 8.01 4.06 14.40 13.61 4.35 7.63 4.77
  - 1998 Dec: 9.51 6.06 13.04 16.95 4.11 7.47 3.98
  - 1999 Dec: 10.51 3.54 12.53 6.75 4.27 6.62 4.23
  - 2000 Jan: 10.51 5.32 12.76 6.65 4.32 4.02 4.18
  - 2000 Feb: 10.51 1.67 12.85 6.63 4.27 6.19 3.95
  - 2000 Mar: 10.41 2.94 12.17 6.93 4.10 6.81 3.96
  - 2000 Apr: 9.81 4.59 12.28 5.32 4.03 6.36 3.81
  - 2000 May: 9.11 2.52 12.18 6.98 3.91 6.00 3.83
  - 2000 Jun: 8.01 3.48 11.69 7.26 3.59 6.75 3.83
  - 2000 Jul: 7.81 1.46 11.30 5.72 3.34 6.40 3.78
  - 2000 Aug: 6.99 9.90 11.21 6.03 3.42 6.43 3.77
  - 2000 Sept: 6.81 0.73 11.64 6.53 3.47 6.67 3.59
  - 2000 Oct: 6.71 0.92 11.60 6.23 3.48 5.77 3.53
  - 2000 Nov: 6.71 0.90 11.34 6.57 3.57 5.64 3.51
  - 2000 Dec: 6.71 0.45 10.74 7.70 3.40 5.54 3.47
  - 2001 Jan: 6.61 0.81 10.26 7.83 3.45 5.19 3.13
  - 2001 Feb: 6.61 0.89 10.27 6.48 3.60 5.22 3.27
  - 2001 Mar: 6.6 8.98 9.82 6.80 3.60 5.64 3.26
  - 2001 Apr: 6.4 8.99 9.81 6.83 3.54 5.40 3.13
  - 2001 May: 6.3 9.32 10.34 7.15 3.32 5.94 3.09
  - 2001 Jun: 5.5 9.88 10.15 6.80 3.18 5.69 2.98
  - 2001 Jul: - 9.39 9.31 6.50 3.04 5.29 2.93
  - 2001 Aug: 4.1 9.27 9.64 6.51 3.11 4.63 2.96
  - 2001 Sept: 5.0 9.46 9.81 6.44 3.10 4.98 2.83
  - 2001 Oct: 5.0 8.53 9.37 5.93 3.06 4.58 2.75
  - 2001 Nov: 4.5 9.56 9.68 5.61 2.99 4.40 2.59
  - 2001 Dec: 3.4 9.51 9.29 5.94 2.76 4.58 2.60
  - 2002 Jan: 3.7 15.28 9.55 8.26 2.48 2.99 2.72
  - 2002 Feb: 3.4 14.28 9.28 7.76 2.32 3.32 2.62
  - 2002 Mar: - 13.47 9.21 6.20 2.02 2.89 2.62
  - 2002 Apr: 3.0 13.42 8.19 6.38 1.94 3.76 2.60
  - 2002 May: 2.7 13.44 8.63 7.48 1.97 2.78 2.57
  - 2002 Jun: 2.2 12.78 8.21 6.71 1.91 3.39 2.58
  - 2002 Jul: 1.9 11.89 8.12 6.48 1.75 3.59 2.59
  - 2002 Aug: 2.0 12.35 7.99 6.55 1.77 3.44 2.59
  - 2002 Sep: 2.0 11.81 8.54 5.79 1.71 3.39 2.56
  - 2002 Oct: 2.0 12.54 8.04 6.18 1.67 3.62 2.52
  - 2002 Nov: 2.1 11.91 8.29 6.46 1.58 3.58 2.54
  - 2002 Dec: 2.1 10.91 8.25 5.91 1.55 2.92 2.55
  - 2003 Jan: 2.1 11.26 8.09 6.19 1.61 3.61 2.54
  - 2003 Feb: 2.1 11.43 8.55 6.33 1.64 3.30 2.50
  - 2003 Mar: 2.2 11.30 8.41 5.70 1.44 3.61 2.37
  - 2003 Apr: 2.2 11.41 8.03 6.55 1.40 3.52 2.36
  - 2003 May: 2.3 11.58 8.07 4.60 1.35 2.98 2.26
  - 2003 Jun: 2.4 11.55 7.68 5.84 1.37 3.61 2.24
  - 2003 Jul: 2.5 11.15 8.05 4.74 1.36 3.25 2.22
  - 2003 Aug: 2.5 12.08 7.96 6.19 1.50 3.14 2.17
  - 2003 Sep: 2.6 11.71 8.12 4.77 1.79 3.40 2.22
  - 2003 Oct: - 12.00 8.09 5.73 1.70 3.37 2.20
  - 2003 Nov: - 12.00 7.39 5.58 1.50 3.25 2.14
  - 2003 Dec: - 11.45 7.07 5.62 1.66 3.48 2.22
  - 2004 Jan: - 12.12 7.00 5.26 1.77 3.67 2.46
  - 2004 Feb: - 12.42 7.99 6.00 5.26 1.89 3.78 2.38
  - 2004 Mar: - 11.76 7.48 4.58 1.98 3.70 2.38

*Source: Croatian National Bank; Central Bureau of Statistics.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/scr/2004/_cr04251.pdf_